{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.46498687574543207,
  "eval_steps": 500,
  "global_step": 150000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 2.3426266074180604,
      "epoch": 3.099912504969547e-05,
      "grad_norm": 15.919363975524902,
      "learning_rate": 4.359197907585005e-09,
      "loss": 0.904,
      "mean_token_accuracy": 0.7392466813325882,
      "num_tokens": 11974.0,
      "step": 10
    },
    {
      "entropy": 2.357245463132858,
      "epoch": 6.199825009939094e-05,
      "grad_norm": 17.161699295043945,
      "learning_rate": 9.202751138235009e-09,
      "loss": 0.9299,
      "mean_token_accuracy": 0.7459014117717743,
      "num_tokens": 23117.0,
      "step": 20
    },
    {
      "entropy": 2.2194840207695963,
      "epoch": 9.299737514908641e-05,
      "grad_norm": 5.266523361206055,
      "learning_rate": 1.4046304368885014e-08,
      "loss": 0.7805,
      "mean_token_accuracy": 0.7642201945185662,
      "num_tokens": 36481.0,
      "step": 30
    },
    {
      "entropy": 2.316211926937103,
      "epoch": 0.00012399650019878189,
      "grad_norm": 17.800691604614258,
      "learning_rate": 1.8889857599535022e-08,
      "loss": 0.901,
      "mean_token_accuracy": 0.756186394393444,
      "num_tokens": 48700.0,
      "step": 40
    },
    {
      "entropy": 2.3188992157578467,
      "epoch": 0.00015499562524847737,
      "grad_norm": 16.84282875061035,
      "learning_rate": 2.3733410830185025e-08,
      "loss": 0.8694,
      "mean_token_accuracy": 0.7459902837872505,
      "num_tokens": 60860.0,
      "step": 50
    },
    {
      "entropy": 2.334391838312149,
      "epoch": 0.00018599475029817283,
      "grad_norm": 17.42536735534668,
      "learning_rate": 2.857696406083503e-08,
      "loss": 0.9114,
      "mean_token_accuracy": 0.744971951842308,
      "num_tokens": 72100.0,
      "step": 60
    },
    {
      "entropy": 2.3774717271327974,
      "epoch": 0.00021699387534786831,
      "grad_norm": 13.821733474731445,
      "learning_rate": 3.3420517291485034e-08,
      "loss": 0.9343,
      "mean_token_accuracy": 0.7515697747468948,
      "num_tokens": 83144.0,
      "step": 70
    },
    {
      "entropy": 2.2282222718000413,
      "epoch": 0.00024799300039756377,
      "grad_norm": 16.813045501708984,
      "learning_rate": 3.8264070522135044e-08,
      "loss": 0.7473,
      "mean_token_accuracy": 0.7650388479232788,
      "num_tokens": 95878.0,
      "step": 80
    },
    {
      "entropy": 2.308245059847832,
      "epoch": 0.00027899212544725923,
      "grad_norm": 9.254444122314453,
      "learning_rate": 4.310762375278505e-08,
      "loss": 0.894,
      "mean_token_accuracy": 0.7462251231074333,
      "num_tokens": 107556.0,
      "step": 90
    },
    {
      "entropy": 2.1781405553221704,
      "epoch": 0.00030999125049695474,
      "grad_norm": 8.745959281921387,
      "learning_rate": 4.795117698343506e-08,
      "loss": 0.7802,
      "mean_token_accuracy": 0.7630234107375145,
      "num_tokens": 121234.0,
      "step": 100
    },
    {
      "entropy": 2.2521888092160225,
      "epoch": 0.0003409903755466502,
      "grad_norm": 9.263279914855957,
      "learning_rate": 5.2794730214085053e-08,
      "loss": 0.8459,
      "mean_token_accuracy": 0.7538067430257798,
      "num_tokens": 133561.0,
      "step": 110
    },
    {
      "entropy": 2.208202359080315,
      "epoch": 0.00037198950059634566,
      "grad_norm": 14.294042587280273,
      "learning_rate": 5.763828344473506e-08,
      "loss": 0.7828,
      "mean_token_accuracy": 0.7738959670066834,
      "num_tokens": 146427.0,
      "step": 120
    },
    {
      "entropy": 2.3306238532066343,
      "epoch": 0.0004029886256460411,
      "grad_norm": 18.315011978149414,
      "learning_rate": 6.248183667538507e-08,
      "loss": 0.935,
      "mean_token_accuracy": 0.7471725791692734,
      "num_tokens": 157320.0,
      "step": 130
    },
    {
      "entropy": 2.1904593795537948,
      "epoch": 0.00043398775069573663,
      "grad_norm": 9.24004077911377,
      "learning_rate": 6.732538990603508e-08,
      "loss": 0.7454,
      "mean_token_accuracy": 0.7584904819726944,
      "num_tokens": 170758.0,
      "step": 140
    },
    {
      "entropy": 2.2158657625317573,
      "epoch": 0.0004649868757454321,
      "grad_norm": 15.915936470031738,
      "learning_rate": 7.216894313668508e-08,
      "loss": 0.8056,
      "mean_token_accuracy": 0.7633484750986099,
      "num_tokens": 183462.0,
      "step": 150
    },
    {
      "entropy": 2.1613963916897774,
      "epoch": 0.0004959860007951275,
      "grad_norm": 16.494924545288086,
      "learning_rate": 7.701249636733508e-08,
      "loss": 0.7496,
      "mean_token_accuracy": 0.7714725866913795,
      "num_tokens": 196484.0,
      "step": 160
    },
    {
      "entropy": 2.2984951809048653,
      "epoch": 0.0005269851258448231,
      "grad_norm": 14.924416542053223,
      "learning_rate": 8.18560495979851e-08,
      "loss": 0.8421,
      "mean_token_accuracy": 0.749389934539795,
      "num_tokens": 208243.0,
      "step": 170
    },
    {
      "entropy": 2.2832500755786898,
      "epoch": 0.0005579842508945185,
      "grad_norm": 15.250982284545898,
      "learning_rate": 8.66996028286351e-08,
      "loss": 0.8231,
      "mean_token_accuracy": 0.7653956174850464,
      "num_tokens": 219830.0,
      "step": 180
    },
    {
      "entropy": 2.3017154157161714,
      "epoch": 0.000588983375944214,
      "grad_norm": 9.919282913208008,
      "learning_rate": 9.154315605928509e-08,
      "loss": 0.8631,
      "mean_token_accuracy": 0.7567846044898033,
      "num_tokens": 231756.0,
      "step": 190
    },
    {
      "entropy": 2.2674851924180985,
      "epoch": 0.0006199825009939095,
      "grad_norm": 15.856183052062988,
      "learning_rate": 9.63867092899351e-08,
      "loss": 0.8038,
      "mean_token_accuracy": 0.7690823331475258,
      "num_tokens": 244149.0,
      "step": 200
    },
    {
      "entropy": 2.2884296044707297,
      "epoch": 0.0006509816260436049,
      "grad_norm": 15.011128425598145,
      "learning_rate": 1.0123026252058511e-07,
      "loss": 0.8224,
      "mean_token_accuracy": 0.7501513630151748,
      "num_tokens": 256276.0,
      "step": 210
    },
    {
      "entropy": 2.3350342929363253,
      "epoch": 0.0006819807510933004,
      "grad_norm": 17.258758544921875,
      "learning_rate": 1.0607381575123511e-07,
      "loss": 0.904,
      "mean_token_accuracy": 0.7543755397200584,
      "num_tokens": 267504.0,
      "step": 220
    },
    {
      "entropy": 2.2519437611103057,
      "epoch": 0.0007129798761429959,
      "grad_norm": 13.8502197265625,
      "learning_rate": 1.1091736898188512e-07,
      "loss": 0.8167,
      "mean_token_accuracy": 0.7757848709821701,
      "num_tokens": 278715.0,
      "step": 230
    },
    {
      "entropy": 2.2798040971159934,
      "epoch": 0.0007439790011926913,
      "grad_norm": 18.540184020996094,
      "learning_rate": 1.1576092221253512e-07,
      "loss": 0.9195,
      "mean_token_accuracy": 0.7466915145516395,
      "num_tokens": 290352.0,
      "step": 240
    },
    {
      "entropy": 2.1401579171419143,
      "epoch": 0.0007749781262423868,
      "grad_norm": 13.905881881713867,
      "learning_rate": 1.2060447544318512e-07,
      "loss": 0.7208,
      "mean_token_accuracy": 0.7657153263688088,
      "num_tokens": 304153.0,
      "step": 250
    },
    {
      "entropy": 2.27926591783762,
      "epoch": 0.0008059772512920822,
      "grad_norm": 15.406006813049316,
      "learning_rate": 1.2544802867383514e-07,
      "loss": 0.8348,
      "mean_token_accuracy": 0.7559728816151619,
      "num_tokens": 316234.0,
      "step": 260
    },
    {
      "entropy": 2.285950776934624,
      "epoch": 0.0008369763763417777,
      "grad_norm": 7.619900703430176,
      "learning_rate": 1.3029158190448516e-07,
      "loss": 0.7548,
      "mean_token_accuracy": 0.7609011575579643,
      "num_tokens": 329019.0,
      "step": 270
    },
    {
      "entropy": 2.331369215250015,
      "epoch": 0.0008679755013914733,
      "grad_norm": 21.633079528808594,
      "learning_rate": 1.3513513513513515e-07,
      "loss": 0.8983,
      "mean_token_accuracy": 0.7470938101410866,
      "num_tokens": 340699.0,
      "step": 280
    },
    {
      "entropy": 2.3679069608449934,
      "epoch": 0.0008989746264411687,
      "grad_norm": 15.301081657409668,
      "learning_rate": 1.3997868836578514e-07,
      "loss": 0.8797,
      "mean_token_accuracy": 0.7614053919911384,
      "num_tokens": 351539.0,
      "step": 290
    },
    {
      "entropy": 2.349914161860943,
      "epoch": 0.0009299737514908642,
      "grad_norm": 13.113327026367188,
      "learning_rate": 1.4482224159643515e-07,
      "loss": 0.8534,
      "mean_token_accuracy": 0.7527757495641708,
      "num_tokens": 362660.0,
      "step": 300
    },
    {
      "entropy": 2.274288383126259,
      "epoch": 0.0009609728765405597,
      "grad_norm": 16.123380661010742,
      "learning_rate": 1.4966579482708517e-07,
      "loss": 0.8189,
      "mean_token_accuracy": 0.7668241068720818,
      "num_tokens": 374530.0,
      "step": 310
    },
    {
      "entropy": 2.2432350262999536,
      "epoch": 0.000991972001590255,
      "grad_norm": 9.562804222106934,
      "learning_rate": 1.5450934805773516e-07,
      "loss": 0.8106,
      "mean_token_accuracy": 0.7698193415999413,
      "num_tokens": 387392.0,
      "step": 320
    },
    {
      "entropy": 2.311431211233139,
      "epoch": 0.0010229711266399506,
      "grad_norm": 16.730819702148438,
      "learning_rate": 1.593529012883852e-07,
      "loss": 0.8178,
      "mean_token_accuracy": 0.7644891291856766,
      "num_tokens": 399180.0,
      "step": 330
    },
    {
      "entropy": 2.349755284190178,
      "epoch": 0.0010539702516896461,
      "grad_norm": 16.725875854492188,
      "learning_rate": 1.641964545190352e-07,
      "loss": 0.8588,
      "mean_token_accuracy": 0.7604990780353547,
      "num_tokens": 410702.0,
      "step": 340
    },
    {
      "entropy": 2.279722733795643,
      "epoch": 0.0010849693767393416,
      "grad_norm": 14.276802062988281,
      "learning_rate": 1.6904000774968518e-07,
      "loss": 0.7989,
      "mean_token_accuracy": 0.7695920899510383,
      "num_tokens": 422591.0,
      "step": 350
    },
    {
      "entropy": 2.1071667522192,
      "epoch": 0.001115968501789037,
      "grad_norm": 12.381896018981934,
      "learning_rate": 1.7388356098033517e-07,
      "loss": 0.6475,
      "mean_token_accuracy": 0.7926038712263107,
      "num_tokens": 436846.0,
      "step": 360
    },
    {
      "entropy": 2.3276678711175918,
      "epoch": 0.0011469676268387324,
      "grad_norm": 17.90573501586914,
      "learning_rate": 1.7872711421098519e-07,
      "loss": 0.8036,
      "mean_token_accuracy": 0.7576885908842087,
      "num_tokens": 448403.0,
      "step": 370
    },
    {
      "entropy": 2.2606518104672433,
      "epoch": 0.001177966751888428,
      "grad_norm": 12.439233779907227,
      "learning_rate": 1.8357066744163518e-07,
      "loss": 0.7721,
      "mean_token_accuracy": 0.7698314875364304,
      "num_tokens": 460244.0,
      "step": 380
    },
    {
      "entropy": 2.338596761226654,
      "epoch": 0.0012089658769381235,
      "grad_norm": 14.934176445007324,
      "learning_rate": 1.884142206722852e-07,
      "loss": 0.8393,
      "mean_token_accuracy": 0.7656854018568993,
      "num_tokens": 470678.0,
      "step": 390
    },
    {
      "entropy": 2.332528692483902,
      "epoch": 0.001239965001987819,
      "grad_norm": 16.811922073364258,
      "learning_rate": 1.932577739029352e-07,
      "loss": 0.8758,
      "mean_token_accuracy": 0.7543434649705887,
      "num_tokens": 482035.0,
      "step": 400
    },
    {
      "entropy": 2.282673454284668,
      "epoch": 0.0012709641270375143,
      "grad_norm": 15.464518547058105,
      "learning_rate": 1.981013271335852e-07,
      "loss": 0.8126,
      "mean_token_accuracy": 0.7647529050707818,
      "num_tokens": 494000.0,
      "step": 410
    },
    {
      "entropy": 2.306594988703728,
      "epoch": 0.0013019632520872098,
      "grad_norm": 16.271461486816406,
      "learning_rate": 2.0294488036423522e-07,
      "loss": 0.8339,
      "mean_token_accuracy": 0.755289240181446,
      "num_tokens": 505996.0,
      "step": 420
    },
    {
      "entropy": 2.3733946830034256,
      "epoch": 0.0013329623771369053,
      "grad_norm": 16.00725746154785,
      "learning_rate": 2.0778843359488523e-07,
      "loss": 0.8955,
      "mean_token_accuracy": 0.7515278026461601,
      "num_tokens": 517285.0,
      "step": 430
    },
    {
      "entropy": 2.2198241636157037,
      "epoch": 0.0013639615021866008,
      "grad_norm": 15.578631401062012,
      "learning_rate": 2.1263198682553522e-07,
      "loss": 0.746,
      "mean_token_accuracy": 0.7685888081789016,
      "num_tokens": 529581.0,
      "step": 440
    },
    {
      "entropy": 2.352111691981554,
      "epoch": 0.0013949606272362963,
      "grad_norm": 14.224849700927734,
      "learning_rate": 2.1747554005618524e-07,
      "loss": 0.859,
      "mean_token_accuracy": 0.7565658301115036,
      "num_tokens": 541039.0,
      "step": 450
    },
    {
      "entropy": 2.17586030960083,
      "epoch": 0.0014259597522859918,
      "grad_norm": 15.60697078704834,
      "learning_rate": 2.2231909328683525e-07,
      "loss": 0.7372,
      "mean_token_accuracy": 0.775518947839737,
      "num_tokens": 554099.0,
      "step": 460
    },
    {
      "entropy": 2.3119050413370132,
      "epoch": 0.0014569588773356871,
      "grad_norm": 14.367731094360352,
      "learning_rate": 2.2716264651748524e-07,
      "loss": 0.8381,
      "mean_token_accuracy": 0.7697366833686828,
      "num_tokens": 564893.0,
      "step": 470
    },
    {
      "entropy": 2.2108234629034995,
      "epoch": 0.0014879580023853826,
      "grad_norm": 17.166481018066406,
      "learning_rate": 2.3200619974813526e-07,
      "loss": 0.7081,
      "mean_token_accuracy": 0.7743594750761986,
      "num_tokens": 577998.0,
      "step": 480
    },
    {
      "entropy": 2.2766464233398436,
      "epoch": 0.0015189571274350781,
      "grad_norm": 16.3574275970459,
      "learning_rate": 2.3684975297878528e-07,
      "loss": 0.7699,
      "mean_token_accuracy": 0.7645501315593719,
      "num_tokens": 589715.0,
      "step": 490
    },
    {
      "entropy": 2.264049357175827,
      "epoch": 0.0015499562524847737,
      "grad_norm": 6.618349075317383,
      "learning_rate": 2.4169330620943527e-07,
      "loss": 0.78,
      "mean_token_accuracy": 0.7658553898334504,
      "num_tokens": 602659.0,
      "step": 500
    },
    {
      "entropy": 2.268606995046139,
      "epoch": 0.0015809553775344692,
      "grad_norm": 15.35413646697998,
      "learning_rate": 2.4653685944008526e-07,
      "loss": 0.8016,
      "mean_token_accuracy": 0.7688265934586525,
      "num_tokens": 615422.0,
      "step": 510
    },
    {
      "entropy": 2.181258149445057,
      "epoch": 0.0016119545025841645,
      "grad_norm": 16.496660232543945,
      "learning_rate": 2.513804126707353e-07,
      "loss": 0.6568,
      "mean_token_accuracy": 0.7991862148046494,
      "num_tokens": 627480.0,
      "step": 520
    },
    {
      "entropy": 2.207347111403942,
      "epoch": 0.00164295362763386,
      "grad_norm": 15.786961555480957,
      "learning_rate": 2.562239659013853e-07,
      "loss": 0.7347,
      "mean_token_accuracy": 0.7753358393907547,
      "num_tokens": 640090.0,
      "step": 530
    },
    {
      "entropy": 2.328162580728531,
      "epoch": 0.0016739527526835555,
      "grad_norm": 16.40909767150879,
      "learning_rate": 2.610675191320353e-07,
      "loss": 0.8802,
      "mean_token_accuracy": 0.7577004536986351,
      "num_tokens": 651238.0,
      "step": 540
    },
    {
      "entropy": 2.285073181986809,
      "epoch": 0.001704951877733251,
      "grad_norm": 15.64692211151123,
      "learning_rate": 2.659110723626853e-07,
      "loss": 0.808,
      "mean_token_accuracy": 0.7738980025053024,
      "num_tokens": 663483.0,
      "step": 550
    },
    {
      "entropy": 2.221916152536869,
      "epoch": 0.0017359510027829465,
      "grad_norm": 16.929515838623047,
      "learning_rate": 2.7075462559333526e-07,
      "loss": 0.7659,
      "mean_token_accuracy": 0.7734002724289895,
      "num_tokens": 676094.0,
      "step": 560
    },
    {
      "entropy": 2.207345911860466,
      "epoch": 0.001766950127832642,
      "grad_norm": 16.074033737182617,
      "learning_rate": 2.755981788239853e-07,
      "loss": 0.6923,
      "mean_token_accuracy": 0.7804467469453812,
      "num_tokens": 688999.0,
      "step": 570
    },
    {
      "entropy": 2.3080645158886908,
      "epoch": 0.0017979492528823373,
      "grad_norm": 14.152597427368164,
      "learning_rate": 2.804417320546353e-07,
      "loss": 0.8044,
      "mean_token_accuracy": 0.7673657357692718,
      "num_tokens": 700787.0,
      "step": 580
    },
    {
      "entropy": 2.3365137010812758,
      "epoch": 0.0018289483779320328,
      "grad_norm": 8.505473136901855,
      "learning_rate": 2.852852852852853e-07,
      "loss": 0.8693,
      "mean_token_accuracy": 0.7561125203967094,
      "num_tokens": 712382.0,
      "step": 590
    },
    {
      "entropy": 2.26595401763916,
      "epoch": 0.0018599475029817283,
      "grad_norm": 15.482884407043457,
      "learning_rate": 2.901288385159353e-07,
      "loss": 0.7499,
      "mean_token_accuracy": 0.7823521599173546,
      "num_tokens": 724324.0,
      "step": 600
    },
    {
      "entropy": 2.2956106424331666,
      "epoch": 0.0018909466280314239,
      "grad_norm": 17.781259536743164,
      "learning_rate": 2.949723917465853e-07,
      "loss": 0.8367,
      "mean_token_accuracy": 0.7637320205569267,
      "num_tokens": 736090.0,
      "step": 610
    },
    {
      "entropy": 2.380499321222305,
      "epoch": 0.0019219457530811194,
      "grad_norm": 13.911691665649414,
      "learning_rate": 2.998159449772353e-07,
      "loss": 0.8561,
      "mean_token_accuracy": 0.7514171093702317,
      "num_tokens": 746693.0,
      "step": 620
    },
    {
      "entropy": 2.3080674216151236,
      "epoch": 0.0019529448781308147,
      "grad_norm": 18.39996337890625,
      "learning_rate": 3.0465949820788535e-07,
      "loss": 0.7974,
      "mean_token_accuracy": 0.7719434529542923,
      "num_tokens": 758380.0,
      "step": 630
    },
    {
      "entropy": 2.33802735209465,
      "epoch": 0.00198394400318051,
      "grad_norm": 14.702959060668945,
      "learning_rate": 3.0950305143853534e-07,
      "loss": 0.8612,
      "mean_token_accuracy": 0.749081814289093,
      "num_tokens": 770107.0,
      "step": 640
    },
    {
      "entropy": 2.3233169302344323,
      "epoch": 0.002014943128230206,
      "grad_norm": 15.600394248962402,
      "learning_rate": 3.143466046691854e-07,
      "loss": 0.8358,
      "mean_token_accuracy": 0.7564076974987983,
      "num_tokens": 781888.0,
      "step": 650
    },
    {
      "entropy": 2.3429936945438383,
      "epoch": 0.002045942253279901,
      "grad_norm": 15.831826210021973,
      "learning_rate": 3.1919015789983537e-07,
      "loss": 0.8509,
      "mean_token_accuracy": 0.7663177505135537,
      "num_tokens": 792407.0,
      "step": 660
    },
    {
      "entropy": 2.2723167344927786,
      "epoch": 0.0020769413783295965,
      "grad_norm": 14.536561012268066,
      "learning_rate": 3.240337111304853e-07,
      "loss": 0.7932,
      "mean_token_accuracy": 0.7709016263484955,
      "num_tokens": 804473.0,
      "step": 670
    },
    {
      "entropy": 2.317927873134613,
      "epoch": 0.0021079405033792922,
      "grad_norm": 15.074233055114746,
      "learning_rate": 3.2887726436113535e-07,
      "loss": 0.7876,
      "mean_token_accuracy": 0.7756809115409851,
      "num_tokens": 815615.0,
      "step": 680
    },
    {
      "entropy": 2.279827579855919,
      "epoch": 0.0021389396284289875,
      "grad_norm": 15.745168685913086,
      "learning_rate": 3.3372081759178534e-07,
      "loss": 0.8729,
      "mean_token_accuracy": 0.7602526545524597,
      "num_tokens": 827866.0,
      "step": 690
    },
    {
      "entropy": 2.2185587152838706,
      "epoch": 0.0021699387534786833,
      "grad_norm": 13.561483383178711,
      "learning_rate": 3.385643708224354e-07,
      "loss": 0.748,
      "mean_token_accuracy": 0.779901297390461,
      "num_tokens": 839797.0,
      "step": 700
    },
    {
      "entropy": 2.234003722667694,
      "epoch": 0.0022009378785283785,
      "grad_norm": 13.80466079711914,
      "learning_rate": 3.4340792405308537e-07,
      "loss": 0.7714,
      "mean_token_accuracy": 0.7791284009814262,
      "num_tokens": 851460.0,
      "step": 710
    },
    {
      "entropy": 2.242032915353775,
      "epoch": 0.002231937003578074,
      "grad_norm": 8.220836639404297,
      "learning_rate": 3.482514772837354e-07,
      "loss": 0.7225,
      "mean_token_accuracy": 0.7788921162486077,
      "num_tokens": 863582.0,
      "step": 720
    },
    {
      "entropy": 2.2308412209153174,
      "epoch": 0.0022629361286277696,
      "grad_norm": 15.163178443908691,
      "learning_rate": 3.5309503051438535e-07,
      "loss": 0.7775,
      "mean_token_accuracy": 0.771105645596981,
      "num_tokens": 875961.0,
      "step": 730
    },
    {
      "entropy": 2.1884254679083823,
      "epoch": 0.002293935253677465,
      "grad_norm": 7.352316856384277,
      "learning_rate": 3.579385837450354e-07,
      "loss": 0.7359,
      "mean_token_accuracy": 0.7716657683253288,
      "num_tokens": 887838.0,
      "step": 740
    },
    {
      "entropy": 2.2045844838023188,
      "epoch": 0.0023249343787271606,
      "grad_norm": 14.715704917907715,
      "learning_rate": 3.627821369756854e-07,
      "loss": 0.7344,
      "mean_token_accuracy": 0.7731250256299973,
      "num_tokens": 900117.0,
      "step": 750
    },
    {
      "entropy": 2.261673641204834,
      "epoch": 0.002355933503776856,
      "grad_norm": 14.88746166229248,
      "learning_rate": 3.676256902063354e-07,
      "loss": 0.7627,
      "mean_token_accuracy": 0.7688748240470886,
      "num_tokens": 912552.0,
      "step": 760
    },
    {
      "entropy": 2.318584197759628,
      "epoch": 0.002386932628826551,
      "grad_norm": 15.337506294250488,
      "learning_rate": 3.7246924343698536e-07,
      "loss": 0.8543,
      "mean_token_accuracy": 0.7685193121433258,
      "num_tokens": 923351.0,
      "step": 770
    },
    {
      "entropy": 2.198628255724907,
      "epoch": 0.002417931753876247,
      "grad_norm": 16.160091400146484,
      "learning_rate": 3.773127966676354e-07,
      "loss": 0.7398,
      "mean_token_accuracy": 0.7804590031504631,
      "num_tokens": 935777.0,
      "step": 780
    },
    {
      "entropy": 2.291598491370678,
      "epoch": 0.002448930878925942,
      "grad_norm": 7.9013800621032715,
      "learning_rate": 3.821563498982854e-07,
      "loss": 0.7758,
      "mean_token_accuracy": 0.7750082641839982,
      "num_tokens": 947492.0,
      "step": 790
    },
    {
      "entropy": 2.3038537070155143,
      "epoch": 0.002479930003975638,
      "grad_norm": 16.498620986938477,
      "learning_rate": 3.8699990312893544e-07,
      "loss": 0.7678,
      "mean_token_accuracy": 0.7833232745528221,
      "num_tokens": 958770.0,
      "step": 800
    },
    {
      "entropy": 2.2682218730449675,
      "epoch": 0.0025109291290253332,
      "grad_norm": 15.202181816101074,
      "learning_rate": 3.9184345635958543e-07,
      "loss": 0.8518,
      "mean_token_accuracy": 0.7557914197444916,
      "num_tokens": 970777.0,
      "step": 810
    },
    {
      "entropy": 2.199041871726513,
      "epoch": 0.0025419282540750285,
      "grad_norm": 13.799867630004883,
      "learning_rate": 3.9668700959023547e-07,
      "loss": 0.7567,
      "mean_token_accuracy": 0.7677057057619094,
      "num_tokens": 982601.0,
      "step": 820
    },
    {
      "entropy": 2.2466016352176665,
      "epoch": 0.0025729273791247243,
      "grad_norm": 14.757229804992676,
      "learning_rate": 4.015305628208854e-07,
      "loss": 0.776,
      "mean_token_accuracy": 0.7803550854325294,
      "num_tokens": 994451.0,
      "step": 830
    },
    {
      "entropy": 2.304102659225464,
      "epoch": 0.0026039265041744196,
      "grad_norm": 16.70046615600586,
      "learning_rate": 4.0637411605153545e-07,
      "loss": 0.8395,
      "mean_token_accuracy": 0.7707262724637985,
      "num_tokens": 1005368.0,
      "step": 840
    },
    {
      "entropy": 2.2025340750813482,
      "epoch": 0.0026349256292241153,
      "grad_norm": 13.70840835571289,
      "learning_rate": 4.1121766928218544e-07,
      "loss": 0.722,
      "mean_token_accuracy": 0.7827607110142708,
      "num_tokens": 1017690.0,
      "step": 850
    },
    {
      "entropy": 2.1886146038770677,
      "epoch": 0.0026659247542738106,
      "grad_norm": 13.6995849609375,
      "learning_rate": 4.160612225128355e-07,
      "loss": 0.6548,
      "mean_token_accuracy": 0.7894467368721962,
      "num_tokens": 1030737.0,
      "step": 860
    },
    {
      "entropy": 2.119298684597015,
      "epoch": 0.002696923879323506,
      "grad_norm": 16.397750854492188,
      "learning_rate": 4.2090477574348547e-07,
      "loss": 0.6458,
      "mean_token_accuracy": 0.7952458307147026,
      "num_tokens": 1044784.0,
      "step": 870
    },
    {
      "entropy": 2.243351861834526,
      "epoch": 0.0027279230043732016,
      "grad_norm": 14.533041954040527,
      "learning_rate": 4.257483289741354e-07,
      "loss": 0.7294,
      "mean_token_accuracy": 0.7760630249977112,
      "num_tokens": 1056942.0,
      "step": 880
    },
    {
      "entropy": 2.3225993037223818,
      "epoch": 0.002758922129422897,
      "grad_norm": 16.506406784057617,
      "learning_rate": 4.3059188220478545e-07,
      "loss": 0.8633,
      "mean_token_accuracy": 0.7579209730029106,
      "num_tokens": 1067838.0,
      "step": 890
    },
    {
      "entropy": 2.0741579085588455,
      "epoch": 0.0027899212544725926,
      "grad_norm": 11.532833099365234,
      "learning_rate": 4.3543543543543544e-07,
      "loss": 0.6632,
      "mean_token_accuracy": 0.7945702284574508,
      "num_tokens": 1082957.0,
      "step": 900
    },
    {
      "entropy": 2.2643200919032096,
      "epoch": 0.002820920379522288,
      "grad_norm": 15.212760925292969,
      "learning_rate": 4.402789886660855e-07,
      "loss": 0.8228,
      "mean_token_accuracy": 0.76753601282835,
      "num_tokens": 1093997.0,
      "step": 910
    },
    {
      "entropy": 2.2407288014888764,
      "epoch": 0.0028519195045719837,
      "grad_norm": 16.18030548095703,
      "learning_rate": 4.451225418967355e-07,
      "loss": 0.8187,
      "mean_token_accuracy": 0.7693384140729904,
      "num_tokens": 1106072.0,
      "step": 920
    },
    {
      "entropy": 2.219868388772011,
      "epoch": 0.002882918629621679,
      "grad_norm": 16.22972297668457,
      "learning_rate": 4.499660951273855e-07,
      "loss": 0.7272,
      "mean_token_accuracy": 0.7875460609793663,
      "num_tokens": 1118504.0,
      "step": 930
    },
    {
      "entropy": 2.300278900563717,
      "epoch": 0.0029139177546713742,
      "grad_norm": 16.145498275756836,
      "learning_rate": 4.5480964835803545e-07,
      "loss": 0.7712,
      "mean_token_accuracy": 0.7815862789750099,
      "num_tokens": 1129720.0,
      "step": 940
    },
    {
      "entropy": 2.3173892736434936,
      "epoch": 0.00294491687972107,
      "grad_norm": 14.980684280395508,
      "learning_rate": 4.596532015886855e-07,
      "loss": 0.9265,
      "mean_token_accuracy": 0.7435617208480835,
      "num_tokens": 1140306.0,
      "step": 950
    },
    {
      "entropy": 2.172358636558056,
      "epoch": 0.0029759160047707653,
      "grad_norm": 14.839875221252441,
      "learning_rate": 4.644967548193355e-07,
      "loss": 0.7082,
      "mean_token_accuracy": 0.7885784819722176,
      "num_tokens": 1152921.0,
      "step": 960
    },
    {
      "entropy": 2.2328238427639007,
      "epoch": 0.003006915129820461,
      "grad_norm": 13.23880386352539,
      "learning_rate": 4.6934030804998553e-07,
      "loss": 0.7795,
      "mean_token_accuracy": 0.7753711074590683,
      "num_tokens": 1164657.0,
      "step": 970
    },
    {
      "entropy": 2.114644192159176,
      "epoch": 0.0030379142548701563,
      "grad_norm": 11.22083854675293,
      "learning_rate": 4.741838612806355e-07,
      "loss": 0.6744,
      "mean_token_accuracy": 0.7775777190923691,
      "num_tokens": 1177945.0,
      "step": 980
    },
    {
      "entropy": 2.1031323984265327,
      "epoch": 0.0030689133799198516,
      "grad_norm": 4.399860382080078,
      "learning_rate": 4.790274145112855e-07,
      "loss": 0.668,
      "mean_token_accuracy": 0.7964637473225593,
      "num_tokens": 1190989.0,
      "step": 990
    },
    {
      "entropy": 2.2336785733699798,
      "epoch": 0.0030999125049695473,
      "grad_norm": 15.71125602722168,
      "learning_rate": 4.838709677419355e-07,
      "loss": 0.772,
      "mean_token_accuracy": 0.7712054967880249,
      "num_tokens": 1201872.0,
      "step": 1000
    },
    {
      "entropy": 2.2714061349630357,
      "epoch": 0.0031309116300192426,
      "grad_norm": 15.550898551940918,
      "learning_rate": 4.887145209725856e-07,
      "loss": 0.795,
      "mean_token_accuracy": 0.7752420023083687,
      "num_tokens": 1213279.0,
      "step": 1010
    },
    {
      "entropy": 2.111206144094467,
      "epoch": 0.0031619107550689383,
      "grad_norm": 14.369123458862305,
      "learning_rate": 4.935580742032355e-07,
      "loss": 0.6963,
      "mean_token_accuracy": 0.7931261286139488,
      "num_tokens": 1226500.0,
      "step": 1020
    },
    {
      "entropy": 2.17624341994524,
      "epoch": 0.0031929098801186336,
      "grad_norm": 18.43202781677246,
      "learning_rate": 4.984016274338856e-07,
      "loss": 0.7139,
      "mean_token_accuracy": 0.7850997045636177,
      "num_tokens": 1238540.0,
      "step": 1030
    },
    {
      "entropy": 2.209955517947674,
      "epoch": 0.003223909005168329,
      "grad_norm": 16.401142120361328,
      "learning_rate": 5.032451806645356e-07,
      "loss": 0.7126,
      "mean_token_accuracy": 0.7926564201712608,
      "num_tokens": 1250753.0,
      "step": 1040
    },
    {
      "entropy": 2.267963781952858,
      "epoch": 0.0032549081302180247,
      "grad_norm": 15.670921325683594,
      "learning_rate": 5.080887338951856e-07,
      "loss": 0.8273,
      "mean_token_accuracy": 0.7623932763934136,
      "num_tokens": 1262037.0,
      "step": 1050
    },
    {
      "entropy": 1.9811606004834175,
      "epoch": 0.00328590725526772,
      "grad_norm": 14.190478324890137,
      "learning_rate": 5.129322871258355e-07,
      "loss": 0.5722,
      "mean_token_accuracy": 0.7981273889541626,
      "num_tokens": 1276993.0,
      "step": 1060
    },
    {
      "entropy": 2.1613825380802156,
      "epoch": 0.0033169063803174157,
      "grad_norm": 7.023738861083984,
      "learning_rate": 5.177758403564856e-07,
      "loss": 0.6904,
      "mean_token_accuracy": 0.7866356909275055,
      "num_tokens": 1289051.0,
      "step": 1070
    },
    {
      "entropy": 2.25160670876503,
      "epoch": 0.003347905505367111,
      "grad_norm": 14.642976760864258,
      "learning_rate": 5.226193935871355e-07,
      "loss": 0.8145,
      "mean_token_accuracy": 0.7689608931541443,
      "num_tokens": 1300620.0,
      "step": 1080
    },
    {
      "entropy": 2.2016865596175195,
      "epoch": 0.0033789046304168063,
      "grad_norm": 15.589518547058105,
      "learning_rate": 5.274629468177855e-07,
      "loss": 0.7511,
      "mean_token_accuracy": 0.7763894140720368,
      "num_tokens": 1312252.0,
      "step": 1090
    },
    {
      "entropy": 2.2270815491676332,
      "epoch": 0.003409903755466502,
      "grad_norm": 14.04956340789795,
      "learning_rate": 5.323065000484356e-07,
      "loss": 0.7445,
      "mean_token_accuracy": 0.7833787277340889,
      "num_tokens": 1323950.0,
      "step": 1100
    },
    {
      "entropy": 2.1960760056972504,
      "epoch": 0.0034409028805161973,
      "grad_norm": 17.72970199584961,
      "learning_rate": 5.371500532790856e-07,
      "loss": 0.7735,
      "mean_token_accuracy": 0.7613553330302238,
      "num_tokens": 1336895.0,
      "step": 1110
    },
    {
      "entropy": 2.0831907019019127,
      "epoch": 0.003471902005565893,
      "grad_norm": 10.339810371398926,
      "learning_rate": 5.419936065097356e-07,
      "loss": 0.6841,
      "mean_token_accuracy": 0.7895815879106521,
      "num_tokens": 1350813.0,
      "step": 1120
    },
    {
      "entropy": 2.1665103510022163,
      "epoch": 0.0035029011306155883,
      "grad_norm": 14.524060249328613,
      "learning_rate": 5.468371597403856e-07,
      "loss": 0.7598,
      "mean_token_accuracy": 0.7754980430006981,
      "num_tokens": 1362111.0,
      "step": 1130
    },
    {
      "entropy": 2.217361180484295,
      "epoch": 0.003533900255665284,
      "grad_norm": 14.30985164642334,
      "learning_rate": 5.516807129710356e-07,
      "loss": 0.7568,
      "mean_token_accuracy": 0.782513102889061,
      "num_tokens": 1373317.0,
      "step": 1140
    },
    {
      "entropy": 2.1852347582578657,
      "epoch": 0.0035648993807149794,
      "grad_norm": 16.35944938659668,
      "learning_rate": 5.565242662016856e-07,
      "loss": 0.7744,
      "mean_token_accuracy": 0.7764955341815949,
      "num_tokens": 1384737.0,
      "step": 1150
    },
    {
      "entropy": 2.2084068596363067,
      "epoch": 0.0035958985057646746,
      "grad_norm": 18.96138572692871,
      "learning_rate": 5.613678194323357e-07,
      "loss": 0.8437,
      "mean_token_accuracy": 0.7699245184659957,
      "num_tokens": 1396007.0,
      "step": 1160
    },
    {
      "entropy": 2.1020758375525475,
      "epoch": 0.0036268976308143704,
      "grad_norm": 7.519232749938965,
      "learning_rate": 5.662113726629856e-07,
      "loss": 0.6637,
      "mean_token_accuracy": 0.7951419726014137,
      "num_tokens": 1408380.0,
      "step": 1170
    },
    {
      "entropy": 2.157642234861851,
      "epoch": 0.0036578967558640657,
      "grad_norm": 15.029388427734375,
      "learning_rate": 5.710549258936356e-07,
      "loss": 0.7325,
      "mean_token_accuracy": 0.785622188448906,
      "num_tokens": 1420661.0,
      "step": 1180
    },
    {
      "entropy": 2.037250469624996,
      "epoch": 0.0036888958809137614,
      "grad_norm": 15.331890106201172,
      "learning_rate": 5.758984791242856e-07,
      "loss": 0.6885,
      "mean_token_accuracy": 0.7827024027705193,
      "num_tokens": 1434538.0,
      "step": 1190
    },
    {
      "entropy": 2.085458919405937,
      "epoch": 0.0037198950059634567,
      "grad_norm": 17.225831985473633,
      "learning_rate": 5.807420323549356e-07,
      "loss": 0.6795,
      "mean_token_accuracy": 0.7806023344397545,
      "num_tokens": 1448017.0,
      "step": 1200
    },
    {
      "entropy": 2.104422791302204,
      "epoch": 0.003750894131013152,
      "grad_norm": 16.510398864746094,
      "learning_rate": 5.855855855855856e-07,
      "loss": 0.7358,
      "mean_token_accuracy": 0.7740836933255195,
      "num_tokens": 1460208.0,
      "step": 1210
    },
    {
      "entropy": 2.1881393998861314,
      "epoch": 0.0037818932560628477,
      "grad_norm": 16.095951080322266,
      "learning_rate": 5.904291388162357e-07,
      "loss": 0.762,
      "mean_token_accuracy": 0.7820860460400582,
      "num_tokens": 1471627.0,
      "step": 1220
    },
    {
      "entropy": 2.092408487200737,
      "epoch": 0.003812892381112543,
      "grad_norm": 16.695289611816406,
      "learning_rate": 5.952726920468856e-07,
      "loss": 0.7157,
      "mean_token_accuracy": 0.787506853044033,
      "num_tokens": 1483004.0,
      "step": 1230
    },
    {
      "entropy": 2.0120772644877434,
      "epoch": 0.0038438915061622387,
      "grad_norm": 15.932051658630371,
      "learning_rate": 6.001162452775357e-07,
      "loss": 0.6887,
      "mean_token_accuracy": 0.7973495230078698,
      "num_tokens": 1496481.0,
      "step": 1240
    },
    {
      "entropy": 2.160550518333912,
      "epoch": 0.003874890631211934,
      "grad_norm": 15.765700340270996,
      "learning_rate": 6.049597985081857e-07,
      "loss": 0.7399,
      "mean_token_accuracy": 0.7820422008633614,
      "num_tokens": 1508377.0,
      "step": 1250
    },
    {
      "entropy": 2.1905842304229735,
      "epoch": 0.0039058897562616293,
      "grad_norm": 16.464012145996094,
      "learning_rate": 6.098033517388357e-07,
      "loss": 0.8825,
      "mean_token_accuracy": 0.7656736433506012,
      "num_tokens": 1519413.0,
      "step": 1260
    },
    {
      "entropy": 2.1988802313804627,
      "epoch": 0.003936888881311325,
      "grad_norm": 13.513689041137695,
      "learning_rate": 6.146469049694856e-07,
      "loss": 0.8144,
      "mean_token_accuracy": 0.7779660537838936,
      "num_tokens": 1530217.0,
      "step": 1270
    },
    {
      "entropy": 2.183208304643631,
      "epoch": 0.00396788800636102,
      "grad_norm": 7.471315383911133,
      "learning_rate": 6.194904582001357e-07,
      "loss": 0.7551,
      "mean_token_accuracy": 0.7802443280816078,
      "num_tokens": 1542024.0,
      "step": 1280
    },
    {
      "entropy": 2.0785870283842085,
      "epoch": 0.003998887131410716,
      "grad_norm": 17.728321075439453,
      "learning_rate": 6.243340114307856e-07,
      "loss": 0.6849,
      "mean_token_accuracy": 0.7881320580840111,
      "num_tokens": 1555038.0,
      "step": 1290
    },
    {
      "entropy": 2.0753221943974496,
      "epoch": 0.004029886256460412,
      "grad_norm": 13.686918258666992,
      "learning_rate": 6.291775646614356e-07,
      "loss": 0.6741,
      "mean_token_accuracy": 0.7906556352972984,
      "num_tokens": 1567185.0,
      "step": 1300
    },
    {
      "entropy": 2.083100973069668,
      "epoch": 0.004060885381510107,
      "grad_norm": 7.613674163818359,
      "learning_rate": 6.340211178920856e-07,
      "loss": 0.6265,
      "mean_token_accuracy": 0.7916320934891701,
      "num_tokens": 1580688.0,
      "step": 1310
    },
    {
      "entropy": 2.227837671339512,
      "epoch": 0.004091884506559802,
      "grad_norm": 17.518901824951172,
      "learning_rate": 6.388646711227357e-07,
      "loss": 0.7923,
      "mean_token_accuracy": 0.7738375559449195,
      "num_tokens": 1592038.0,
      "step": 1320
    },
    {
      "entropy": 2.1594600453972816,
      "epoch": 0.004122883631609498,
      "grad_norm": 15.655396461486816,
      "learning_rate": 6.437082243533857e-07,
      "loss": 0.7167,
      "mean_token_accuracy": 0.7852950558066368,
      "num_tokens": 1603983.0,
      "step": 1330
    },
    {
      "entropy": 2.0584792092442514,
      "epoch": 0.004153882756659193,
      "grad_norm": 12.117267608642578,
      "learning_rate": 6.485517775840357e-07,
      "loss": 0.6728,
      "mean_token_accuracy": 0.7903219699859619,
      "num_tokens": 1616249.0,
      "step": 1340
    },
    {
      "entropy": 2.088547757267952,
      "epoch": 0.004184881881708889,
      "grad_norm": 16.858585357666016,
      "learning_rate": 6.533953308146857e-07,
      "loss": 0.7391,
      "mean_token_accuracy": 0.778919804096222,
      "num_tokens": 1628549.0,
      "step": 1350
    },
    {
      "entropy": 2.1754756391048433,
      "epoch": 0.0042158810067585845,
      "grad_norm": 15.385615348815918,
      "learning_rate": 6.582388840453358e-07,
      "loss": 0.7799,
      "mean_token_accuracy": 0.7731549352407455,
      "num_tokens": 1639876.0,
      "step": 1360
    },
    {
      "entropy": 2.127278658747673,
      "epoch": 0.004246880131808279,
      "grad_norm": 8.056036949157715,
      "learning_rate": 6.630824372759858e-07,
      "loss": 0.7034,
      "mean_token_accuracy": 0.7827352419495582,
      "num_tokens": 1652097.0,
      "step": 1370
    },
    {
      "entropy": 2.0123574048280717,
      "epoch": 0.004277879256857975,
      "grad_norm": 6.172573566436768,
      "learning_rate": 6.679259905066357e-07,
      "loss": 0.6069,
      "mean_token_accuracy": 0.7994264915585518,
      "num_tokens": 1665043.0,
      "step": 1380
    },
    {
      "entropy": 2.096929004788399,
      "epoch": 0.004308878381907671,
      "grad_norm": 8.49801254272461,
      "learning_rate": 6.727695437372856e-07,
      "loss": 0.7136,
      "mean_token_accuracy": 0.7863008975982666,
      "num_tokens": 1677725.0,
      "step": 1390
    },
    {
      "entropy": 2.0824546411633493,
      "epoch": 0.0043398775069573665,
      "grad_norm": 14.584894180297852,
      "learning_rate": 6.776130969679358e-07,
      "loss": 0.7036,
      "mean_token_accuracy": 0.7845976158976555,
      "num_tokens": 1690896.0,
      "step": 1400
    },
    {
      "entropy": 2.13474909812212,
      "epoch": 0.004370876632007061,
      "grad_norm": 13.656248092651367,
      "learning_rate": 6.824566501985857e-07,
      "loss": 0.732,
      "mean_token_accuracy": 0.7771820738911629,
      "num_tokens": 1702660.0,
      "step": 1410
    },
    {
      "entropy": 2.0404009863734247,
      "epoch": 0.004401875757056757,
      "grad_norm": 13.051438331604004,
      "learning_rate": 6.873002034292357e-07,
      "loss": 0.6294,
      "mean_token_accuracy": 0.7985868468880654,
      "num_tokens": 1714725.0,
      "step": 1420
    },
    {
      "entropy": 2.0748074553906917,
      "epoch": 0.004432874882106453,
      "grad_norm": 13.570876121520996,
      "learning_rate": 6.921437566598857e-07,
      "loss": 0.6639,
      "mean_token_accuracy": 0.7878038331866264,
      "num_tokens": 1728017.0,
      "step": 1430
    },
    {
      "entropy": 2.072512838244438,
      "epoch": 0.004463874007156148,
      "grad_norm": 7.851343154907227,
      "learning_rate": 6.969873098905358e-07,
      "loss": 0.7054,
      "mean_token_accuracy": 0.7887277230620384,
      "num_tokens": 1740469.0,
      "step": 1440
    },
    {
      "entropy": 2.15796595364809,
      "epoch": 0.004494873132205843,
      "grad_norm": 13.530000686645508,
      "learning_rate": 7.018308631211858e-07,
      "loss": 0.7452,
      "mean_token_accuracy": 0.7796575650572777,
      "num_tokens": 1751979.0,
      "step": 1450
    },
    {
      "entropy": 2.0756077498197554,
      "epoch": 0.004525872257255539,
      "grad_norm": 14.499431610107422,
      "learning_rate": 7.066744163518358e-07,
      "loss": 0.7363,
      "mean_token_accuracy": 0.7842569574713707,
      "num_tokens": 1764493.0,
      "step": 1460
    },
    {
      "entropy": 1.988140694797039,
      "epoch": 0.004556871382305234,
      "grad_norm": 6.652562618255615,
      "learning_rate": 7.115179695824858e-07,
      "loss": 0.6319,
      "mean_token_accuracy": 0.7997438430786132,
      "num_tokens": 1777296.0,
      "step": 1470
    },
    {
      "entropy": 2.0755784660577774,
      "epoch": 0.00458787050735493,
      "grad_norm": 13.8985013961792,
      "learning_rate": 7.163615228131359e-07,
      "loss": 0.7402,
      "mean_token_accuracy": 0.7709438994526863,
      "num_tokens": 1789171.0,
      "step": 1480
    },
    {
      "entropy": 2.0258029848337173,
      "epoch": 0.0046188696324046255,
      "grad_norm": 7.679947376251221,
      "learning_rate": 7.212050760437858e-07,
      "loss": 0.6942,
      "mean_token_accuracy": 0.785995414853096,
      "num_tokens": 1801926.0,
      "step": 1490
    },
    {
      "entropy": 2.0404342114925385,
      "epoch": 0.004649868757454321,
      "grad_norm": 6.81480598449707,
      "learning_rate": 7.260486292744357e-07,
      "loss": 0.6672,
      "mean_token_accuracy": 0.7987658172845841,
      "num_tokens": 1814939.0,
      "step": 1500
    },
    {
      "entropy": 2.071673668920994,
      "epoch": 0.004680867882504016,
      "grad_norm": 15.106314659118652,
      "learning_rate": 7.308921825050857e-07,
      "loss": 0.6894,
      "mean_token_accuracy": 0.7804762050509453,
      "num_tokens": 1827340.0,
      "step": 1510
    },
    {
      "entropy": 2.0692993998527527,
      "epoch": 0.004711867007553712,
      "grad_norm": 15.853338241577148,
      "learning_rate": 7.357357357357357e-07,
      "loss": 0.7129,
      "mean_token_accuracy": 0.7840728506445884,
      "num_tokens": 1839510.0,
      "step": 1520
    },
    {
      "entropy": 2.135234770178795,
      "epoch": 0.0047428661326034075,
      "grad_norm": 6.517123699188232,
      "learning_rate": 7.405792889663858e-07,
      "loss": 0.8329,
      "mean_token_accuracy": 0.7573605760931968,
      "num_tokens": 1851013.0,
      "step": 1530
    },
    {
      "entropy": 2.036342756450176,
      "epoch": 0.004773865257653102,
      "grad_norm": 13.540575981140137,
      "learning_rate": 7.454228421970358e-07,
      "loss": 0.6977,
      "mean_token_accuracy": 0.7853451654314995,
      "num_tokens": 1863818.0,
      "step": 1540
    },
    {
      "entropy": 2.087058800458908,
      "epoch": 0.004804864382702798,
      "grad_norm": 8.103325843811035,
      "learning_rate": 7.502663954276858e-07,
      "loss": 0.7266,
      "mean_token_accuracy": 0.7897397369146347,
      "num_tokens": 1875751.0,
      "step": 1550
    },
    {
      "entropy": 2.091054494678974,
      "epoch": 0.004835863507752494,
      "grad_norm": 16.008771896362305,
      "learning_rate": 7.551099486583358e-07,
      "loss": 0.7302,
      "mean_token_accuracy": 0.7836149573326111,
      "num_tokens": 1887588.0,
      "step": 1560
    },
    {
      "entropy": 2.0431090027093886,
      "epoch": 0.0048668626328021896,
      "grad_norm": 8.24061393737793,
      "learning_rate": 7.599535018889859e-07,
      "loss": 0.6644,
      "mean_token_accuracy": 0.8029666885733604,
      "num_tokens": 1899870.0,
      "step": 1570
    },
    {
      "entropy": 1.8285046428442002,
      "epoch": 0.004897861757851884,
      "grad_norm": 7.983487606048584,
      "learning_rate": 7.647970551196359e-07,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8131875425577164,
      "num_tokens": 1915664.0,
      "step": 1580
    },
    {
      "entropy": 2.086928182840347,
      "epoch": 0.00492886088290158,
      "grad_norm": 13.69112777709961,
      "learning_rate": 7.696406083502859e-07,
      "loss": 0.6791,
      "mean_token_accuracy": 0.7944577828049659,
      "num_tokens": 1927081.0,
      "step": 1590
    },
    {
      "entropy": 2.0402188524603844,
      "epoch": 0.004959860007951276,
      "grad_norm": 15.185161590576172,
      "learning_rate": 7.744841615809357e-07,
      "loss": 0.7159,
      "mean_token_accuracy": 0.7936604365706443,
      "num_tokens": 1940073.0,
      "step": 1600
    },
    {
      "entropy": 2.149085035920143,
      "epoch": 0.004990859133000971,
      "grad_norm": 16.92563247680664,
      "learning_rate": 7.793277148115859e-07,
      "loss": 0.7675,
      "mean_token_accuracy": 0.7764922738075256,
      "num_tokens": 1951448.0,
      "step": 1610
    },
    {
      "entropy": 2.099419781565666,
      "epoch": 0.0050218582580506665,
      "grad_norm": 15.367118835449219,
      "learning_rate": 7.841712680422358e-07,
      "loss": 0.7569,
      "mean_token_accuracy": 0.7802120968699455,
      "num_tokens": 1962645.0,
      "step": 1620
    },
    {
      "entropy": 2.0213639751076697,
      "epoch": 0.005052857383100362,
      "grad_norm": 13.852250099182129,
      "learning_rate": 7.890148212728858e-07,
      "loss": 0.6431,
      "mean_token_accuracy": 0.7958111792802811,
      "num_tokens": 1974941.0,
      "step": 1630
    },
    {
      "entropy": 2.1066105023026465,
      "epoch": 0.005083856508150057,
      "grad_norm": 15.502884864807129,
      "learning_rate": 7.938583745035358e-07,
      "loss": 0.6782,
      "mean_token_accuracy": 0.7936943307518959,
      "num_tokens": 1987185.0,
      "step": 1640
    },
    {
      "entropy": 2.0182456478476523,
      "epoch": 0.005114855633199753,
      "grad_norm": 7.306271076202393,
      "learning_rate": 7.987019277341859e-07,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8059568896889686,
      "num_tokens": 2000487.0,
      "step": 1650
    },
    {
      "entropy": 2.147824813425541,
      "epoch": 0.0051458547582494485,
      "grad_norm": 18.491527557373047,
      "learning_rate": 8.035454809648359e-07,
      "loss": 0.8096,
      "mean_token_accuracy": 0.7748027428984642,
      "num_tokens": 2011914.0,
      "step": 1660
    },
    {
      "entropy": 1.9622011199593543,
      "epoch": 0.005176853883299144,
      "grad_norm": 13.780349731445312,
      "learning_rate": 8.083890341954859e-07,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8053760439157486,
      "num_tokens": 2025757.0,
      "step": 1670
    },
    {
      "entropy": 2.098005874454975,
      "epoch": 0.005207853008348839,
      "grad_norm": 14.81534194946289,
      "learning_rate": 8.132325874261359e-07,
      "loss": 0.7484,
      "mean_token_accuracy": 0.7781560063362122,
      "num_tokens": 2037711.0,
      "step": 1680
    },
    {
      "entropy": 2.052612027525902,
      "epoch": 0.005238852133398535,
      "grad_norm": 8.133528709411621,
      "learning_rate": 8.18076140656786e-07,
      "loss": 0.6822,
      "mean_token_accuracy": 0.79173034876585,
      "num_tokens": 2049738.0,
      "step": 1690
    },
    {
      "entropy": 1.9735546708106995,
      "epoch": 0.005269851258448231,
      "grad_norm": 16.37797737121582,
      "learning_rate": 8.22919693887436e-07,
      "loss": 0.6594,
      "mean_token_accuracy": 0.7950135871767998,
      "num_tokens": 2062316.0,
      "step": 1700
    },
    {
      "entropy": 2.029072532057762,
      "epoch": 0.0053008503834979254,
      "grad_norm": 5.160717010498047,
      "learning_rate": 8.277632471180859e-07,
      "loss": 0.6559,
      "mean_token_accuracy": 0.7948900654911994,
      "num_tokens": 2075028.0,
      "step": 1710
    },
    {
      "entropy": 2.0763241961598395,
      "epoch": 0.005331849508547621,
      "grad_norm": 15.46190357208252,
      "learning_rate": 8.326068003487358e-07,
      "loss": 0.684,
      "mean_token_accuracy": 0.7929447039961814,
      "num_tokens": 2086702.0,
      "step": 1720
    },
    {
      "entropy": 2.141001485288143,
      "epoch": 0.005362848633597317,
      "grad_norm": 13.897162437438965,
      "learning_rate": 8.374503535793858e-07,
      "loss": 0.7299,
      "mean_token_accuracy": 0.7951222896575928,
      "num_tokens": 2097706.0,
      "step": 1730
    },
    {
      "entropy": 2.0640996396541595,
      "epoch": 0.005393847758647012,
      "grad_norm": 13.737348556518555,
      "learning_rate": 8.422939068100359e-07,
      "loss": 0.7175,
      "mean_token_accuracy": 0.7896956667304039,
      "num_tokens": 2108391.0,
      "step": 1740
    },
    {
      "entropy": 1.9356930390000344,
      "epoch": 0.0054248468836967075,
      "grad_norm": 8.433479309082031,
      "learning_rate": 8.471374600406859e-07,
      "loss": 0.631,
      "mean_token_accuracy": 0.7932530760765075,
      "num_tokens": 2120471.0,
      "step": 1750
    },
    {
      "entropy": 2.0442377462983132,
      "epoch": 0.005455846008746403,
      "grad_norm": 18.229915618896484,
      "learning_rate": 8.519810132713359e-07,
      "loss": 0.7221,
      "mean_token_accuracy": 0.7952359080314636,
      "num_tokens": 2132167.0,
      "step": 1760
    },
    {
      "entropy": 2.04462625682354,
      "epoch": 0.005486845133796099,
      "grad_norm": 14.919930458068848,
      "learning_rate": 8.568245665019859e-07,
      "loss": 0.699,
      "mean_token_accuracy": 0.7886338919401169,
      "num_tokens": 2143973.0,
      "step": 1770
    },
    {
      "entropy": 2.0158157765865328,
      "epoch": 0.005517844258845794,
      "grad_norm": 19.1096248626709,
      "learning_rate": 8.61668119732636e-07,
      "loss": 0.7034,
      "mean_token_accuracy": 0.7857451155781746,
      "num_tokens": 2156439.0,
      "step": 1780
    },
    {
      "entropy": 2.1503632470965384,
      "epoch": 0.0055488433838954895,
      "grad_norm": 17.376968383789062,
      "learning_rate": 8.66511672963286e-07,
      "loss": 0.742,
      "mean_token_accuracy": 0.7804848656058312,
      "num_tokens": 2167869.0,
      "step": 1790
    },
    {
      "entropy": 2.0076410725712774,
      "epoch": 0.005579842508945185,
      "grad_norm": 18.342008590698242,
      "learning_rate": 8.71355226193936e-07,
      "loss": 0.642,
      "mean_token_accuracy": 0.8009541600942611,
      "num_tokens": 2180188.0,
      "step": 1800
    },
    {
      "entropy": 2.0049708664417265,
      "epoch": 0.00561084163399488,
      "grad_norm": 8.827497482299805,
      "learning_rate": 8.761987794245858e-07,
      "loss": 0.6828,
      "mean_token_accuracy": 0.7910264268517494,
      "num_tokens": 2192579.0,
      "step": 1810
    },
    {
      "entropy": 2.085582806169987,
      "epoch": 0.005641840759044576,
      "grad_norm": 17.707990646362305,
      "learning_rate": 8.81042332655236e-07,
      "loss": 0.7866,
      "mean_token_accuracy": 0.7726237401366234,
      "num_tokens": 2204679.0,
      "step": 1820
    },
    {
      "entropy": 2.0235436514019964,
      "epoch": 0.005672839884094272,
      "grad_norm": 9.473557472229004,
      "learning_rate": 8.858858858858859e-07,
      "loss": 0.7397,
      "mean_token_accuracy": 0.7853599205613137,
      "num_tokens": 2216727.0,
      "step": 1830
    },
    {
      "entropy": 2.0955518752336504,
      "epoch": 0.005703839009143967,
      "grad_norm": 14.616801261901855,
      "learning_rate": 8.907294391165359e-07,
      "loss": 0.7601,
      "mean_token_accuracy": 0.7755588620901108,
      "num_tokens": 2228738.0,
      "step": 1840
    },
    {
      "entropy": 2.0462396532297134,
      "epoch": 0.005734838134193662,
      "grad_norm": 12.282236099243164,
      "learning_rate": 8.955729923471859e-07,
      "loss": 0.7061,
      "mean_token_accuracy": 0.7882364228367805,
      "num_tokens": 2240796.0,
      "step": 1850
    },
    {
      "entropy": 2.0632716536521913,
      "epoch": 0.005765837259243358,
      "grad_norm": 18.294025421142578,
      "learning_rate": 9.00416545577836e-07,
      "loss": 0.72,
      "mean_token_accuracy": 0.7841200634837151,
      "num_tokens": 2252408.0,
      "step": 1860
    },
    {
      "entropy": 2.0464861541986465,
      "epoch": 0.005796836384293054,
      "grad_norm": 13.904800415039062,
      "learning_rate": 9.05260098808486e-07,
      "loss": 0.7355,
      "mean_token_accuracy": 0.7857467010617256,
      "num_tokens": 2263031.0,
      "step": 1870
    },
    {
      "entropy": 1.9634292259812356,
      "epoch": 0.0058278355093427485,
      "grad_norm": 7.740158557891846,
      "learning_rate": 9.10103652039136e-07,
      "loss": 0.6566,
      "mean_token_accuracy": 0.7917238727211953,
      "num_tokens": 2275474.0,
      "step": 1880
    },
    {
      "entropy": 1.934257398545742,
      "epoch": 0.005858834634392444,
      "grad_norm": 7.317460060119629,
      "learning_rate": 9.14947205269786e-07,
      "loss": 0.6318,
      "mean_token_accuracy": 0.8091923862695694,
      "num_tokens": 2287535.0,
      "step": 1890
    },
    {
      "entropy": 2.0293245404958724,
      "epoch": 0.00588983375944214,
      "grad_norm": 13.60265064239502,
      "learning_rate": 9.197907585004361e-07,
      "loss": 0.6919,
      "mean_token_accuracy": 0.7935610696673393,
      "num_tokens": 2299006.0,
      "step": 1900
    },
    {
      "entropy": 1.9936009973287583,
      "epoch": 0.005920832884491835,
      "grad_norm": 7.678701400756836,
      "learning_rate": 9.246343117310861e-07,
      "loss": 0.6424,
      "mean_token_accuracy": 0.7983625203371048,
      "num_tokens": 2310660.0,
      "step": 1910
    },
    {
      "entropy": 1.910171502828598,
      "epoch": 0.0059518320095415305,
      "grad_norm": 4.376688480377197,
      "learning_rate": 9.29477864961736e-07,
      "loss": 0.6749,
      "mean_token_accuracy": 0.791431900858879,
      "num_tokens": 2323660.0,
      "step": 1920
    },
    {
      "entropy": 1.9754113674163818,
      "epoch": 0.005982831134591226,
      "grad_norm": 16.712688446044922,
      "learning_rate": 9.343214181923859e-07,
      "loss": 0.6853,
      "mean_token_accuracy": 0.7904626414179802,
      "num_tokens": 2335479.0,
      "step": 1930
    },
    {
      "entropy": 1.9469898149371148,
      "epoch": 0.006013830259640922,
      "grad_norm": 7.1344733238220215,
      "learning_rate": 9.391649714230359e-07,
      "loss": 0.6714,
      "mean_token_accuracy": 0.7852405294775963,
      "num_tokens": 2347294.0,
      "step": 1940
    },
    {
      "entropy": 2.031823492050171,
      "epoch": 0.006044829384690617,
      "grad_norm": 14.974668502807617,
      "learning_rate": 9.44008524653686e-07,
      "loss": 0.7696,
      "mean_token_accuracy": 0.7771394848823547,
      "num_tokens": 2358885.0,
      "step": 1950
    },
    {
      "entropy": 2.001576192677021,
      "epoch": 0.006075828509740313,
      "grad_norm": 14.197503089904785,
      "learning_rate": 9.48852077884336e-07,
      "loss": 0.7037,
      "mean_token_accuracy": 0.7951447427272796,
      "num_tokens": 2370132.0,
      "step": 1960
    },
    {
      "entropy": 1.9794816732406617,
      "epoch": 0.006106827634790008,
      "grad_norm": 14.473783493041992,
      "learning_rate": 9.53695631114986e-07,
      "loss": 0.7147,
      "mean_token_accuracy": 0.7915472134947776,
      "num_tokens": 2381547.0,
      "step": 1970
    },
    {
      "entropy": 2.093907243013382,
      "epoch": 0.006137826759839703,
      "grad_norm": 14.241476058959961,
      "learning_rate": 9.585391843456359e-07,
      "loss": 0.8013,
      "mean_token_accuracy": 0.7786402046680451,
      "num_tokens": 2392334.0,
      "step": 1980
    },
    {
      "entropy": 1.9560458779335022,
      "epoch": 0.006168825884889399,
      "grad_norm": 13.721399307250977,
      "learning_rate": 9.633827375762862e-07,
      "loss": 0.6542,
      "mean_token_accuracy": 0.7954897537827492,
      "num_tokens": 2405128.0,
      "step": 1990
    },
    {
      "entropy": 2.0026605874300003,
      "epoch": 0.006199825009939095,
      "grad_norm": 16.695531845092773,
      "learning_rate": 9.68226290806936e-07,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7941533356904984,
      "num_tokens": 2416699.0,
      "step": 2000
    },
    {
      "entropy": 1.9149562925100327,
      "epoch": 0.00623082413498879,
      "grad_norm": 13.72096061706543,
      "learning_rate": 9.73069844037586e-07,
      "loss": 0.6126,
      "mean_token_accuracy": 0.7917631477117538,
      "num_tokens": 2429162.0,
      "step": 2010
    },
    {
      "entropy": 1.981172299385071,
      "epoch": 0.006261823260038485,
      "grad_norm": 15.749645233154297,
      "learning_rate": 9.77913397268236e-07,
      "loss": 0.7357,
      "mean_token_accuracy": 0.7887249678373337,
      "num_tokens": 2440565.0,
      "step": 2020
    },
    {
      "entropy": 1.9732341334223746,
      "epoch": 0.006292822385088181,
      "grad_norm": 16.25255012512207,
      "learning_rate": 9.827569504988861e-07,
      "loss": 0.6904,
      "mean_token_accuracy": 0.7926795452833175,
      "num_tokens": 2452318.0,
      "step": 2030
    },
    {
      "entropy": 1.9686682283878327,
      "epoch": 0.006323821510137877,
      "grad_norm": 7.215946197509766,
      "learning_rate": 9.87600503729536e-07,
      "loss": 0.6178,
      "mean_token_accuracy": 0.7987834200263023,
      "num_tokens": 2464480.0,
      "step": 2040
    },
    {
      "entropy": 2.0191121250391006,
      "epoch": 0.0063548206351875716,
      "grad_norm": 7.287489891052246,
      "learning_rate": 9.924440569601861e-07,
      "loss": 0.751,
      "mean_token_accuracy": 0.7695384725928307,
      "num_tokens": 2477286.0,
      "step": 2050
    },
    {
      "entropy": 1.9891038194298745,
      "epoch": 0.006385819760237267,
      "grad_norm": 15.255844116210938,
      "learning_rate": 9.97287610190836e-07,
      "loss": 0.6402,
      "mean_token_accuracy": 0.7988035991787911,
      "num_tokens": 2489470.0,
      "step": 2060
    },
    {
      "entropy": 2.0310388937592507,
      "epoch": 0.006416818885286963,
      "grad_norm": 18.825220108032227,
      "learning_rate": 1.0021311634214861e-06,
      "loss": 0.6939,
      "mean_token_accuracy": 0.790933045744896,
      "num_tokens": 2500898.0,
      "step": 2070
    },
    {
      "entropy": 1.9929038733243942,
      "epoch": 0.006447818010336658,
      "grad_norm": 12.69325065612793,
      "learning_rate": 1.0069747166521362e-06,
      "loss": 0.7442,
      "mean_token_accuracy": 0.7899109661579132,
      "num_tokens": 2511757.0,
      "step": 2080
    },
    {
      "entropy": 1.956763118505478,
      "epoch": 0.006478817135386354,
      "grad_norm": 11.434943199157715,
      "learning_rate": 1.011818269882786e-06,
      "loss": 0.7163,
      "mean_token_accuracy": 0.7898407876491547,
      "num_tokens": 2522888.0,
      "step": 2090
    },
    {
      "entropy": 1.97008508592844,
      "epoch": 0.006509816260436049,
      "grad_norm": 15.685938835144043,
      "learning_rate": 1.016661823113436e-06,
      "loss": 0.7396,
      "mean_token_accuracy": 0.785623998939991,
      "num_tokens": 2534453.0,
      "step": 2100
    },
    {
      "entropy": 1.9547904863953591,
      "epoch": 0.006540815385485745,
      "grad_norm": 13.821385383605957,
      "learning_rate": 1.021505376344086e-06,
      "loss": 0.7402,
      "mean_token_accuracy": 0.785305617749691,
      "num_tokens": 2546617.0,
      "step": 2110
    },
    {
      "entropy": 1.8934344440698623,
      "epoch": 0.00657181451053544,
      "grad_norm": 13.396963119506836,
      "learning_rate": 1.0263489295747362e-06,
      "loss": 0.627,
      "mean_token_accuracy": 0.8024709656834602,
      "num_tokens": 2559429.0,
      "step": 2120
    },
    {
      "entropy": 1.9755502566695213,
      "epoch": 0.006602813635585136,
      "grad_norm": 13.538336753845215,
      "learning_rate": 1.031192482805386e-06,
      "loss": 0.7059,
      "mean_token_accuracy": 0.7905629277229309,
      "num_tokens": 2570593.0,
      "step": 2130
    },
    {
      "entropy": 1.949264821410179,
      "epoch": 0.006633812760634831,
      "grad_norm": 14.021260261535645,
      "learning_rate": 1.0360360360360361e-06,
      "loss": 0.6666,
      "mean_token_accuracy": 0.7925188347697258,
      "num_tokens": 2582602.0,
      "step": 2140
    },
    {
      "entropy": 1.9639265209436416,
      "epoch": 0.006664811885684526,
      "grad_norm": 14.326904296875,
      "learning_rate": 1.040879589266686e-06,
      "loss": 0.7264,
      "mean_token_accuracy": 0.7851147443056107,
      "num_tokens": 2594982.0,
      "step": 2150
    },
    {
      "entropy": 1.908056390285492,
      "epoch": 0.006695811010734222,
      "grad_norm": 15.651636123657227,
      "learning_rate": 1.0457231424973361e-06,
      "loss": 0.6616,
      "mean_token_accuracy": 0.7910144224762916,
      "num_tokens": 2607900.0,
      "step": 2160
    },
    {
      "entropy": 2.064947435259819,
      "epoch": 0.006726810135783918,
      "grad_norm": 13.445158004760742,
      "learning_rate": 1.0505666957279862e-06,
      "loss": 0.7854,
      "mean_token_accuracy": 0.7805202215909958,
      "num_tokens": 2618173.0,
      "step": 2170
    },
    {
      "entropy": 1.9829859480261802,
      "epoch": 0.0067578092608336126,
      "grad_norm": 17.02464485168457,
      "learning_rate": 1.055410248958636e-06,
      "loss": 0.7521,
      "mean_token_accuracy": 0.7822321712970733,
      "num_tokens": 2629832.0,
      "step": 2180
    },
    {
      "entropy": 1.9977169573307036,
      "epoch": 0.006788808385883308,
      "grad_norm": 16.560504913330078,
      "learning_rate": 1.060253802189286e-06,
      "loss": 0.6964,
      "mean_token_accuracy": 0.793058268725872,
      "num_tokens": 2641302.0,
      "step": 2190
    },
    {
      "entropy": 2.0350762784481047,
      "epoch": 0.006819807510933004,
      "grad_norm": 16.779306411743164,
      "learning_rate": 1.0650973554199363e-06,
      "loss": 0.754,
      "mean_token_accuracy": 0.7842149198055267,
      "num_tokens": 2651948.0,
      "step": 2200
    },
    {
      "entropy": 1.9923300370573997,
      "epoch": 0.0068508066359827,
      "grad_norm": 16.477792739868164,
      "learning_rate": 1.0699409086505862e-06,
      "loss": 0.7457,
      "mean_token_accuracy": 0.7842650130391121,
      "num_tokens": 2663131.0,
      "step": 2210
    },
    {
      "entropy": 1.7988253235816956,
      "epoch": 0.006881805761032395,
      "grad_norm": 6.730621814727783,
      "learning_rate": 1.074784461881236e-06,
      "loss": 0.5973,
      "mean_token_accuracy": 0.80888991355896,
      "num_tokens": 2676405.0,
      "step": 2220
    },
    {
      "entropy": 1.9884469777345657,
      "epoch": 0.00691280488608209,
      "grad_norm": 15.052291870117188,
      "learning_rate": 1.0796280151118862e-06,
      "loss": 0.6849,
      "mean_token_accuracy": 0.7967025130987168,
      "num_tokens": 2688102.0,
      "step": 2230
    },
    {
      "entropy": 2.003503252565861,
      "epoch": 0.006943804011131786,
      "grad_norm": 7.750051498413086,
      "learning_rate": 1.0844715683425363e-06,
      "loss": 0.7153,
      "mean_token_accuracy": 0.7897115662693978,
      "num_tokens": 2699848.0,
      "step": 2240
    },
    {
      "entropy": 1.8995970517396927,
      "epoch": 0.006974803136181481,
      "grad_norm": 7.199272632598877,
      "learning_rate": 1.0893151215731861e-06,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8024276942014694,
      "num_tokens": 2713648.0,
      "step": 2250
    },
    {
      "entropy": 2.017641696333885,
      "epoch": 0.007005802261231177,
      "grad_norm": 15.489433288574219,
      "learning_rate": 1.0941586748038362e-06,
      "loss": 0.7434,
      "mean_token_accuracy": 0.7756871730089188,
      "num_tokens": 2725047.0,
      "step": 2260
    },
    {
      "entropy": 1.895441135764122,
      "epoch": 0.007036801386280872,
      "grad_norm": 17.170063018798828,
      "learning_rate": 1.0990022280344861e-06,
      "loss": 0.6771,
      "mean_token_accuracy": 0.7931119576096535,
      "num_tokens": 2737777.0,
      "step": 2270
    },
    {
      "entropy": 1.8989380843937398,
      "epoch": 0.007067800511330568,
      "grad_norm": 13.440524101257324,
      "learning_rate": 1.1038457812651362e-06,
      "loss": 0.6491,
      "mean_token_accuracy": 0.806152519583702,
      "num_tokens": 2749458.0,
      "step": 2280
    },
    {
      "entropy": 1.9610642537474632,
      "epoch": 0.007098799636380263,
      "grad_norm": 17.207420349121094,
      "learning_rate": 1.1086893344957863e-06,
      "loss": 0.7137,
      "mean_token_accuracy": 0.7841380387544632,
      "num_tokens": 2760376.0,
      "step": 2290
    },
    {
      "entropy": 1.9233600437641143,
      "epoch": 0.007129798761429959,
      "grad_norm": 14.002983093261719,
      "learning_rate": 1.1135328877264362e-06,
      "loss": 0.6691,
      "mean_token_accuracy": 0.7973776832222939,
      "num_tokens": 2771954.0,
      "step": 2300
    },
    {
      "entropy": 1.9587767377495766,
      "epoch": 0.0071607978864796544,
      "grad_norm": 14.966231346130371,
      "learning_rate": 1.118376440957086e-06,
      "loss": 0.6645,
      "mean_token_accuracy": 0.7941723063588142,
      "num_tokens": 2784375.0,
      "step": 2310
    },
    {
      "entropy": 1.994602382183075,
      "epoch": 0.007191797011529349,
      "grad_norm": 15.076648712158203,
      "learning_rate": 1.1232199941877364e-06,
      "loss": 0.7767,
      "mean_token_accuracy": 0.7866624847054482,
      "num_tokens": 2795409.0,
      "step": 2320
    },
    {
      "entropy": 1.871568574011326,
      "epoch": 0.007222796136579045,
      "grad_norm": 15.472980499267578,
      "learning_rate": 1.1280635474183863e-06,
      "loss": 0.6379,
      "mean_token_accuracy": 0.7967082962393761,
      "num_tokens": 2807885.0,
      "step": 2330
    },
    {
      "entropy": 1.9470938667654991,
      "epoch": 0.007253795261628741,
      "grad_norm": 8.13938045501709,
      "learning_rate": 1.1329071006490362e-06,
      "loss": 0.6471,
      "mean_token_accuracy": 0.796507653594017,
      "num_tokens": 2820252.0,
      "step": 2340
    },
    {
      "entropy": 1.92869935631752,
      "epoch": 0.007284794386678436,
      "grad_norm": 15.049967765808105,
      "learning_rate": 1.1377506538796862e-06,
      "loss": 0.705,
      "mean_token_accuracy": 0.7862442404031753,
      "num_tokens": 2832038.0,
      "step": 2350
    },
    {
      "entropy": 1.914162775874138,
      "epoch": 0.007315793511728131,
      "grad_norm": 14.12210750579834,
      "learning_rate": 1.1425942071103361e-06,
      "loss": 0.6729,
      "mean_token_accuracy": 0.7895276993513107,
      "num_tokens": 2844886.0,
      "step": 2360
    },
    {
      "entropy": 1.967710091173649,
      "epoch": 0.007346792636777827,
      "grad_norm": 7.458662033081055,
      "learning_rate": 1.1474377603409862e-06,
      "loss": 0.7499,
      "mean_token_accuracy": 0.7771386727690697,
      "num_tokens": 2857547.0,
      "step": 2370
    },
    {
      "entropy": 1.998754534125328,
      "epoch": 0.007377791761827523,
      "grad_norm": 14.489635467529297,
      "learning_rate": 1.1522813135716363e-06,
      "loss": 0.7481,
      "mean_token_accuracy": 0.7853496834635735,
      "num_tokens": 2868673.0,
      "step": 2380
    },
    {
      "entropy": 1.9616554602980614,
      "epoch": 0.007408790886877218,
      "grad_norm": 12.88488483428955,
      "learning_rate": 1.1571248668022862e-06,
      "loss": 0.7012,
      "mean_token_accuracy": 0.7969075560569763,
      "num_tokens": 2880536.0,
      "step": 2390
    },
    {
      "entropy": 2.0036663502454757,
      "epoch": 0.007439790011926913,
      "grad_norm": 13.311254501342773,
      "learning_rate": 1.161968420032936e-06,
      "loss": 0.7523,
      "mean_token_accuracy": 0.7829051271080971,
      "num_tokens": 2891783.0,
      "step": 2400
    },
    {
      "entropy": 1.9757195591926575,
      "epoch": 0.007470789136976609,
      "grad_norm": 16.514055252075195,
      "learning_rate": 1.1668119732635864e-06,
      "loss": 0.716,
      "mean_token_accuracy": 0.7831418976187706,
      "num_tokens": 2903383.0,
      "step": 2410
    },
    {
      "entropy": 1.897585642337799,
      "epoch": 0.007501788262026304,
      "grad_norm": 14.471832275390625,
      "learning_rate": 1.1716555264942363e-06,
      "loss": 0.6846,
      "mean_token_accuracy": 0.7894420132040978,
      "num_tokens": 2915336.0,
      "step": 2420
    },
    {
      "entropy": 2.0040734350681304,
      "epoch": 0.007532787387076,
      "grad_norm": 15.47757339477539,
      "learning_rate": 1.1764990797248862e-06,
      "loss": 0.7554,
      "mean_token_accuracy": 0.7723087310791016,
      "num_tokens": 2926824.0,
      "step": 2430
    },
    {
      "entropy": 1.9852148294448853,
      "epoch": 0.0075637865121256954,
      "grad_norm": 14.812381744384766,
      "learning_rate": 1.1813426329555363e-06,
      "loss": 0.7039,
      "mean_token_accuracy": 0.7904437810182572,
      "num_tokens": 2938564.0,
      "step": 2440
    },
    {
      "entropy": 1.9800173118710518,
      "epoch": 0.00759478563717539,
      "grad_norm": 7.551329612731934,
      "learning_rate": 1.1861861861861864e-06,
      "loss": 0.7097,
      "mean_token_accuracy": 0.7884911954402923,
      "num_tokens": 2950233.0,
      "step": 2450
    },
    {
      "entropy": 1.9487542375922202,
      "epoch": 0.007625784762225086,
      "grad_norm": 15.525565147399902,
      "learning_rate": 1.1910297394168362e-06,
      "loss": 0.706,
      "mean_token_accuracy": 0.7928291246294975,
      "num_tokens": 2962418.0,
      "step": 2460
    },
    {
      "entropy": 1.9103297770023346,
      "epoch": 0.007656783887274782,
      "grad_norm": 16.958786010742188,
      "learning_rate": 1.1958732926474863e-06,
      "loss": 0.6459,
      "mean_token_accuracy": 0.7916588813066483,
      "num_tokens": 2974375.0,
      "step": 2470
    },
    {
      "entropy": 1.8603628262877465,
      "epoch": 0.0076877830123244775,
      "grad_norm": 15.165175437927246,
      "learning_rate": 1.2007168458781362e-06,
      "loss": 0.6034,
      "mean_token_accuracy": 0.8035720109939575,
      "num_tokens": 2986722.0,
      "step": 2480
    },
    {
      "entropy": 1.8906757444143296,
      "epoch": 0.007718782137374172,
      "grad_norm": 14.799652099609375,
      "learning_rate": 1.2055603991087863e-06,
      "loss": 0.721,
      "mean_token_accuracy": 0.7916418060660362,
      "num_tokens": 2998939.0,
      "step": 2490
    },
    {
      "entropy": 1.8013859555125236,
      "epoch": 0.007749781262423868,
      "grad_norm": 6.574148654937744,
      "learning_rate": 1.2104039523394364e-06,
      "loss": 0.591,
      "mean_token_accuracy": 0.8007106646895409,
      "num_tokens": 3011678.0,
      "step": 2500
    },
    {
      "entropy": 1.9229096472263336,
      "epoch": 0.007780780387473564,
      "grad_norm": 15.573455810546875,
      "learning_rate": 1.2152475055700863e-06,
      "loss": 0.6444,
      "mean_token_accuracy": 0.7954755112528801,
      "num_tokens": 3024579.0,
      "step": 2510
    },
    {
      "entropy": 1.938393659889698,
      "epoch": 0.007811779512523259,
      "grad_norm": 14.614435195922852,
      "learning_rate": 1.2200910588007362e-06,
      "loss": 0.7364,
      "mean_token_accuracy": 0.7778101339936256,
      "num_tokens": 3036102.0,
      "step": 2520
    },
    {
      "entropy": 2.013949643075466,
      "epoch": 0.007842778637572954,
      "grad_norm": 8.017887115478516,
      "learning_rate": 1.2249346120313865e-06,
      "loss": 0.7157,
      "mean_token_accuracy": 0.7902801647782326,
      "num_tokens": 3047786.0,
      "step": 2530
    },
    {
      "entropy": 1.9791011467576027,
      "epoch": 0.00787377776262265,
      "grad_norm": 13.628707885742188,
      "learning_rate": 1.2297781652620364e-06,
      "loss": 0.7443,
      "mean_token_accuracy": 0.7846587881445884,
      "num_tokens": 3059102.0,
      "step": 2540
    },
    {
      "entropy": 1.9235740274190902,
      "epoch": 0.007904776887672346,
      "grad_norm": 14.778630256652832,
      "learning_rate": 1.2346217184926863e-06,
      "loss": 0.6579,
      "mean_token_accuracy": 0.7987881228327751,
      "num_tokens": 3071029.0,
      "step": 2550
    },
    {
      "entropy": 1.873728483915329,
      "epoch": 0.00793577601272204,
      "grad_norm": 5.178359508514404,
      "learning_rate": 1.2394652717233364e-06,
      "loss": 0.6452,
      "mean_token_accuracy": 0.7905326500535012,
      "num_tokens": 3083691.0,
      "step": 2560
    },
    {
      "entropy": 1.9002502277493476,
      "epoch": 0.007966775137771736,
      "grad_norm": 6.418622016906738,
      "learning_rate": 1.2443088249539862e-06,
      "loss": 0.6569,
      "mean_token_accuracy": 0.7951910138130188,
      "num_tokens": 3096979.0,
      "step": 2570
    },
    {
      "entropy": 2.094506660103798,
      "epoch": 0.007997774262821432,
      "grad_norm": 14.790390968322754,
      "learning_rate": 1.2491523781846363e-06,
      "loss": 0.7683,
      "mean_token_accuracy": 0.7839270010590553,
      "num_tokens": 3107589.0,
      "step": 2580
    },
    {
      "entropy": 1.9785408988595008,
      "epoch": 0.008028773387871127,
      "grad_norm": 13.493193626403809,
      "learning_rate": 1.2539959314152864e-06,
      "loss": 0.6176,
      "mean_token_accuracy": 0.7813256174325943,
      "num_tokens": 3120357.0,
      "step": 2590
    },
    {
      "entropy": 1.8982400611042975,
      "epoch": 0.008059772512920824,
      "grad_norm": 8.22580623626709,
      "learning_rate": 1.2588394846459365e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.8012925937771798,
      "num_tokens": 3134146.0,
      "step": 2600
    },
    {
      "entropy": 1.9128680050373077,
      "epoch": 0.008090771637970519,
      "grad_norm": 12.298065185546875,
      "learning_rate": 1.2636830378765864e-06,
      "loss": 0.6659,
      "mean_token_accuracy": 0.7981240004301071,
      "num_tokens": 3146528.0,
      "step": 2610
    },
    {
      "entropy": 1.8557716280221939,
      "epoch": 0.008121770763020213,
      "grad_norm": 13.299572944641113,
      "learning_rate": 1.2685265911072365e-06,
      "loss": 0.5324,
      "mean_token_accuracy": 0.811582088470459,
      "num_tokens": 3160369.0,
      "step": 2620
    },
    {
      "entropy": 1.9083554148674011,
      "epoch": 0.00815276988806991,
      "grad_norm": 12.815226554870605,
      "learning_rate": 1.2733701443378862e-06,
      "loss": 0.7,
      "mean_token_accuracy": 0.7826020732522011,
      "num_tokens": 3171895.0,
      "step": 2630
    },
    {
      "entropy": 1.858922117948532,
      "epoch": 0.008183769013119605,
      "grad_norm": 7.371999740600586,
      "learning_rate": 1.2782136975685363e-06,
      "loss": 0.6308,
      "mean_token_accuracy": 0.7977426454424859,
      "num_tokens": 3184167.0,
      "step": 2640
    },
    {
      "entropy": 1.9080826759338378,
      "epoch": 0.0082147681381693,
      "grad_norm": 15.016209602355957,
      "learning_rate": 1.2830572507991866e-06,
      "loss": 0.6998,
      "mean_token_accuracy": 0.7945051088929176,
      "num_tokens": 3196310.0,
      "step": 2650
    },
    {
      "entropy": 2.000731149315834,
      "epoch": 0.008245767263218996,
      "grad_norm": 7.419742107391357,
      "learning_rate": 1.2879008040298362e-06,
      "loss": 0.7669,
      "mean_token_accuracy": 0.7797714114189148,
      "num_tokens": 3207853.0,
      "step": 2660
    },
    {
      "entropy": 1.9436663463711739,
      "epoch": 0.008276766388268691,
      "grad_norm": 14.453858375549316,
      "learning_rate": 1.2927443572604863e-06,
      "loss": 0.7125,
      "mean_token_accuracy": 0.7874180287122726,
      "num_tokens": 3219704.0,
      "step": 2670
    },
    {
      "entropy": 1.8370084151625634,
      "epoch": 0.008307765513318386,
      "grad_norm": 13.891809463500977,
      "learning_rate": 1.2975879104911364e-06,
      "loss": 0.6604,
      "mean_token_accuracy": 0.7957833841443062,
      "num_tokens": 3232139.0,
      "step": 2680
    },
    {
      "entropy": 1.8790874809026719,
      "epoch": 0.008338764638368083,
      "grad_norm": 6.238895893096924,
      "learning_rate": 1.3024314637217863e-06,
      "loss": 0.6469,
      "mean_token_accuracy": 0.8013461381196976,
      "num_tokens": 3244302.0,
      "step": 2690
    },
    {
      "entropy": 1.8268845871090889,
      "epoch": 0.008369763763417777,
      "grad_norm": 8.341919898986816,
      "learning_rate": 1.3072750169524364e-06,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8057503595948219,
      "num_tokens": 3258280.0,
      "step": 2700
    },
    {
      "entropy": 1.9424984157085419,
      "epoch": 0.008400762888467472,
      "grad_norm": 14.885581970214844,
      "learning_rate": 1.3121185701830863e-06,
      "loss": 0.6694,
      "mean_token_accuracy": 0.7825981408357621,
      "num_tokens": 3270381.0,
      "step": 2710
    },
    {
      "entropy": 1.954266221821308,
      "epoch": 0.008431762013517169,
      "grad_norm": 14.373212814331055,
      "learning_rate": 1.3169621234137364e-06,
      "loss": 0.6892,
      "mean_token_accuracy": 0.7864295184612274,
      "num_tokens": 3282087.0,
      "step": 2720
    },
    {
      "entropy": 1.9162860542535782,
      "epoch": 0.008462761138566864,
      "grad_norm": 12.995490074157715,
      "learning_rate": 1.3218056766443865e-06,
      "loss": 0.7248,
      "mean_token_accuracy": 0.7878184035420418,
      "num_tokens": 3295059.0,
      "step": 2730
    },
    {
      "entropy": 1.8371574386954308,
      "epoch": 0.008493760263616559,
      "grad_norm": 14.610986709594727,
      "learning_rate": 1.3266492298750364e-06,
      "loss": 0.6114,
      "mean_token_accuracy": 0.7994052410125733,
      "num_tokens": 3307978.0,
      "step": 2740
    },
    {
      "entropy": 1.850772686302662,
      "epoch": 0.008524759388666255,
      "grad_norm": 14.59720516204834,
      "learning_rate": 1.3314927831056865e-06,
      "loss": 0.675,
      "mean_token_accuracy": 0.7939030557870865,
      "num_tokens": 3320392.0,
      "step": 2750
    },
    {
      "entropy": 1.8222134336829185,
      "epoch": 0.00855575851371595,
      "grad_norm": 5.238037109375,
      "learning_rate": 1.3363363363363366e-06,
      "loss": 0.5953,
      "mean_token_accuracy": 0.8046846255660057,
      "num_tokens": 3333208.0,
      "step": 2760
    },
    {
      "entropy": 1.9551292091608048,
      "epoch": 0.008586757638765647,
      "grad_norm": 16.98749351501465,
      "learning_rate": 1.3411798895669865e-06,
      "loss": 0.77,
      "mean_token_accuracy": 0.7961449101567268,
      "num_tokens": 3344004.0,
      "step": 2770
    },
    {
      "entropy": 1.9047151386737824,
      "epoch": 0.008617756763815342,
      "grad_norm": 15.543542861938477,
      "learning_rate": 1.3460234427976366e-06,
      "loss": 0.7398,
      "mean_token_accuracy": 0.7901824861764908,
      "num_tokens": 3354724.0,
      "step": 2780
    },
    {
      "entropy": 1.9245003148913384,
      "epoch": 0.008648755888865036,
      "grad_norm": 12.885390281677246,
      "learning_rate": 1.3508669960282864e-06,
      "loss": 0.7131,
      "mean_token_accuracy": 0.7908341690897942,
      "num_tokens": 3366242.0,
      "step": 2790
    },
    {
      "entropy": 1.873139852285385,
      "epoch": 0.008679755013914733,
      "grad_norm": 12.843483924865723,
      "learning_rate": 1.3557105492589365e-06,
      "loss": 0.6774,
      "mean_token_accuracy": 0.8066575720906257,
      "num_tokens": 3377837.0,
      "step": 2800
    },
    {
      "entropy": 1.8930615320801736,
      "epoch": 0.008710754138964428,
      "grad_norm": 10.335051536560059,
      "learning_rate": 1.3605541024895866e-06,
      "loss": 0.6729,
      "mean_token_accuracy": 0.789339256286621,
      "num_tokens": 3390578.0,
      "step": 2810
    },
    {
      "entropy": 1.9254612401127815,
      "epoch": 0.008741753264014123,
      "grad_norm": 13.433055877685547,
      "learning_rate": 1.3653976557202365e-06,
      "loss": 0.7126,
      "mean_token_accuracy": 0.7813933417201042,
      "num_tokens": 3401808.0,
      "step": 2820
    },
    {
      "entropy": 1.8774896636605263,
      "epoch": 0.00877275238906382,
      "grad_norm": 15.659320831298828,
      "learning_rate": 1.3702412089508866e-06,
      "loss": 0.6582,
      "mean_token_accuracy": 0.7960716530680656,
      "num_tokens": 3412922.0,
      "step": 2830
    },
    {
      "entropy": 1.8881728678941727,
      "epoch": 0.008803751514113514,
      "grad_norm": 14.669523239135742,
      "learning_rate": 1.3750847621815363e-06,
      "loss": 0.6416,
      "mean_token_accuracy": 0.8034508541226387,
      "num_tokens": 3425285.0,
      "step": 2840
    },
    {
      "entropy": 1.907693549990654,
      "epoch": 0.008834750639163209,
      "grad_norm": 17.293590545654297,
      "learning_rate": 1.3799283154121864e-06,
      "loss": 0.6792,
      "mean_token_accuracy": 0.797477675974369,
      "num_tokens": 3436697.0,
      "step": 2850
    },
    {
      "entropy": 1.9830518007278441,
      "epoch": 0.008865749764212906,
      "grad_norm": 15.576813697814941,
      "learning_rate": 1.3847718686428367e-06,
      "loss": 0.7526,
      "mean_token_accuracy": 0.779937069118023,
      "num_tokens": 3447898.0,
      "step": 2860
    },
    {
      "entropy": 2.013961046934128,
      "epoch": 0.0088967488892626,
      "grad_norm": 19.190113067626953,
      "learning_rate": 1.3896154218734864e-06,
      "loss": 0.7515,
      "mean_token_accuracy": 0.7880441144108772,
      "num_tokens": 3458894.0,
      "step": 2870
    },
    {
      "entropy": 1.9314459562301636,
      "epoch": 0.008927748014312295,
      "grad_norm": 12.630243301391602,
      "learning_rate": 1.3944589751041364e-06,
      "loss": 0.6666,
      "mean_token_accuracy": 0.7965318202972412,
      "num_tokens": 3471083.0,
      "step": 2880
    },
    {
      "entropy": 1.9051019206643105,
      "epoch": 0.008958747139361992,
      "grad_norm": 12.55810546875,
      "learning_rate": 1.3993025283347865e-06,
      "loss": 0.678,
      "mean_token_accuracy": 0.7915308341383934,
      "num_tokens": 3483322.0,
      "step": 2890
    },
    {
      "entropy": 1.889833214879036,
      "epoch": 0.008989746264411687,
      "grad_norm": 11.794564247131348,
      "learning_rate": 1.4041460815654364e-06,
      "loss": 0.5963,
      "mean_token_accuracy": 0.7985853642225266,
      "num_tokens": 3496188.0,
      "step": 2900
    },
    {
      "entropy": 1.8793418243527413,
      "epoch": 0.009020745389461382,
      "grad_norm": 6.941933631896973,
      "learning_rate": 1.4089896347960865e-06,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8100902363657951,
      "num_tokens": 3508412.0,
      "step": 2910
    },
    {
      "entropy": 1.7804090306162834,
      "epoch": 0.009051744514511078,
      "grad_norm": 7.316125392913818,
      "learning_rate": 1.4138331880267364e-06,
      "loss": 0.5352,
      "mean_token_accuracy": 0.8099708408117294,
      "num_tokens": 3522455.0,
      "step": 2920
    },
    {
      "entropy": 1.9128621608018874,
      "epoch": 0.009082743639560773,
      "grad_norm": 17.679824829101562,
      "learning_rate": 1.4186767412573865e-06,
      "loss": 0.684,
      "mean_token_accuracy": 0.7950308054685593,
      "num_tokens": 3534077.0,
      "step": 2930
    },
    {
      "entropy": 1.951967991888523,
      "epoch": 0.009113742764610468,
      "grad_norm": 15.615214347839355,
      "learning_rate": 1.4235202944880366e-06,
      "loss": 0.6822,
      "mean_token_accuracy": 0.7895113542675972,
      "num_tokens": 3545717.0,
      "step": 2940
    },
    {
      "entropy": 1.993370993435383,
      "epoch": 0.009144741889660165,
      "grad_norm": 6.757071018218994,
      "learning_rate": 1.4283638477186865e-06,
      "loss": 0.7091,
      "mean_token_accuracy": 0.7836599707603454,
      "num_tokens": 3556645.0,
      "step": 2950
    },
    {
      "entropy": 2.0306035369634627,
      "epoch": 0.00917574101470986,
      "grad_norm": 14.662017822265625,
      "learning_rate": 1.4332074009493366e-06,
      "loss": 0.7833,
      "mean_token_accuracy": 0.7853764742612839,
      "num_tokens": 3568121.0,
      "step": 2960
    },
    {
      "entropy": 1.981213203072548,
      "epoch": 0.009206740139759556,
      "grad_norm": 15.54949951171875,
      "learning_rate": 1.4380509541799867e-06,
      "loss": 0.7714,
      "mean_token_accuracy": 0.7741693139076233,
      "num_tokens": 3578617.0,
      "step": 2970
    },
    {
      "entropy": 1.872342900186777,
      "epoch": 0.009237739264809251,
      "grad_norm": 17.143173217773438,
      "learning_rate": 1.4428945074106366e-06,
      "loss": 0.6521,
      "mean_token_accuracy": 0.7934080705046653,
      "num_tokens": 3590839.0,
      "step": 2980
    },
    {
      "entropy": 1.9597981229424477,
      "epoch": 0.009268738389858946,
      "grad_norm": 14.204910278320312,
      "learning_rate": 1.4477380606412867e-06,
      "loss": 0.7372,
      "mean_token_accuracy": 0.7796640574932099,
      "num_tokens": 3602197.0,
      "step": 2990
    },
    {
      "entropy": 1.9134353682398797,
      "epoch": 0.009299737514908642,
      "grad_norm": 15.377642631530762,
      "learning_rate": 1.4525816138719365e-06,
      "loss": 0.5693,
      "mean_token_accuracy": 0.80574119836092,
      "num_tokens": 3614912.0,
      "step": 3000
    },
    {
      "entropy": 1.978382122516632,
      "epoch": 0.009330736639958337,
      "grad_norm": 7.089384078979492,
      "learning_rate": 1.4574251671025866e-06,
      "loss": 0.7043,
      "mean_token_accuracy": 0.7872982352972031,
      "num_tokens": 3626738.0,
      "step": 3010
    },
    {
      "entropy": 1.9824548229575156,
      "epoch": 0.009361735765008032,
      "grad_norm": 7.4606781005859375,
      "learning_rate": 1.4622687203332367e-06,
      "loss": 0.7379,
      "mean_token_accuracy": 0.7830901652574539,
      "num_tokens": 3638714.0,
      "step": 3020
    },
    {
      "entropy": 1.8985080510377883,
      "epoch": 0.009392734890057729,
      "grad_norm": 13.994378089904785,
      "learning_rate": 1.4671122735638866e-06,
      "loss": 0.6141,
      "mean_token_accuracy": 0.8004487827420235,
      "num_tokens": 3651514.0,
      "step": 3030
    },
    {
      "entropy": 1.9109556749463081,
      "epoch": 0.009423734015107424,
      "grad_norm": 11.190482139587402,
      "learning_rate": 1.4719558267945367e-06,
      "loss": 0.6391,
      "mean_token_accuracy": 0.8045189529657364,
      "num_tokens": 3662697.0,
      "step": 3040
    },
    {
      "entropy": 1.8760196268558502,
      "epoch": 0.009454733140157118,
      "grad_norm": 12.307433128356934,
      "learning_rate": 1.4767993800251864e-06,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8026507899165154,
      "num_tokens": 3674503.0,
      "step": 3050
    },
    {
      "entropy": 1.8742030158638954,
      "epoch": 0.009485732265206815,
      "grad_norm": 7.280045509338379,
      "learning_rate": 1.4816429332558365e-06,
      "loss": 0.6784,
      "mean_token_accuracy": 0.7885152325034142,
      "num_tokens": 3686269.0,
      "step": 3060
    },
    {
      "entropy": 1.7706302866339683,
      "epoch": 0.00951673139025651,
      "grad_norm": 16.125728607177734,
      "learning_rate": 1.4864864864864868e-06,
      "loss": 0.6367,
      "mean_token_accuracy": 0.8014728352427483,
      "num_tokens": 3699713.0,
      "step": 3070
    },
    {
      "entropy": 1.9473395302891732,
      "epoch": 0.009547730515306205,
      "grad_norm": 13.948101997375488,
      "learning_rate": 1.4913300397171365e-06,
      "loss": 0.7302,
      "mean_token_accuracy": 0.7855488672852516,
      "num_tokens": 3710852.0,
      "step": 3080
    },
    {
      "entropy": 1.9612311527132988,
      "epoch": 0.009578729640355901,
      "grad_norm": 13.97789192199707,
      "learning_rate": 1.4961735929477866e-06,
      "loss": 0.6629,
      "mean_token_accuracy": 0.7972382113337517,
      "num_tokens": 3722160.0,
      "step": 3090
    },
    {
      "entropy": 1.960909178853035,
      "epoch": 0.009609728765405596,
      "grad_norm": 17.231172561645508,
      "learning_rate": 1.5010171461784366e-06,
      "loss": 0.754,
      "mean_token_accuracy": 0.7888594791293144,
      "num_tokens": 3732729.0,
      "step": 3100
    },
    {
      "entropy": 1.8868347622454167,
      "epoch": 0.009640727890455291,
      "grad_norm": 15.940637588500977,
      "learning_rate": 1.5058606994090865e-06,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8036533623933793,
      "num_tokens": 3746194.0,
      "step": 3110
    },
    {
      "entropy": 2.00607987344265,
      "epoch": 0.009671727015504988,
      "grad_norm": 16.22367286682129,
      "learning_rate": 1.5107042526397366e-06,
      "loss": 0.7286,
      "mean_token_accuracy": 0.7928143292665482,
      "num_tokens": 3756751.0,
      "step": 3120
    },
    {
      "entropy": 1.9029984012246133,
      "epoch": 0.009702726140554683,
      "grad_norm": 6.391202926635742,
      "learning_rate": 1.5155478058703865e-06,
      "loss": 0.6529,
      "mean_token_accuracy": 0.7976860404014587,
      "num_tokens": 3768774.0,
      "step": 3130
    },
    {
      "entropy": 1.900516900420189,
      "epoch": 0.009733725265604379,
      "grad_norm": 8.11662483215332,
      "learning_rate": 1.5203913591010366e-06,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8091256007552147,
      "num_tokens": 3780820.0,
      "step": 3140
    },
    {
      "entropy": 1.9626210525631904,
      "epoch": 0.009764724390654074,
      "grad_norm": 12.520034790039062,
      "learning_rate": 1.5252349123316867e-06,
      "loss": 0.7038,
      "mean_token_accuracy": 0.7925601005554199,
      "num_tokens": 3792530.0,
      "step": 3150
    },
    {
      "entropy": 1.9671666517853736,
      "epoch": 0.009795723515703769,
      "grad_norm": 13.528938293457031,
      "learning_rate": 1.5300784655623366e-06,
      "loss": 0.6599,
      "mean_token_accuracy": 0.7918861374258995,
      "num_tokens": 3804625.0,
      "step": 3160
    },
    {
      "entropy": 2.005161625146866,
      "epoch": 0.009826722640753465,
      "grad_norm": 15.400350570678711,
      "learning_rate": 1.5349220187929867e-06,
      "loss": 0.7848,
      "mean_token_accuracy": 0.7739899069070816,
      "num_tokens": 3815444.0,
      "step": 3170
    },
    {
      "entropy": 1.9359246730804442,
      "epoch": 0.00985772176580316,
      "grad_norm": 12.780713081359863,
      "learning_rate": 1.5397655720236368e-06,
      "loss": 0.6371,
      "mean_token_accuracy": 0.7938413843512535,
      "num_tokens": 3827684.0,
      "step": 3180
    },
    {
      "entropy": 1.8368206307291985,
      "epoch": 0.009888720890852855,
      "grad_norm": 14.72050952911377,
      "learning_rate": 1.5446091252542867e-06,
      "loss": 0.6173,
      "mean_token_accuracy": 0.8116529732942581,
      "num_tokens": 3840934.0,
      "step": 3190
    },
    {
      "entropy": 1.9378165647387504,
      "epoch": 0.009919720015902552,
      "grad_norm": 14.93758487701416,
      "learning_rate": 1.5494526784849368e-06,
      "loss": 0.7211,
      "mean_token_accuracy": 0.7911474660038949,
      "num_tokens": 3852242.0,
      "step": 3200
    },
    {
      "entropy": 1.9542915537953376,
      "epoch": 0.009950719140952247,
      "grad_norm": 16.133121490478516,
      "learning_rate": 1.5542962317155866e-06,
      "loss": 0.7008,
      "mean_token_accuracy": 0.7842934697866439,
      "num_tokens": 3864015.0,
      "step": 3210
    },
    {
      "entropy": 1.9251963838934898,
      "epoch": 0.009981718266001941,
      "grad_norm": 15.101007461547852,
      "learning_rate": 1.5591397849462367e-06,
      "loss": 0.6018,
      "mean_token_accuracy": 0.8036449015140533,
      "num_tokens": 3876673.0,
      "step": 3220
    },
    {
      "entropy": 1.9787523686885833,
      "epoch": 0.010012717391051638,
      "grad_norm": 11.887308120727539,
      "learning_rate": 1.5639833381768868e-06,
      "loss": 0.7158,
      "mean_token_accuracy": 0.79226556122303,
      "num_tokens": 3888063.0,
      "step": 3230
    },
    {
      "entropy": 1.99595515280962,
      "epoch": 0.010043716516101333,
      "grad_norm": 12.0714693069458,
      "learning_rate": 1.5688268914075367e-06,
      "loss": 0.7203,
      "mean_token_accuracy": 0.7814503669738769,
      "num_tokens": 3899339.0,
      "step": 3240
    },
    {
      "entropy": 1.92583971619606,
      "epoch": 0.010074715641151028,
      "grad_norm": 14.31470775604248,
      "learning_rate": 1.5736704446381868e-06,
      "loss": 0.6961,
      "mean_token_accuracy": 0.7950286611914634,
      "num_tokens": 3911124.0,
      "step": 3250
    },
    {
      "entropy": 1.9035099178552628,
      "epoch": 0.010105714766200724,
      "grad_norm": 14.206976890563965,
      "learning_rate": 1.5785139978688365e-06,
      "loss": 0.6835,
      "mean_token_accuracy": 0.7935141950845719,
      "num_tokens": 3923110.0,
      "step": 3260
    },
    {
      "entropy": 1.8103605896234511,
      "epoch": 0.01013671389125042,
      "grad_norm": 11.845626831054688,
      "learning_rate": 1.5833575510994868e-06,
      "loss": 0.5957,
      "mean_token_accuracy": 0.8122652113437653,
      "num_tokens": 3936835.0,
      "step": 3270
    },
    {
      "entropy": 1.9789897188544274,
      "epoch": 0.010167713016300114,
      "grad_norm": 16.31435775756836,
      "learning_rate": 1.5882011043301369e-06,
      "loss": 0.6931,
      "mean_token_accuracy": 0.790991485118866,
      "num_tokens": 3948691.0,
      "step": 3280
    },
    {
      "entropy": 1.9635705202817917,
      "epoch": 0.01019871214134981,
      "grad_norm": 18.17036247253418,
      "learning_rate": 1.5930446575607866e-06,
      "loss": 0.7027,
      "mean_token_accuracy": 0.7976090654730796,
      "num_tokens": 3960588.0,
      "step": 3290
    },
    {
      "entropy": 1.8178722724318503,
      "epoch": 0.010229711266399506,
      "grad_norm": 4.42231559753418,
      "learning_rate": 1.5978882107914367e-06,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8080889210104942,
      "num_tokens": 3974030.0,
      "step": 3300
    },
    {
      "entropy": 2.014702117443085,
      "epoch": 0.010260710391449202,
      "grad_norm": 13.163649559020996,
      "learning_rate": 1.6027317640220868e-06,
      "loss": 0.7217,
      "mean_token_accuracy": 0.7939134731888771,
      "num_tokens": 3984810.0,
      "step": 3310
    },
    {
      "entropy": 1.9950988233089446,
      "epoch": 0.010291709516498897,
      "grad_norm": 15.815617561340332,
      "learning_rate": 1.6075753172527366e-06,
      "loss": 0.7374,
      "mean_token_accuracy": 0.7863194987177848,
      "num_tokens": 3996191.0,
      "step": 3320
    },
    {
      "entropy": 1.9132106304168701,
      "epoch": 0.010322708641548592,
      "grad_norm": 15.737754821777344,
      "learning_rate": 1.6124188704833867e-06,
      "loss": 0.7396,
      "mean_token_accuracy": 0.7841628924012184,
      "num_tokens": 4007285.0,
      "step": 3330
    },
    {
      "entropy": 1.8114167541265487,
      "epoch": 0.010353707766598289,
      "grad_norm": 13.689631462097168,
      "learning_rate": 1.6172624237140366e-06,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8101338252425194,
      "num_tokens": 4020382.0,
      "step": 3340
    },
    {
      "entropy": 1.8398823648691178,
      "epoch": 0.010384706891647983,
      "grad_norm": 14.286563873291016,
      "learning_rate": 1.6221059769446867e-06,
      "loss": 0.6389,
      "mean_token_accuracy": 0.8049512296915055,
      "num_tokens": 4032030.0,
      "step": 3350
    },
    {
      "entropy": 1.9326018333435058,
      "epoch": 0.010415706016697678,
      "grad_norm": 16.373798370361328,
      "learning_rate": 1.6269495301753368e-06,
      "loss": 0.7155,
      "mean_token_accuracy": 0.7966372415423393,
      "num_tokens": 4043118.0,
      "step": 3360
    },
    {
      "entropy": 1.905664649605751,
      "epoch": 0.010446705141747375,
      "grad_norm": 12.13680648803711,
      "learning_rate": 1.6317930834059867e-06,
      "loss": 0.7183,
      "mean_token_accuracy": 0.7907021954655647,
      "num_tokens": 4055085.0,
      "step": 3370
    },
    {
      "entropy": 1.7940703511238099,
      "epoch": 0.01047770426679707,
      "grad_norm": 6.879446983337402,
      "learning_rate": 1.6366366366366368e-06,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8221948266029357,
      "num_tokens": 4067929.0,
      "step": 3380
    },
    {
      "entropy": 1.964170092344284,
      "epoch": 0.010508703391846765,
      "grad_norm": 17.012205123901367,
      "learning_rate": 1.6414801898672869e-06,
      "loss": 0.7205,
      "mean_token_accuracy": 0.7949927791953086,
      "num_tokens": 4079191.0,
      "step": 3390
    },
    {
      "entropy": 1.9278406336903573,
      "epoch": 0.010539702516896461,
      "grad_norm": 15.772573471069336,
      "learning_rate": 1.6463237430979368e-06,
      "loss": 0.6634,
      "mean_token_accuracy": 0.7951213136315346,
      "num_tokens": 4090571.0,
      "step": 3400
    },
    {
      "entropy": 1.894165250658989,
      "epoch": 0.010570701641946156,
      "grad_norm": 7.302316188812256,
      "learning_rate": 1.6511672963285869e-06,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8126410812139511,
      "num_tokens": 4102478.0,
      "step": 3410
    },
    {
      "entropy": 1.8725802972912788,
      "epoch": 0.010601700766995851,
      "grad_norm": 12.873794555664062,
      "learning_rate": 1.6560108495592367e-06,
      "loss": 0.6578,
      "mean_token_accuracy": 0.808172582089901,
      "num_tokens": 4114349.0,
      "step": 3420
    },
    {
      "entropy": 1.9037815779447556,
      "epoch": 0.010632699892045547,
      "grad_norm": 6.517296314239502,
      "learning_rate": 1.6608544027898868e-06,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8006631091237069,
      "num_tokens": 4126609.0,
      "step": 3430
    },
    {
      "entropy": 1.9055060788989067,
      "epoch": 0.010663699017095242,
      "grad_norm": 15.855356216430664,
      "learning_rate": 1.665697956020537e-06,
      "loss": 0.7089,
      "mean_token_accuracy": 0.7951321870088577,
      "num_tokens": 4138476.0,
      "step": 3440
    },
    {
      "entropy": 1.9123858138918877,
      "epoch": 0.010694698142144937,
      "grad_norm": 14.749032020568848,
      "learning_rate": 1.6705415092511868e-06,
      "loss": 0.6421,
      "mean_token_accuracy": 0.799940338730812,
      "num_tokens": 4150474.0,
      "step": 3450
    },
    {
      "entropy": 1.9116358309984207,
      "epoch": 0.010725697267194634,
      "grad_norm": 13.992262840270996,
      "learning_rate": 1.675385062481837e-06,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8097973421216011,
      "num_tokens": 4162977.0,
      "step": 3460
    },
    {
      "entropy": 1.9737422183156013,
      "epoch": 0.010756696392244329,
      "grad_norm": 19.65887451171875,
      "learning_rate": 1.6802286157124866e-06,
      "loss": 0.7513,
      "mean_token_accuracy": 0.786674790084362,
      "num_tokens": 4174740.0,
      "step": 3470
    },
    {
      "entropy": 1.9084539324045182,
      "epoch": 0.010787695517294024,
      "grad_norm": 7.143436908721924,
      "learning_rate": 1.685072168943137e-06,
      "loss": 0.6752,
      "mean_token_accuracy": 0.8009995728731155,
      "num_tokens": 4187420.0,
      "step": 3480
    },
    {
      "entropy": 1.9033571183681488,
      "epoch": 0.01081869464234372,
      "grad_norm": 12.777674674987793,
      "learning_rate": 1.689915722173787e-06,
      "loss": 0.6786,
      "mean_token_accuracy": 0.8100510269403458,
      "num_tokens": 4198138.0,
      "step": 3490
    },
    {
      "entropy": 1.8302487596869468,
      "epoch": 0.010849693767393415,
      "grad_norm": 14.17345905303955,
      "learning_rate": 1.6947592754044367e-06,
      "loss": 0.6558,
      "mean_token_accuracy": 0.7974511593580246,
      "num_tokens": 4210639.0,
      "step": 3500
    },
    {
      "entropy": 1.8793094977736473,
      "epoch": 0.010880692892443112,
      "grad_norm": 15.137555122375488,
      "learning_rate": 1.6996028286350868e-06,
      "loss": 0.6217,
      "mean_token_accuracy": 0.8036959484219551,
      "num_tokens": 4222201.0,
      "step": 3510
    },
    {
      "entropy": 1.8393305256962775,
      "epoch": 0.010911692017492806,
      "grad_norm": 12.150157928466797,
      "learning_rate": 1.704446381865737e-06,
      "loss": 0.6265,
      "mean_token_accuracy": 0.8053354471921921,
      "num_tokens": 4234129.0,
      "step": 3520
    },
    {
      "entropy": 1.8129374995827674,
      "epoch": 0.010942691142542501,
      "grad_norm": 6.902397155761719,
      "learning_rate": 1.7092899350963867e-06,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8074716746807098,
      "num_tokens": 4247360.0,
      "step": 3530
    },
    {
      "entropy": 1.847067180275917,
      "epoch": 0.010973690267592198,
      "grad_norm": 9.06917953491211,
      "learning_rate": 1.7141334883270368e-06,
      "loss": 0.6657,
      "mean_token_accuracy": 0.7977814689278603,
      "num_tokens": 4260605.0,
      "step": 3540
    },
    {
      "entropy": 1.935109367966652,
      "epoch": 0.011004689392641893,
      "grad_norm": 13.947689056396484,
      "learning_rate": 1.7189770415576867e-06,
      "loss": 0.6719,
      "mean_token_accuracy": 0.8019600152969361,
      "num_tokens": 4271987.0,
      "step": 3550
    },
    {
      "entropy": 1.8618975296616553,
      "epoch": 0.011035688517691588,
      "grad_norm": 7.604375839233398,
      "learning_rate": 1.7238205947883368e-06,
      "loss": 0.6746,
      "mean_token_accuracy": 0.7953737512230873,
      "num_tokens": 4283981.0,
      "step": 3560
    },
    {
      "entropy": 1.9873407423496245,
      "epoch": 0.011066687642741284,
      "grad_norm": 12.513638496398926,
      "learning_rate": 1.728664148018987e-06,
      "loss": 0.8001,
      "mean_token_accuracy": 0.7810007587075234,
      "num_tokens": 4295314.0,
      "step": 3570
    },
    {
      "entropy": 1.8661947965621948,
      "epoch": 0.011097686767790979,
      "grad_norm": 8.476912498474121,
      "learning_rate": 1.7335077012496368e-06,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8108328223228455,
      "num_tokens": 4306919.0,
      "step": 3580
    },
    {
      "entropy": 1.9405100882053374,
      "epoch": 0.011128685892840674,
      "grad_norm": 11.374967575073242,
      "learning_rate": 1.7383512544802869e-06,
      "loss": 0.7597,
      "mean_token_accuracy": 0.7805906578898429,
      "num_tokens": 4318405.0,
      "step": 3590
    },
    {
      "entropy": 1.8536500945687293,
      "epoch": 0.01115968501789037,
      "grad_norm": 17.95821762084961,
      "learning_rate": 1.743194807710937e-06,
      "loss": 0.6538,
      "mean_token_accuracy": 0.7971635937690735,
      "num_tokens": 4330730.0,
      "step": 3600
    },
    {
      "entropy": 1.8885389596223832,
      "epoch": 0.011190684142940065,
      "grad_norm": 15.392064094543457,
      "learning_rate": 1.7480383609415869e-06,
      "loss": 0.6708,
      "mean_token_accuracy": 0.8034435287117958,
      "num_tokens": 4341732.0,
      "step": 3610
    },
    {
      "entropy": 1.9072374895215034,
      "epoch": 0.01122168326798976,
      "grad_norm": 15.804481506347656,
      "learning_rate": 1.752881914172237e-06,
      "loss": 0.7122,
      "mean_token_accuracy": 0.7888311371207237,
      "num_tokens": 4352944.0,
      "step": 3620
    },
    {
      "entropy": 1.8851828515529632,
      "epoch": 0.011252682393039457,
      "grad_norm": 14.782668113708496,
      "learning_rate": 1.7577254674028869e-06,
      "loss": 0.6998,
      "mean_token_accuracy": 0.7909768223762512,
      "num_tokens": 4363879.0,
      "step": 3630
    },
    {
      "entropy": 1.8380015999078751,
      "epoch": 0.011283681518089152,
      "grad_norm": 14.773683547973633,
      "learning_rate": 1.762569020633537e-06,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8070567846298218,
      "num_tokens": 4375759.0,
      "step": 3640
    },
    {
      "entropy": 1.8058712944388389,
      "epoch": 0.011314680643138847,
      "grad_norm": 13.945756912231445,
      "learning_rate": 1.767412573864187e-06,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8059628069400787,
      "num_tokens": 4389341.0,
      "step": 3650
    },
    {
      "entropy": 1.8813235729932785,
      "epoch": 0.011345679768188543,
      "grad_norm": 12.711499214172363,
      "learning_rate": 1.772256127094837e-06,
      "loss": 0.6448,
      "mean_token_accuracy": 0.8153657332062721,
      "num_tokens": 4400801.0,
      "step": 3660
    },
    {
      "entropy": 1.849845513701439,
      "epoch": 0.011376678893238238,
      "grad_norm": 15.230743408203125,
      "learning_rate": 1.777099680325487e-06,
      "loss": 0.5924,
      "mean_token_accuracy": 0.8115788295865058,
      "num_tokens": 4413458.0,
      "step": 3670
    },
    {
      "entropy": 1.9601149141788483,
      "epoch": 0.011407678018287935,
      "grad_norm": 13.785551071166992,
      "learning_rate": 1.7819432335561367e-06,
      "loss": 0.7572,
      "mean_token_accuracy": 0.7834953621029854,
      "num_tokens": 4424999.0,
      "step": 3680
    },
    {
      "entropy": 1.8368350595235825,
      "epoch": 0.01143867714333763,
      "grad_norm": 7.308245658874512,
      "learning_rate": 1.786786786786787e-06,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8099960327148438,
      "num_tokens": 4436854.0,
      "step": 3690
    },
    {
      "entropy": 1.8301602736115457,
      "epoch": 0.011469676268387324,
      "grad_norm": 13.396066665649414,
      "learning_rate": 1.791630340017437e-06,
      "loss": 0.61,
      "mean_token_accuracy": 0.8089475125074387,
      "num_tokens": 4449388.0,
      "step": 3700
    },
    {
      "entropy": 1.9232326075434685,
      "epoch": 0.011500675393437021,
      "grad_norm": 14.820940971374512,
      "learning_rate": 1.7964738932480868e-06,
      "loss": 0.7029,
      "mean_token_accuracy": 0.7972525000572205,
      "num_tokens": 4460601.0,
      "step": 3710
    },
    {
      "entropy": 1.9275172814726829,
      "epoch": 0.011531674518486716,
      "grad_norm": 13.760834693908691,
      "learning_rate": 1.8013174464787369e-06,
      "loss": 0.67,
      "mean_token_accuracy": 0.7982899993658066,
      "num_tokens": 4472147.0,
      "step": 3720
    },
    {
      "entropy": 1.841733305156231,
      "epoch": 0.01156267364353641,
      "grad_norm": 4.695168495178223,
      "learning_rate": 1.8061609997093872e-06,
      "loss": 0.604,
      "mean_token_accuracy": 0.8182388171553612,
      "num_tokens": 4484477.0,
      "step": 3730
    },
    {
      "entropy": 1.8292774349451064,
      "epoch": 0.011593672768586107,
      "grad_norm": 5.965690612792969,
      "learning_rate": 1.8110045529400368e-06,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8187378898262978,
      "num_tokens": 4497212.0,
      "step": 3740
    },
    {
      "entropy": 1.8096903830766677,
      "epoch": 0.011624671893635802,
      "grad_norm": 6.74788761138916,
      "learning_rate": 1.815848106170687e-06,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8042061701416969,
      "num_tokens": 4509570.0,
      "step": 3750
    },
    {
      "entropy": 1.8733750000596046,
      "epoch": 0.011655671018685497,
      "grad_norm": 15.247547149658203,
      "learning_rate": 1.8206916594013368e-06,
      "loss": 0.6363,
      "mean_token_accuracy": 0.7944400921463967,
      "num_tokens": 4521995.0,
      "step": 3760
    },
    {
      "entropy": 1.9379871144890786,
      "epoch": 0.011686670143735194,
      "grad_norm": 5.622687816619873,
      "learning_rate": 1.825535212631987e-06,
      "loss": 0.6895,
      "mean_token_accuracy": 0.7910193100571632,
      "num_tokens": 4533546.0,
      "step": 3770
    },
    {
      "entropy": 1.8144460648298264,
      "epoch": 0.011717669268784888,
      "grad_norm": 7.539439678192139,
      "learning_rate": 1.830378765862637e-06,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8145319998264313,
      "num_tokens": 4546173.0,
      "step": 3780
    },
    {
      "entropy": 1.914628729224205,
      "epoch": 0.011748668393834583,
      "grad_norm": 15.919760704040527,
      "learning_rate": 1.835222319093287e-06,
      "loss": 0.6686,
      "mean_token_accuracy": 0.8031175434589386,
      "num_tokens": 4557017.0,
      "step": 3790
    },
    {
      "entropy": 1.873679694533348,
      "epoch": 0.01177966751888428,
      "grad_norm": 15.300971031188965,
      "learning_rate": 1.840065872323937e-06,
      "loss": 0.6355,
      "mean_token_accuracy": 0.7994208186864853,
      "num_tokens": 4569294.0,
      "step": 3800
    },
    {
      "entropy": 1.9676134511828423,
      "epoch": 0.011810666643933975,
      "grad_norm": 14.174129486083984,
      "learning_rate": 1.844909425554587e-06,
      "loss": 0.709,
      "mean_token_accuracy": 0.7975045815110207,
      "num_tokens": 4580270.0,
      "step": 3810
    },
    {
      "entropy": 1.7934204161167144,
      "epoch": 0.01184166576898367,
      "grad_norm": 14.214337348937988,
      "learning_rate": 1.849752978785237e-06,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8222482308745385,
      "num_tokens": 4593333.0,
      "step": 3820
    },
    {
      "entropy": 1.880036075413227,
      "epoch": 0.011872664894033366,
      "grad_norm": 13.970633506774902,
      "learning_rate": 1.854596532015887e-06,
      "loss": 0.6634,
      "mean_token_accuracy": 0.7958371922373771,
      "num_tokens": 4605903.0,
      "step": 3830
    },
    {
      "entropy": 1.8512480556964874,
      "epoch": 0.011903664019083061,
      "grad_norm": 11.63100814819336,
      "learning_rate": 1.859440085246537e-06,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8205496445298195,
      "num_tokens": 4618539.0,
      "step": 3840
    },
    {
      "entropy": 1.9032649219036102,
      "epoch": 0.011934663144132758,
      "grad_norm": 14.786650657653809,
      "learning_rate": 1.864283638477187e-06,
      "loss": 0.6271,
      "mean_token_accuracy": 0.7999899938702584,
      "num_tokens": 4630679.0,
      "step": 3850
    },
    {
      "entropy": 1.8507294937968255,
      "epoch": 0.011965662269182453,
      "grad_norm": 15.505373001098633,
      "learning_rate": 1.8691271917078371e-06,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8145019263029099,
      "num_tokens": 4644442.0,
      "step": 3860
    },
    {
      "entropy": 2.043610119819641,
      "epoch": 0.011996661394232147,
      "grad_norm": 17.056150436401367,
      "learning_rate": 1.873970744938487e-06,
      "loss": 0.719,
      "mean_token_accuracy": 0.7810810253024101,
      "num_tokens": 4654902.0,
      "step": 3870
    },
    {
      "entropy": 1.9894573852419852,
      "epoch": 0.012027660519281844,
      "grad_norm": 14.494179725646973,
      "learning_rate": 1.8788142981691371e-06,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8132184654474258,
      "num_tokens": 4665669.0,
      "step": 3880
    },
    {
      "entropy": 2.0583216458559037,
      "epoch": 0.012058659644331539,
      "grad_norm": 12.334035873413086,
      "learning_rate": 1.883657851399787e-06,
      "loss": 0.7878,
      "mean_token_accuracy": 0.7764063253998756,
      "num_tokens": 4676308.0,
      "step": 3890
    },
    {
      "entropy": 1.8608665503561497,
      "epoch": 0.012089658769381234,
      "grad_norm": 6.846446514129639,
      "learning_rate": 1.8885014046304371e-06,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8085096433758736,
      "num_tokens": 4689421.0,
      "step": 3900
    },
    {
      "entropy": 1.957814833521843,
      "epoch": 0.01212065789443093,
      "grad_norm": 12.504427909851074,
      "learning_rate": 1.8933449578610872e-06,
      "loss": 0.6852,
      "mean_token_accuracy": 0.8001207739114762,
      "num_tokens": 4701492.0,
      "step": 3910
    },
    {
      "entropy": 1.8347266063094139,
      "epoch": 0.012151657019480625,
      "grad_norm": 15.001429557800293,
      "learning_rate": 1.8981885110917369e-06,
      "loss": 0.5869,
      "mean_token_accuracy": 0.81041249781847,
      "num_tokens": 4714252.0,
      "step": 3920
    },
    {
      "entropy": 1.9000362247228622,
      "epoch": 0.01218265614453032,
      "grad_norm": 15.871277809143066,
      "learning_rate": 1.903032064322387e-06,
      "loss": 0.6574,
      "mean_token_accuracy": 0.8029514774680138,
      "num_tokens": 4725743.0,
      "step": 3930
    },
    {
      "entropy": 1.9085357084870338,
      "epoch": 0.012213655269580017,
      "grad_norm": 11.861626625061035,
      "learning_rate": 1.907875617553037e-06,
      "loss": 0.6232,
      "mean_token_accuracy": 0.8074530705809593,
      "num_tokens": 4737602.0,
      "step": 3940
    },
    {
      "entropy": 1.8550396144390107,
      "epoch": 0.012244654394629712,
      "grad_norm": 8.315388679504395,
      "learning_rate": 1.912719170783687e-06,
      "loss": 0.6392,
      "mean_token_accuracy": 0.7964627891778946,
      "num_tokens": 4749843.0,
      "step": 3950
    },
    {
      "entropy": 1.8868352055549622,
      "epoch": 0.012275653519679406,
      "grad_norm": 6.1222028732299805,
      "learning_rate": 1.9175627240143373e-06,
      "loss": 0.6339,
      "mean_token_accuracy": 0.8049155220389366,
      "num_tokens": 4762341.0,
      "step": 3960
    },
    {
      "entropy": 1.9090922087430955,
      "epoch": 0.012306652644729103,
      "grad_norm": 13.033318519592285,
      "learning_rate": 1.922406277244987e-06,
      "loss": 0.7009,
      "mean_token_accuracy": 0.7983160421252251,
      "num_tokens": 4774156.0,
      "step": 3970
    },
    {
      "entropy": 1.8667214959859848,
      "epoch": 0.012337651769778798,
      "grad_norm": 13.20643424987793,
      "learning_rate": 1.927249830475637e-06,
      "loss": 0.6541,
      "mean_token_accuracy": 0.7958778321743012,
      "num_tokens": 4786644.0,
      "step": 3980
    },
    {
      "entropy": 1.8116755485534668,
      "epoch": 0.012368650894828493,
      "grad_norm": 4.642879962921143,
      "learning_rate": 1.9320933837062873e-06,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8170793354511261,
      "num_tokens": 4799594.0,
      "step": 3990
    },
    {
      "entropy": 1.7941434875130653,
      "epoch": 0.01239965001987819,
      "grad_norm": 13.760114669799805,
      "learning_rate": 1.9369369369369372e-06,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8124065786600113,
      "num_tokens": 4811994.0,
      "step": 4000
    },
    {
      "entropy": 1.9247236236929894,
      "epoch": 0.012430649144927884,
      "grad_norm": 13.147157669067383,
      "learning_rate": 1.941780490167587e-06,
      "loss": 0.7134,
      "mean_token_accuracy": 0.7897293627262115,
      "num_tokens": 4823158.0,
      "step": 4010
    },
    {
      "entropy": 1.853769588470459,
      "epoch": 0.01246164826997758,
      "grad_norm": 11.97884750366211,
      "learning_rate": 1.9466240433982374e-06,
      "loss": 0.6361,
      "mean_token_accuracy": 0.8066916093230247,
      "num_tokens": 4834679.0,
      "step": 4020
    },
    {
      "entropy": 1.8805758222937583,
      "epoch": 0.012492647395027276,
      "grad_norm": 13.70005989074707,
      "learning_rate": 1.951467596628887e-06,
      "loss": 0.684,
      "mean_token_accuracy": 0.7904590263962745,
      "num_tokens": 4846234.0,
      "step": 4030
    },
    {
      "entropy": 1.9497404143214225,
      "epoch": 0.01252364652007697,
      "grad_norm": 5.448077201843262,
      "learning_rate": 1.956311149859537e-06,
      "loss": 0.7173,
      "mean_token_accuracy": 0.7891417220234871,
      "num_tokens": 4857890.0,
      "step": 4040
    },
    {
      "entropy": 1.8685417965054512,
      "epoch": 0.012554645645126667,
      "grad_norm": 13.170268058776855,
      "learning_rate": 1.961154703090187e-06,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8130503103137017,
      "num_tokens": 4869779.0,
      "step": 4050
    },
    {
      "entropy": 1.91714718490839,
      "epoch": 0.012585644770176362,
      "grad_norm": 13.95510196685791,
      "learning_rate": 1.965998256320837e-06,
      "loss": 0.6498,
      "mean_token_accuracy": 0.7988581836223603,
      "num_tokens": 4881872.0,
      "step": 4060
    },
    {
      "entropy": 1.8488874793052674,
      "epoch": 0.012616643895226057,
      "grad_norm": 15.533134460449219,
      "learning_rate": 1.9708418095514873e-06,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8086504280567169,
      "num_tokens": 4893930.0,
      "step": 4070
    },
    {
      "entropy": 1.827307391166687,
      "epoch": 0.012647643020275753,
      "grad_norm": 14.92998218536377,
      "learning_rate": 1.975685362782137e-06,
      "loss": 0.6516,
      "mean_token_accuracy": 0.7971120476722717,
      "num_tokens": 4905551.0,
      "step": 4080
    },
    {
      "entropy": 1.8286997452378273,
      "epoch": 0.012678642145325448,
      "grad_norm": 14.836496353149414,
      "learning_rate": 1.980528916012787e-06,
      "loss": 0.6449,
      "mean_token_accuracy": 0.7994356840848923,
      "num_tokens": 4917025.0,
      "step": 4090
    },
    {
      "entropy": 1.8299384236335754,
      "epoch": 0.012709641270375143,
      "grad_norm": 13.094915390014648,
      "learning_rate": 1.985372469243437e-06,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8010865300893784,
      "num_tokens": 4930128.0,
      "step": 4100
    },
    {
      "entropy": 1.895867148041725,
      "epoch": 0.01274064039542484,
      "grad_norm": 13.526707649230957,
      "learning_rate": 1.990216022474087e-06,
      "loss": 0.7135,
      "mean_token_accuracy": 0.7984941571950912,
      "num_tokens": 4941517.0,
      "step": 4110
    },
    {
      "entropy": 1.89191782027483,
      "epoch": 0.012771639520474535,
      "grad_norm": 14.294553756713867,
      "learning_rate": 1.995059575704737e-06,
      "loss": 0.6931,
      "mean_token_accuracy": 0.7926407441496849,
      "num_tokens": 4952810.0,
      "step": 4120
    },
    {
      "entropy": 1.8057468429207801,
      "epoch": 0.01280263864552423,
      "grad_norm": 12.936646461486816,
      "learning_rate": 1.999903128935387e-06,
      "loss": 0.6215,
      "mean_token_accuracy": 0.806446696817875,
      "num_tokens": 4965259.0,
      "step": 4130
    },
    {
      "entropy": 1.8559681817889213,
      "epoch": 0.012833637770573926,
      "grad_norm": 14.09201431274414,
      "learning_rate": 2.0047466821660373e-06,
      "loss": 0.6436,
      "mean_token_accuracy": 0.8033310145139694,
      "num_tokens": 4976636.0,
      "step": 4140
    },
    {
      "entropy": 1.754169662296772,
      "epoch": 0.012864636895623621,
      "grad_norm": 6.45542573928833,
      "learning_rate": 2.009590235396687e-06,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8193281710147857,
      "num_tokens": 4989310.0,
      "step": 4150
    },
    {
      "entropy": 1.755221924185753,
      "epoch": 0.012895636020673316,
      "grad_norm": 6.969062328338623,
      "learning_rate": 2.014433788627337e-06,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8212425723671913,
      "num_tokens": 5001740.0,
      "step": 4160
    },
    {
      "entropy": 1.923591212928295,
      "epoch": 0.012926635145723012,
      "grad_norm": 16.79177474975586,
      "learning_rate": 2.0192773418579874e-06,
      "loss": 0.7115,
      "mean_token_accuracy": 0.7964399144053459,
      "num_tokens": 5012837.0,
      "step": 4170
    },
    {
      "entropy": 1.789121389389038,
      "epoch": 0.012957634270772707,
      "grad_norm": 6.560935974121094,
      "learning_rate": 2.0241208950886372e-06,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8209047600626945,
      "num_tokens": 5026059.0,
      "step": 4180
    },
    {
      "entropy": 1.8681050762534142,
      "epoch": 0.012988633395822402,
      "grad_norm": 13.890046119689941,
      "learning_rate": 2.028964448319287e-06,
      "loss": 0.614,
      "mean_token_accuracy": 0.7979257851839066,
      "num_tokens": 5038464.0,
      "step": 4190
    },
    {
      "entropy": 1.844995491206646,
      "epoch": 0.013019632520872099,
      "grad_norm": 13.706587791442871,
      "learning_rate": 2.0338080015499374e-06,
      "loss": 0.655,
      "mean_token_accuracy": 0.8055862948298455,
      "num_tokens": 5051036.0,
      "step": 4200
    },
    {
      "entropy": 1.8614793449640274,
      "epoch": 0.013050631645921794,
      "grad_norm": 14.581295013427734,
      "learning_rate": 2.0386515547805873e-06,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8066158786416053,
      "num_tokens": 5063162.0,
      "step": 4210
    },
    {
      "entropy": 1.8593030095100402,
      "epoch": 0.01308163077097149,
      "grad_norm": 13.553071022033691,
      "learning_rate": 2.043495108011237e-06,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8146178603172303,
      "num_tokens": 5075545.0,
      "step": 4220
    },
    {
      "entropy": 1.8848890259861946,
      "epoch": 0.013112629896021185,
      "grad_norm": 17.947132110595703,
      "learning_rate": 2.0483386612418875e-06,
      "loss": 0.6514,
      "mean_token_accuracy": 0.7945050925016404,
      "num_tokens": 5087372.0,
      "step": 4230
    },
    {
      "entropy": 1.880325546860695,
      "epoch": 0.01314362902107088,
      "grad_norm": 15.84373950958252,
      "learning_rate": 2.053182214472537e-06,
      "loss": 0.6546,
      "mean_token_accuracy": 0.8056602075695991,
      "num_tokens": 5098820.0,
      "step": 4240
    },
    {
      "entropy": 1.9198369443416596,
      "epoch": 0.013174628146120576,
      "grad_norm": 14.227551460266113,
      "learning_rate": 2.0580257677031873e-06,
      "loss": 0.6509,
      "mean_token_accuracy": 0.8100201249122619,
      "num_tokens": 5109708.0,
      "step": 4250
    },
    {
      "entropy": 1.95897875726223,
      "epoch": 0.013205627271170271,
      "grad_norm": 15.566633224487305,
      "learning_rate": 2.062869320933837e-06,
      "loss": 0.6949,
      "mean_token_accuracy": 0.7956610158085823,
      "num_tokens": 5121454.0,
      "step": 4260
    },
    {
      "entropy": 1.8369690001010894,
      "epoch": 0.013236626396219966,
      "grad_norm": 16.143478393554688,
      "learning_rate": 2.067712874164487e-06,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8159258916974068,
      "num_tokens": 5134141.0,
      "step": 4270
    },
    {
      "entropy": 1.912563818693161,
      "epoch": 0.013267625521269663,
      "grad_norm": 15.098668098449707,
      "learning_rate": 2.0725564273951374e-06,
      "loss": 0.6755,
      "mean_token_accuracy": 0.7963416025042533,
      "num_tokens": 5146353.0,
      "step": 4280
    },
    {
      "entropy": 1.9880844503641129,
      "epoch": 0.013298624646319358,
      "grad_norm": 14.18931770324707,
      "learning_rate": 2.0773999806257872e-06,
      "loss": 0.7678,
      "mean_token_accuracy": 0.7802196651697159,
      "num_tokens": 5157703.0,
      "step": 4290
    },
    {
      "entropy": 1.8342043563723565,
      "epoch": 0.013329623771369052,
      "grad_norm": 6.213775157928467,
      "learning_rate": 2.082243533856437e-06,
      "loss": 0.5523,
      "mean_token_accuracy": 0.8177143961191178,
      "num_tokens": 5170695.0,
      "step": 4300
    },
    {
      "entropy": 1.8414741292595864,
      "epoch": 0.013360622896418749,
      "grad_norm": 12.719599723815918,
      "learning_rate": 2.087087087087087e-06,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8134493753314018,
      "num_tokens": 5183018.0,
      "step": 4310
    },
    {
      "entropy": 1.8449064493179321,
      "epoch": 0.013391622021468444,
      "grad_norm": 12.744540214538574,
      "learning_rate": 2.0919306403177373e-06,
      "loss": 0.6141,
      "mean_token_accuracy": 0.7993644192814827,
      "num_tokens": 5195549.0,
      "step": 4320
    },
    {
      "entropy": 1.9135398477315904,
      "epoch": 0.013422621146518139,
      "grad_norm": 15.205565452575684,
      "learning_rate": 2.096774193548387e-06,
      "loss": 0.6636,
      "mean_token_accuracy": 0.8040877833962441,
      "num_tokens": 5207595.0,
      "step": 4330
    },
    {
      "entropy": 1.892987634241581,
      "epoch": 0.013453620271567835,
      "grad_norm": 14.970274925231934,
      "learning_rate": 2.101617746779037e-06,
      "loss": 0.6144,
      "mean_token_accuracy": 0.8050429001450539,
      "num_tokens": 5220012.0,
      "step": 4340
    },
    {
      "entropy": 1.933681371808052,
      "epoch": 0.01348461939661753,
      "grad_norm": 16.078292846679688,
      "learning_rate": 2.1064613000096874e-06,
      "loss": 0.6802,
      "mean_token_accuracy": 0.7953403264284133,
      "num_tokens": 5232144.0,
      "step": 4350
    },
    {
      "entropy": 1.8837040960788727,
      "epoch": 0.013515618521667225,
      "grad_norm": 12.591126441955566,
      "learning_rate": 2.1113048532403373e-06,
      "loss": 0.6592,
      "mean_token_accuracy": 0.8096277773380279,
      "num_tokens": 5244523.0,
      "step": 4360
    },
    {
      "entropy": 1.8900502383708955,
      "epoch": 0.013546617646716922,
      "grad_norm": 7.1288604736328125,
      "learning_rate": 2.116148406470987e-06,
      "loss": 0.6995,
      "mean_token_accuracy": 0.797728767991066,
      "num_tokens": 5255816.0,
      "step": 4370
    },
    {
      "entropy": 1.8959971249103547,
      "epoch": 0.013577616771766617,
      "grad_norm": 8.134716987609863,
      "learning_rate": 2.1209919597016375e-06,
      "loss": 0.6631,
      "mean_token_accuracy": 0.7931605547666549,
      "num_tokens": 5268217.0,
      "step": 4380
    },
    {
      "entropy": 1.8685301005840302,
      "epoch": 0.013608615896816313,
      "grad_norm": 14.050363540649414,
      "learning_rate": 2.1258355129322874e-06,
      "loss": 0.5861,
      "mean_token_accuracy": 0.7998355850577354,
      "num_tokens": 5280627.0,
      "step": 4390
    },
    {
      "entropy": 1.9573094069957733,
      "epoch": 0.013639615021866008,
      "grad_norm": 12.361193656921387,
      "learning_rate": 2.1306790661629372e-06,
      "loss": 0.6558,
      "mean_token_accuracy": 0.797396968305111,
      "num_tokens": 5291443.0,
      "step": 4400
    },
    {
      "entropy": 1.8597141653299332,
      "epoch": 0.013670614146915703,
      "grad_norm": 16.280147552490234,
      "learning_rate": 2.1355226193935875e-06,
      "loss": 0.6441,
      "mean_token_accuracy": 0.7984641641378403,
      "num_tokens": 5303414.0,
      "step": 4410
    },
    {
      "entropy": 1.8121001735329627,
      "epoch": 0.0137016132719654,
      "grad_norm": 19.821773529052734,
      "learning_rate": 2.1403661726242374e-06,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8085174813866616,
      "num_tokens": 5316024.0,
      "step": 4420
    },
    {
      "entropy": 1.8680142611265182,
      "epoch": 0.013732612397015094,
      "grad_norm": 14.235099792480469,
      "learning_rate": 2.1452097258548873e-06,
      "loss": 0.6586,
      "mean_token_accuracy": 0.7996479585766793,
      "num_tokens": 5327955.0,
      "step": 4430
    },
    {
      "entropy": 1.8850964441895486,
      "epoch": 0.01376361152206479,
      "grad_norm": 12.817394256591797,
      "learning_rate": 2.1500532790855376e-06,
      "loss": 0.6506,
      "mean_token_accuracy": 0.7959688842296601,
      "num_tokens": 5340123.0,
      "step": 4440
    },
    {
      "entropy": 1.9236231684684753,
      "epoch": 0.013794610647114486,
      "grad_norm": 13.93416976928711,
      "learning_rate": 2.1548968323161875e-06,
      "loss": 0.6866,
      "mean_token_accuracy": 0.8000624299049377,
      "num_tokens": 5351902.0,
      "step": 4450
    },
    {
      "entropy": 1.8246183797717095,
      "epoch": 0.01382560977216418,
      "grad_norm": 6.33852481842041,
      "learning_rate": 2.1597403855468374e-06,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8101627513766289,
      "num_tokens": 5364681.0,
      "step": 4460
    },
    {
      "entropy": 1.896468523144722,
      "epoch": 0.013856608897213876,
      "grad_norm": 15.372577667236328,
      "learning_rate": 2.1645839387774873e-06,
      "loss": 0.6683,
      "mean_token_accuracy": 0.8046876505017281,
      "num_tokens": 5376647.0,
      "step": 4470
    },
    {
      "entropy": 1.8945748567581178,
      "epoch": 0.013887608022263572,
      "grad_norm": 14.416013717651367,
      "learning_rate": 2.169427492008137e-06,
      "loss": 0.6216,
      "mean_token_accuracy": 0.8085326358675957,
      "num_tokens": 5388517.0,
      "step": 4480
    },
    {
      "entropy": 1.9351640045642853,
      "epoch": 0.013918607147313267,
      "grad_norm": 11.383118629455566,
      "learning_rate": 2.1742710452387875e-06,
      "loss": 0.7617,
      "mean_token_accuracy": 0.7835462704300881,
      "num_tokens": 5399458.0,
      "step": 4490
    },
    {
      "entropy": 1.873738704621792,
      "epoch": 0.013949606272362962,
      "grad_norm": 14.66506576538086,
      "learning_rate": 2.1791145984694373e-06,
      "loss": 0.6014,
      "mean_token_accuracy": 0.8175080880522728,
      "num_tokens": 5411219.0,
      "step": 4500
    },
    {
      "entropy": 1.9110942378640174,
      "epoch": 0.013980605397412658,
      "grad_norm": 7.845452308654785,
      "learning_rate": 2.1839581517000872e-06,
      "loss": 0.6462,
      "mean_token_accuracy": 0.799024523794651,
      "num_tokens": 5423441.0,
      "step": 4510
    },
    {
      "entropy": 1.8728701308369637,
      "epoch": 0.014011604522462353,
      "grad_norm": 14.299894332885742,
      "learning_rate": 2.188801704930737e-06,
      "loss": 0.6333,
      "mean_token_accuracy": 0.7987611889839172,
      "num_tokens": 5435798.0,
      "step": 4520
    },
    {
      "entropy": 1.9454208359122276,
      "epoch": 0.014042603647512048,
      "grad_norm": 14.531904220581055,
      "learning_rate": 2.1936452581613874e-06,
      "loss": 0.6534,
      "mean_token_accuracy": 0.8002553582191467,
      "num_tokens": 5447898.0,
      "step": 4530
    },
    {
      "entropy": 1.93377585709095,
      "epoch": 0.014073602772561745,
      "grad_norm": 12.258691787719727,
      "learning_rate": 2.1984888113920373e-06,
      "loss": 0.6732,
      "mean_token_accuracy": 0.7952050551772117,
      "num_tokens": 5459752.0,
      "step": 4540
    },
    {
      "entropy": 1.9374273508787154,
      "epoch": 0.01410460189761144,
      "grad_norm": 6.370416641235352,
      "learning_rate": 2.203332364622687e-06,
      "loss": 0.6438,
      "mean_token_accuracy": 0.8051278084516525,
      "num_tokens": 5470921.0,
      "step": 4550
    },
    {
      "entropy": 1.9412765011191369,
      "epoch": 0.014135601022661136,
      "grad_norm": 6.380893230438232,
      "learning_rate": 2.2081759178533375e-06,
      "loss": 0.6152,
      "mean_token_accuracy": 0.811233189702034,
      "num_tokens": 5482832.0,
      "step": 4560
    },
    {
      "entropy": 1.954646387696266,
      "epoch": 0.014166600147710831,
      "grad_norm": 11.699299812316895,
      "learning_rate": 2.2130194710839874e-06,
      "loss": 0.6937,
      "mean_token_accuracy": 0.7911556661128998,
      "num_tokens": 5494419.0,
      "step": 4570
    },
    {
      "entropy": 1.8840202033519744,
      "epoch": 0.014197599272760526,
      "grad_norm": 14.447190284729004,
      "learning_rate": 2.2178630243146373e-06,
      "loss": 0.5864,
      "mean_token_accuracy": 0.8123130038380623,
      "num_tokens": 5507775.0,
      "step": 4580
    },
    {
      "entropy": 1.9366067603230477,
      "epoch": 0.014228598397810223,
      "grad_norm": 6.459699630737305,
      "learning_rate": 2.2227065775452876e-06,
      "loss": 0.6687,
      "mean_token_accuracy": 0.7912245571613312,
      "num_tokens": 5519641.0,
      "step": 4590
    },
    {
      "entropy": 1.8860922396183013,
      "epoch": 0.014259597522859917,
      "grad_norm": 7.023352146148682,
      "learning_rate": 2.2275501307759375e-06,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8207052007317543,
      "num_tokens": 5532518.0,
      "step": 4600
    },
    {
      "entropy": 1.8792527481913566,
      "epoch": 0.014290596647909612,
      "grad_norm": 14.053731918334961,
      "learning_rate": 2.2323936840065873e-06,
      "loss": 0.6267,
      "mean_token_accuracy": 0.8102740809321404,
      "num_tokens": 5545434.0,
      "step": 4610
    },
    {
      "entropy": 1.8577488988637925,
      "epoch": 0.014321595772959309,
      "grad_norm": 13.818965911865234,
      "learning_rate": 2.2372372372372376e-06,
      "loss": 0.607,
      "mean_token_accuracy": 0.8134890556335449,
      "num_tokens": 5557659.0,
      "step": 4620
    },
    {
      "entropy": 1.8798778101801872,
      "epoch": 0.014352594898009004,
      "grad_norm": 15.251192092895508,
      "learning_rate": 2.2420807904678875e-06,
      "loss": 0.7015,
      "mean_token_accuracy": 0.8011739581823349,
      "num_tokens": 5569313.0,
      "step": 4630
    },
    {
      "entropy": 1.9401899367570876,
      "epoch": 0.014383594023058699,
      "grad_norm": 14.222334861755371,
      "learning_rate": 2.2469243436985374e-06,
      "loss": 0.6341,
      "mean_token_accuracy": 0.8067525029182434,
      "num_tokens": 5580286.0,
      "step": 4640
    },
    {
      "entropy": 1.791366559267044,
      "epoch": 0.014414593148108395,
      "grad_norm": 15.209399223327637,
      "learning_rate": 2.2517678969291877e-06,
      "loss": 0.621,
      "mean_token_accuracy": 0.8049216285347939,
      "num_tokens": 5593090.0,
      "step": 4650
    },
    {
      "entropy": 1.8601211935281754,
      "epoch": 0.01444559227315809,
      "grad_norm": 6.989866733551025,
      "learning_rate": 2.2566114501598376e-06,
      "loss": 0.6226,
      "mean_token_accuracy": 0.7975085467100144,
      "num_tokens": 5606291.0,
      "step": 4660
    },
    {
      "entropy": 1.8646612569689751,
      "epoch": 0.014476591398207785,
      "grad_norm": 6.809642314910889,
      "learning_rate": 2.2614550033904875e-06,
      "loss": 0.6267,
      "mean_token_accuracy": 0.801960551738739,
      "num_tokens": 5618542.0,
      "step": 4670
    },
    {
      "entropy": 1.8657331004738809,
      "epoch": 0.014507590523257482,
      "grad_norm": 15.695577621459961,
      "learning_rate": 2.2662985566211374e-06,
      "loss": 0.6266,
      "mean_token_accuracy": 0.8036523833870888,
      "num_tokens": 5630622.0,
      "step": 4680
    },
    {
      "entropy": 1.8683514580130578,
      "epoch": 0.014538589648307176,
      "grad_norm": 14.665810585021973,
      "learning_rate": 2.2711421098517873e-06,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8168015897274017,
      "num_tokens": 5642626.0,
      "step": 4690
    },
    {
      "entropy": 1.910755640268326,
      "epoch": 0.014569588773356871,
      "grad_norm": 14.815983772277832,
      "learning_rate": 2.2759856630824376e-06,
      "loss": 0.6303,
      "mean_token_accuracy": 0.7996454507112503,
      "num_tokens": 5654704.0,
      "step": 4700
    },
    {
      "entropy": 1.9073377847671509,
      "epoch": 0.014600587898406568,
      "grad_norm": 12.390499114990234,
      "learning_rate": 2.2808292163130874e-06,
      "loss": 0.6777,
      "mean_token_accuracy": 0.7973004877567291,
      "num_tokens": 5666771.0,
      "step": 4710
    },
    {
      "entropy": 1.8886594474315643,
      "epoch": 0.014631587023456263,
      "grad_norm": 15.042806625366211,
      "learning_rate": 2.2856727695437373e-06,
      "loss": 0.666,
      "mean_token_accuracy": 0.7995603799819946,
      "num_tokens": 5679125.0,
      "step": 4720
    },
    {
      "entropy": 1.8458405777812004,
      "epoch": 0.014662586148505958,
      "grad_norm": 13.178945541381836,
      "learning_rate": 2.2905163227743872e-06,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8050804138183594,
      "num_tokens": 5691263.0,
      "step": 4730
    },
    {
      "entropy": 1.8889022842049599,
      "epoch": 0.014693585273555654,
      "grad_norm": 13.8629150390625,
      "learning_rate": 2.2953598760050375e-06,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8008949771523476,
      "num_tokens": 5703903.0,
      "step": 4740
    },
    {
      "entropy": 1.8346245408058166,
      "epoch": 0.014724584398605349,
      "grad_norm": 13.840920448303223,
      "learning_rate": 2.3002034292356874e-06,
      "loss": 0.6284,
      "mean_token_accuracy": 0.8062951177358627,
      "num_tokens": 5715573.0,
      "step": 4750
    },
    {
      "entropy": 1.797354480624199,
      "epoch": 0.014755583523655046,
      "grad_norm": 7.390272617340088,
      "learning_rate": 2.3050469824663373e-06,
      "loss": 0.5823,
      "mean_token_accuracy": 0.81457539498806,
      "num_tokens": 5728120.0,
      "step": 4760
    },
    {
      "entropy": 1.787947428226471,
      "epoch": 0.01478658264870474,
      "grad_norm": 13.413910865783691,
      "learning_rate": 2.3098905356969876e-06,
      "loss": 0.6103,
      "mean_token_accuracy": 0.8159710958600044,
      "num_tokens": 5739981.0,
      "step": 4770
    },
    {
      "entropy": 1.7457917869091033,
      "epoch": 0.014817581773754435,
      "grad_norm": 14.069392204284668,
      "learning_rate": 2.3147340889276375e-06,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8245069414377213,
      "num_tokens": 5753136.0,
      "step": 4780
    },
    {
      "entropy": 1.8335485979914665,
      "epoch": 0.014848580898804132,
      "grad_norm": 12.345169067382812,
      "learning_rate": 2.3195776421582874e-06,
      "loss": 0.6593,
      "mean_token_accuracy": 0.8006348922848702,
      "num_tokens": 5765376.0,
      "step": 4790
    },
    {
      "entropy": 1.8139121234416962,
      "epoch": 0.014879580023853827,
      "grad_norm": 15.01768970489502,
      "learning_rate": 2.3244211953889377e-06,
      "loss": 0.6159,
      "mean_token_accuracy": 0.8078545331954956,
      "num_tokens": 5778012.0,
      "step": 4800
    },
    {
      "entropy": 1.8669386252760887,
      "epoch": 0.014910579148903522,
      "grad_norm": 14.147043228149414,
      "learning_rate": 2.3292647486195876e-06,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8018081709742546,
      "num_tokens": 5789805.0,
      "step": 4810
    },
    {
      "entropy": 1.784064681828022,
      "epoch": 0.014941578273953218,
      "grad_norm": 13.715585708618164,
      "learning_rate": 2.3341083018502374e-06,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8130191281437874,
      "num_tokens": 5802546.0,
      "step": 4820
    },
    {
      "entropy": 1.7632321387529373,
      "epoch": 0.014972577399002913,
      "grad_norm": 14.334385871887207,
      "learning_rate": 2.3389518550808878e-06,
      "loss": 0.6187,
      "mean_token_accuracy": 0.8080756813287735,
      "num_tokens": 5815760.0,
      "step": 4830
    },
    {
      "entropy": 1.826922358572483,
      "epoch": 0.015003576524052608,
      "grad_norm": 12.539878845214844,
      "learning_rate": 2.3437954083115376e-06,
      "loss": 0.6262,
      "mean_token_accuracy": 0.8105258822441102,
      "num_tokens": 5827438.0,
      "step": 4840
    },
    {
      "entropy": 1.788956308364868,
      "epoch": 0.015034575649102305,
      "grad_norm": 11.371108055114746,
      "learning_rate": 2.3486389615421875e-06,
      "loss": 0.5732,
      "mean_token_accuracy": 0.817362517118454,
      "num_tokens": 5839410.0,
      "step": 4850
    },
    {
      "entropy": 1.734141993522644,
      "epoch": 0.015065574774152,
      "grad_norm": 6.056836128234863,
      "learning_rate": 2.353482514772838e-06,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8292117938399315,
      "num_tokens": 5852222.0,
      "step": 4860
    },
    {
      "entropy": 1.9274750858545304,
      "epoch": 0.015096573899201694,
      "grad_norm": 14.039139747619629,
      "learning_rate": 2.3583260680034877e-06,
      "loss": 0.6941,
      "mean_token_accuracy": 0.7952279299497604,
      "num_tokens": 5864055.0,
      "step": 4870
    },
    {
      "entropy": 1.8699121579527855,
      "epoch": 0.015127573024251391,
      "grad_norm": 15.34942626953125,
      "learning_rate": 2.3631696212341376e-06,
      "loss": 0.669,
      "mean_token_accuracy": 0.7950506001710892,
      "num_tokens": 5876040.0,
      "step": 4880
    },
    {
      "entropy": 1.9636721938848496,
      "epoch": 0.015158572149301086,
      "grad_norm": 16.529897689819336,
      "learning_rate": 2.3680131744647875e-06,
      "loss": 0.7517,
      "mean_token_accuracy": 0.7848048597574234,
      "num_tokens": 5887191.0,
      "step": 4890
    },
    {
      "entropy": 1.7576201945543288,
      "epoch": 0.01518957127435078,
      "grad_norm": 11.712950706481934,
      "learning_rate": 2.3728567276954374e-06,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8087734043598175,
      "num_tokens": 5901313.0,
      "step": 4900
    },
    {
      "entropy": 1.8790557369589806,
      "epoch": 0.015220570399400477,
      "grad_norm": 13.049158096313477,
      "learning_rate": 2.3777002809260877e-06,
      "loss": 0.6831,
      "mean_token_accuracy": 0.7949941202998161,
      "num_tokens": 5913169.0,
      "step": 4910
    },
    {
      "entropy": 1.8304862260818482,
      "epoch": 0.015251569524450172,
      "grad_norm": 14.758519172668457,
      "learning_rate": 2.3825438341567376e-06,
      "loss": 0.6111,
      "mean_token_accuracy": 0.7983208373188972,
      "num_tokens": 5925901.0,
      "step": 4920
    },
    {
      "entropy": 1.7591045543551445,
      "epoch": 0.015282568649499869,
      "grad_norm": 7.14612340927124,
      "learning_rate": 2.3873873873873874e-06,
      "loss": 0.608,
      "mean_token_accuracy": 0.8068576440215111,
      "num_tokens": 5939642.0,
      "step": 4930
    },
    {
      "entropy": 1.9316822454333304,
      "epoch": 0.015313567774549564,
      "grad_norm": 11.606863021850586,
      "learning_rate": 2.3922309406180373e-06,
      "loss": 0.6747,
      "mean_token_accuracy": 0.7950692802667618,
      "num_tokens": 5951806.0,
      "step": 4940
    },
    {
      "entropy": 1.8316121339797973,
      "epoch": 0.015344566899599258,
      "grad_norm": 12.468024253845215,
      "learning_rate": 2.3970744938486876e-06,
      "loss": 0.6445,
      "mean_token_accuracy": 0.7969418242573738,
      "num_tokens": 5963560.0,
      "step": 4950
    },
    {
      "entropy": 1.8592291116714477,
      "epoch": 0.015375566024648955,
      "grad_norm": 10.784393310546875,
      "learning_rate": 2.4019180470793375e-06,
      "loss": 0.584,
      "mean_token_accuracy": 0.815529865026474,
      "num_tokens": 5975576.0,
      "step": 4960
    },
    {
      "entropy": 1.8554728999733925,
      "epoch": 0.01540656514969865,
      "grad_norm": 6.568962574005127,
      "learning_rate": 2.4067616003099874e-06,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8106912553310395,
      "num_tokens": 5987431.0,
      "step": 4970
    },
    {
      "entropy": 1.9579525411128997,
      "epoch": 0.015437564274748345,
      "grad_norm": 15.040802001953125,
      "learning_rate": 2.4116051535406377e-06,
      "loss": 0.7144,
      "mean_token_accuracy": 0.7940159112215042,
      "num_tokens": 5998986.0,
      "step": 4980
    },
    {
      "entropy": 1.8558162599802017,
      "epoch": 0.015468563399798041,
      "grad_norm": 12.90155029296875,
      "learning_rate": 2.4164487067712876e-06,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8087948963046074,
      "num_tokens": 6011316.0,
      "step": 4990
    },
    {
      "entropy": 1.9120135977864265,
      "epoch": 0.015499562524847736,
      "grad_norm": 13.731481552124023,
      "learning_rate": 2.4212922600019375e-06,
      "loss": 0.6358,
      "mean_token_accuracy": 0.808275742828846,
      "num_tokens": 6022791.0,
      "step": 5000
    },
    {
      "entropy": 1.9652427747845649,
      "epoch": 0.015530561649897431,
      "grad_norm": 10.624104499816895,
      "learning_rate": 2.4261358132325878e-06,
      "loss": 0.7283,
      "mean_token_accuracy": 0.7885913565754891,
      "num_tokens": 6034662.0,
      "step": 5010
    },
    {
      "entropy": 1.8572456985712051,
      "epoch": 0.015561560774947128,
      "grad_norm": 13.485920906066895,
      "learning_rate": 2.4309793664632377e-06,
      "loss": 0.6653,
      "mean_token_accuracy": 0.8142799854278564,
      "num_tokens": 6046637.0,
      "step": 5020
    },
    {
      "entropy": 1.8741802372038365,
      "epoch": 0.015592559899996822,
      "grad_norm": 13.382155418395996,
      "learning_rate": 2.4358229196938875e-06,
      "loss": 0.673,
      "mean_token_accuracy": 0.7978162422776223,
      "num_tokens": 6058910.0,
      "step": 5030
    },
    {
      "entropy": 1.8276728346943856,
      "epoch": 0.015623559025046517,
      "grad_norm": 13.904196739196777,
      "learning_rate": 2.440666472924538e-06,
      "loss": 0.6098,
      "mean_token_accuracy": 0.80892014503479,
      "num_tokens": 6071136.0,
      "step": 5040
    },
    {
      "entropy": 1.8636898145079612,
      "epoch": 0.015654558150096212,
      "grad_norm": 12.65003490447998,
      "learning_rate": 2.4455100261551877e-06,
      "loss": 0.6052,
      "mean_token_accuracy": 0.8113350749015809,
      "num_tokens": 6082834.0,
      "step": 5050
    },
    {
      "entropy": 1.9150115132331849,
      "epoch": 0.01568555727514591,
      "grad_norm": 12.964512825012207,
      "learning_rate": 2.4503535793858376e-06,
      "loss": 0.6974,
      "mean_token_accuracy": 0.7987506031990051,
      "num_tokens": 6093808.0,
      "step": 5060
    },
    {
      "entropy": 1.8618119135499,
      "epoch": 0.015716556400195605,
      "grad_norm": 5.460781097412109,
      "learning_rate": 2.455197132616488e-06,
      "loss": 0.6885,
      "mean_token_accuracy": 0.8034725025296211,
      "num_tokens": 6105718.0,
      "step": 5070
    },
    {
      "entropy": 1.9206832945346832,
      "epoch": 0.0157475555252453,
      "grad_norm": 6.1644134521484375,
      "learning_rate": 2.460040685847138e-06,
      "loss": 0.6132,
      "mean_token_accuracy": 0.8128495365381241,
      "num_tokens": 6117606.0,
      "step": 5080
    },
    {
      "entropy": 1.8482828214764595,
      "epoch": 0.015778554650294995,
      "grad_norm": 16.253582000732422,
      "learning_rate": 2.4648842390777877e-06,
      "loss": 0.5834,
      "mean_token_accuracy": 0.8172304585576058,
      "num_tokens": 6130415.0,
      "step": 5090
    },
    {
      "entropy": 1.998060804605484,
      "epoch": 0.01580955377534469,
      "grad_norm": 12.287386894226074,
      "learning_rate": 2.4697277923084376e-06,
      "loss": 0.7151,
      "mean_token_accuracy": 0.8040244802832603,
      "num_tokens": 6140984.0,
      "step": 5100
    },
    {
      "entropy": 1.8899462610483169,
      "epoch": 0.015840552900394385,
      "grad_norm": 14.631074905395508,
      "learning_rate": 2.4745713455390875e-06,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8050005823373795,
      "num_tokens": 6152531.0,
      "step": 5110
    },
    {
      "entropy": 1.9535317480564118,
      "epoch": 0.01587155202544408,
      "grad_norm": 17.399723052978516,
      "learning_rate": 2.4794148987697378e-06,
      "loss": 0.714,
      "mean_token_accuracy": 0.7943957537412644,
      "num_tokens": 6163431.0,
      "step": 5120
    },
    {
      "entropy": 1.8350759640336036,
      "epoch": 0.015902551150493778,
      "grad_norm": 6.640737056732178,
      "learning_rate": 2.4842584520003877e-06,
      "loss": 0.5901,
      "mean_token_accuracy": 0.8108617261052131,
      "num_tokens": 6176031.0,
      "step": 5130
    },
    {
      "entropy": 1.8753447353839874,
      "epoch": 0.01593355027554347,
      "grad_norm": 12.087038040161133,
      "learning_rate": 2.4891020052310375e-06,
      "loss": 0.6479,
      "mean_token_accuracy": 0.8096778213977813,
      "num_tokens": 6187445.0,
      "step": 5140
    },
    {
      "entropy": 1.9582001134753226,
      "epoch": 0.015964549400593168,
      "grad_norm": 14.169228553771973,
      "learning_rate": 2.4939455584616874e-06,
      "loss": 0.6759,
      "mean_token_accuracy": 0.8033049464225769,
      "num_tokens": 6199080.0,
      "step": 5150
    },
    {
      "entropy": 1.9630192652344705,
      "epoch": 0.015995548525642864,
      "grad_norm": 15.533053398132324,
      "learning_rate": 2.4987891116923377e-06,
      "loss": 0.6722,
      "mean_token_accuracy": 0.7981782972812652,
      "num_tokens": 6210218.0,
      "step": 5160
    },
    {
      "entropy": 1.9423971146345138,
      "epoch": 0.016026547650692557,
      "grad_norm": 15.028423309326172,
      "learning_rate": 2.5036326649229876e-06,
      "loss": 0.6986,
      "mean_token_accuracy": 0.7927390620112419,
      "num_tokens": 6221926.0,
      "step": 5170
    },
    {
      "entropy": 1.9331501245498657,
      "epoch": 0.016057546775742254,
      "grad_norm": 14.322336196899414,
      "learning_rate": 2.5084762181536375e-06,
      "loss": 0.669,
      "mean_token_accuracy": 0.7981711998581886,
      "num_tokens": 6233615.0,
      "step": 5180
    },
    {
      "entropy": 1.920173704624176,
      "epoch": 0.01608854590079195,
      "grad_norm": 13.353188514709473,
      "learning_rate": 2.513319771384288e-06,
      "loss": 0.6021,
      "mean_token_accuracy": 0.804129633307457,
      "num_tokens": 6245365.0,
      "step": 5190
    },
    {
      "entropy": 1.9042732536792755,
      "epoch": 0.016119545025841647,
      "grad_norm": 6.624873161315918,
      "learning_rate": 2.5181633246149377e-06,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8022510379552841,
      "num_tokens": 6257307.0,
      "step": 5200
    },
    {
      "entropy": 1.9409416317939758,
      "epoch": 0.01615054415089134,
      "grad_norm": 16.728931427001953,
      "learning_rate": 2.523006877845588e-06,
      "loss": 0.6694,
      "mean_token_accuracy": 0.8007285013794899,
      "num_tokens": 6268908.0,
      "step": 5210
    },
    {
      "entropy": 1.9607915192842484,
      "epoch": 0.016181543275941037,
      "grad_norm": 12.1609525680542,
      "learning_rate": 2.5278504310762375e-06,
      "loss": 0.6902,
      "mean_token_accuracy": 0.7996085345745086,
      "num_tokens": 6279792.0,
      "step": 5220
    },
    {
      "entropy": 1.9298648953437805,
      "epoch": 0.016212542400990734,
      "grad_norm": 14.782210350036621,
      "learning_rate": 2.5326939843068878e-06,
      "loss": 0.6543,
      "mean_token_accuracy": 0.7992920339107513,
      "num_tokens": 6291405.0,
      "step": 5230
    },
    {
      "entropy": 1.8961961045861244,
      "epoch": 0.016243541526040427,
      "grad_norm": 15.001334190368652,
      "learning_rate": 2.5375375375375377e-06,
      "loss": 0.619,
      "mean_token_accuracy": 0.8142734676599502,
      "num_tokens": 6303357.0,
      "step": 5240
    },
    {
      "entropy": 1.8733942970633506,
      "epoch": 0.016274540651090123,
      "grad_norm": 11.908313751220703,
      "learning_rate": 2.542381090768188e-06,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8036831006407738,
      "num_tokens": 6316147.0,
      "step": 5250
    },
    {
      "entropy": 1.8637539759278297,
      "epoch": 0.01630553977613982,
      "grad_norm": 13.72813606262207,
      "learning_rate": 2.547224643998838e-06,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8079806834459304,
      "num_tokens": 6328593.0,
      "step": 5260
    },
    {
      "entropy": 1.9503967970609666,
      "epoch": 0.016336538901189513,
      "grad_norm": 13.402142524719238,
      "learning_rate": 2.5520681972294873e-06,
      "loss": 0.7049,
      "mean_token_accuracy": 0.7919132471084595,
      "num_tokens": 6340259.0,
      "step": 5270
    },
    {
      "entropy": 1.8650514677166938,
      "epoch": 0.01636753802623921,
      "grad_norm": 17.254850387573242,
      "learning_rate": 2.5569117504601376e-06,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8054231390357017,
      "num_tokens": 6351914.0,
      "step": 5280
    },
    {
      "entropy": 1.861028863489628,
      "epoch": 0.016398537151288906,
      "grad_norm": 14.428080558776855,
      "learning_rate": 2.561755303690788e-06,
      "loss": 0.681,
      "mean_token_accuracy": 0.7947863683104515,
      "num_tokens": 6364405.0,
      "step": 5290
    },
    {
      "entropy": 1.9019827499985695,
      "epoch": 0.0164295362763386,
      "grad_norm": 14.161713600158691,
      "learning_rate": 2.566598856921438e-06,
      "loss": 0.6824,
      "mean_token_accuracy": 0.8035579726099968,
      "num_tokens": 6375673.0,
      "step": 5300
    },
    {
      "entropy": 1.8992380529642106,
      "epoch": 0.016460535401388296,
      "grad_norm": 13.563090324401855,
      "learning_rate": 2.571442410152088e-06,
      "loss": 0.6121,
      "mean_token_accuracy": 0.8122377499938012,
      "num_tokens": 6387160.0,
      "step": 5310
    },
    {
      "entropy": 1.8584588319063187,
      "epoch": 0.016491534526437993,
      "grad_norm": 7.150653839111328,
      "learning_rate": 2.576285963382738e-06,
      "loss": 0.6296,
      "mean_token_accuracy": 0.8052732735872269,
      "num_tokens": 6399189.0,
      "step": 5320
    },
    {
      "entropy": 1.822690936923027,
      "epoch": 0.016522533651487686,
      "grad_norm": 13.757638931274414,
      "learning_rate": 2.5811295166133875e-06,
      "loss": 0.6489,
      "mean_token_accuracy": 0.8050631493330002,
      "num_tokens": 6411077.0,
      "step": 5330
    },
    {
      "entropy": 1.86706335991621,
      "epoch": 0.016553532776537382,
      "grad_norm": 12.249866485595703,
      "learning_rate": 2.5859730698440378e-06,
      "loss": 0.6379,
      "mean_token_accuracy": 0.8055833503603935,
      "num_tokens": 6422795.0,
      "step": 5340
    },
    {
      "entropy": 1.914227731525898,
      "epoch": 0.01658453190158708,
      "grad_norm": 13.125116348266602,
      "learning_rate": 2.5908166230746876e-06,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8045208930969239,
      "num_tokens": 6434827.0,
      "step": 5350
    },
    {
      "entropy": 1.8957944259047508,
      "epoch": 0.016615531026636772,
      "grad_norm": 16.094440460205078,
      "learning_rate": 2.595660176305338e-06,
      "loss": 0.6912,
      "mean_token_accuracy": 0.8000255629420281,
      "num_tokens": 6446181.0,
      "step": 5360
    },
    {
      "entropy": 1.8262048691511155,
      "epoch": 0.01664653015168647,
      "grad_norm": 7.709341526031494,
      "learning_rate": 2.6005037295359883e-06,
      "loss": 0.5977,
      "mean_token_accuracy": 0.8138811901211739,
      "num_tokens": 6458511.0,
      "step": 5370
    },
    {
      "entropy": 1.9274291083216668,
      "epoch": 0.016677529276736165,
      "grad_norm": 15.519542694091797,
      "learning_rate": 2.6053472827666377e-06,
      "loss": 0.7491,
      "mean_token_accuracy": 0.7906691998243331,
      "num_tokens": 6470175.0,
      "step": 5380
    },
    {
      "entropy": 1.8656795375049113,
      "epoch": 0.01670852840178586,
      "grad_norm": 15.985265731811523,
      "learning_rate": 2.6101908359972876e-06,
      "loss": 0.6216,
      "mean_token_accuracy": 0.8056905150413514,
      "num_tokens": 6482873.0,
      "step": 5390
    },
    {
      "entropy": 1.9562640219926835,
      "epoch": 0.016739527526835555,
      "grad_norm": 12.090657234191895,
      "learning_rate": 2.615034389227938e-06,
      "loss": 0.6927,
      "mean_token_accuracy": 0.7986342236399651,
      "num_tokens": 6494045.0,
      "step": 5400
    },
    {
      "entropy": 1.919235098361969,
      "epoch": 0.01677052665188525,
      "grad_norm": 12.741799354553223,
      "learning_rate": 2.619877942458588e-06,
      "loss": 0.6563,
      "mean_token_accuracy": 0.8064717799425125,
      "num_tokens": 6504934.0,
      "step": 5410
    },
    {
      "entropy": 1.8321070238947867,
      "epoch": 0.016801525776934945,
      "grad_norm": 13.527047157287598,
      "learning_rate": 2.624721495689238e-06,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8185374692082406,
      "num_tokens": 6517746.0,
      "step": 5420
    },
    {
      "entropy": 1.8565064251422883,
      "epoch": 0.01683252490198464,
      "grad_norm": 6.626979351043701,
      "learning_rate": 2.6295650489198876e-06,
      "loss": 0.6632,
      "mean_token_accuracy": 0.801714438199997,
      "num_tokens": 6530512.0,
      "step": 5430
    },
    {
      "entropy": 1.815597450733185,
      "epoch": 0.016863524027034338,
      "grad_norm": 14.231537818908691,
      "learning_rate": 2.634408602150538e-06,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8043477773666382,
      "num_tokens": 6543281.0,
      "step": 5440
    },
    {
      "entropy": 1.847208908200264,
      "epoch": 0.01689452315208403,
      "grad_norm": 14.332513809204102,
      "learning_rate": 2.6392521553811878e-06,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8129609242081642,
      "num_tokens": 6556103.0,
      "step": 5450
    },
    {
      "entropy": 1.8287439972162247,
      "epoch": 0.016925522277133728,
      "grad_norm": 17.275390625,
      "learning_rate": 2.644095708611838e-06,
      "loss": 0.5664,
      "mean_token_accuracy": 0.8119298800826072,
      "num_tokens": 6568831.0,
      "step": 5460
    },
    {
      "entropy": 1.9554360315203667,
      "epoch": 0.016956521402183424,
      "grad_norm": 16.211164474487305,
      "learning_rate": 2.648939261842488e-06,
      "loss": 0.7553,
      "mean_token_accuracy": 0.790603817999363,
      "num_tokens": 6580730.0,
      "step": 5470
    },
    {
      "entropy": 1.8777785643935203,
      "epoch": 0.016987520527233117,
      "grad_norm": 7.529537677764893,
      "learning_rate": 2.6537828150731374e-06,
      "loss": 0.5822,
      "mean_token_accuracy": 0.81802958548069,
      "num_tokens": 6593151.0,
      "step": 5480
    },
    {
      "entropy": 1.83996739089489,
      "epoch": 0.017018519652282814,
      "grad_norm": 16.056753158569336,
      "learning_rate": 2.6586263683037877e-06,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8094681099057197,
      "num_tokens": 6605508.0,
      "step": 5490
    },
    {
      "entropy": 1.8156257584691047,
      "epoch": 0.01704951877733251,
      "grad_norm": 13.18197250366211,
      "learning_rate": 2.663469921534438e-06,
      "loss": 0.5445,
      "mean_token_accuracy": 0.818387684226036,
      "num_tokens": 6617913.0,
      "step": 5500
    },
    {
      "entropy": 1.830448153614998,
      "epoch": 0.017080517902382204,
      "grad_norm": 17.966718673706055,
      "learning_rate": 2.668313474765088e-06,
      "loss": 0.591,
      "mean_token_accuracy": 0.8055599793791771,
      "num_tokens": 6630804.0,
      "step": 5510
    },
    {
      "entropy": 1.9519992083311082,
      "epoch": 0.0171115170274319,
      "grad_norm": 15.51613998413086,
      "learning_rate": 2.6731570279957382e-06,
      "loss": 0.7094,
      "mean_token_accuracy": 0.7885363206267357,
      "num_tokens": 6641484.0,
      "step": 5520
    },
    {
      "entropy": 1.787497617304325,
      "epoch": 0.017142516152481597,
      "grad_norm": 13.95913028717041,
      "learning_rate": 2.678000581226388e-06,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8152831554412842,
      "num_tokens": 6655013.0,
      "step": 5530
    },
    {
      "entropy": 1.8791037619113922,
      "epoch": 0.017173515277531293,
      "grad_norm": 12.858197212219238,
      "learning_rate": 2.6828441344570376e-06,
      "loss": 0.708,
      "mean_token_accuracy": 0.7970382794737816,
      "num_tokens": 6665889.0,
      "step": 5540
    },
    {
      "entropy": 1.8964224010705948,
      "epoch": 0.017204514402580987,
      "grad_norm": 14.71556282043457,
      "learning_rate": 2.687687687687688e-06,
      "loss": 0.7067,
      "mean_token_accuracy": 0.8008865773677826,
      "num_tokens": 6676444.0,
      "step": 5550
    },
    {
      "entropy": 1.868991169333458,
      "epoch": 0.017235513527630683,
      "grad_norm": 12.040255546569824,
      "learning_rate": 2.6925312409183378e-06,
      "loss": 0.6944,
      "mean_token_accuracy": 0.7887442111968994,
      "num_tokens": 6687681.0,
      "step": 5560
    },
    {
      "entropy": 1.7748029723763465,
      "epoch": 0.01726651265268038,
      "grad_norm": 16.154150009155273,
      "learning_rate": 2.697374794148988e-06,
      "loss": 0.558,
      "mean_token_accuracy": 0.818262355029583,
      "num_tokens": 6700056.0,
      "step": 5570
    },
    {
      "entropy": 1.8267305195331573,
      "epoch": 0.017297511777730073,
      "grad_norm": 13.24530029296875,
      "learning_rate": 2.7022183473796384e-06,
      "loss": 0.6354,
      "mean_token_accuracy": 0.8041493833065033,
      "num_tokens": 6712309.0,
      "step": 5580
    },
    {
      "entropy": 1.801226018369198,
      "epoch": 0.01732851090277977,
      "grad_norm": 13.650617599487305,
      "learning_rate": 2.707061900610288e-06,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8037795275449753,
      "num_tokens": 6725171.0,
      "step": 5590
    },
    {
      "entropy": 1.9054549396038056,
      "epoch": 0.017359510027829466,
      "grad_norm": 13.89938735961914,
      "learning_rate": 2.7119054538409377e-06,
      "loss": 0.673,
      "mean_token_accuracy": 0.794456647336483,
      "num_tokens": 6735768.0,
      "step": 5600
    },
    {
      "entropy": 1.8770520448684693,
      "epoch": 0.01739050915287916,
      "grad_norm": 11.668399810791016,
      "learning_rate": 2.716749007071588e-06,
      "loss": 0.6098,
      "mean_token_accuracy": 0.8073992922902107,
      "num_tokens": 6748114.0,
      "step": 5610
    },
    {
      "entropy": 1.8336502104997634,
      "epoch": 0.017421508277928856,
      "grad_norm": 15.977221488952637,
      "learning_rate": 2.721592560302238e-06,
      "loss": 0.5509,
      "mean_token_accuracy": 0.7992581978440285,
      "num_tokens": 6761005.0,
      "step": 5620
    },
    {
      "entropy": 1.868009014427662,
      "epoch": 0.017452507402978552,
      "grad_norm": 14.214484214782715,
      "learning_rate": 2.726436113532888e-06,
      "loss": 0.6505,
      "mean_token_accuracy": 0.794841094315052,
      "num_tokens": 6773183.0,
      "step": 5630
    },
    {
      "entropy": 1.8636821389198304,
      "epoch": 0.017483506528028245,
      "grad_norm": 12.620464324951172,
      "learning_rate": 2.7312796667635377e-06,
      "loss": 0.6509,
      "mean_token_accuracy": 0.8034439250826836,
      "num_tokens": 6785345.0,
      "step": 5640
    },
    {
      "entropy": 1.9341711491346358,
      "epoch": 0.017514505653077942,
      "grad_norm": 15.093106269836426,
      "learning_rate": 2.736123219994188e-06,
      "loss": 0.6546,
      "mean_token_accuracy": 0.80091482847929,
      "num_tokens": 6796959.0,
      "step": 5650
    },
    {
      "entropy": 1.8229330405592918,
      "epoch": 0.01754550477812764,
      "grad_norm": 5.629141807556152,
      "learning_rate": 2.740966773224838e-06,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8240866810083389,
      "num_tokens": 6809608.0,
      "step": 5660
    },
    {
      "entropy": 1.8584133207798004,
      "epoch": 0.017576503903177332,
      "grad_norm": 14.71494197845459,
      "learning_rate": 2.745810326455488e-06,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8071165978908539,
      "num_tokens": 6821161.0,
      "step": 5670
    },
    {
      "entropy": 1.9193618685007094,
      "epoch": 0.01760750302822703,
      "grad_norm": 6.501834392547607,
      "learning_rate": 2.750653879686138e-06,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8095588624477387,
      "num_tokens": 6832907.0,
      "step": 5680
    },
    {
      "entropy": 1.9340075165033341,
      "epoch": 0.017638502153276725,
      "grad_norm": 16.142831802368164,
      "learning_rate": 2.755497432916788e-06,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8106139227747917,
      "num_tokens": 6844618.0,
      "step": 5690
    },
    {
      "entropy": 1.9251418694853784,
      "epoch": 0.017669501278326418,
      "grad_norm": 5.188778877258301,
      "learning_rate": 2.760340986147438e-06,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8212108716368676,
      "num_tokens": 6856467.0,
      "step": 5700
    },
    {
      "entropy": 2.0215213894844055,
      "epoch": 0.017700500403376115,
      "grad_norm": 14.539006233215332,
      "learning_rate": 2.765184539378088e-06,
      "loss": 0.7512,
      "mean_token_accuracy": 0.7919975519180298,
      "num_tokens": 6867647.0,
      "step": 5710
    },
    {
      "entropy": 1.8021862357854843,
      "epoch": 0.01773149952842581,
      "grad_norm": 12.78034782409668,
      "learning_rate": 2.770028092608738e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8308590099215507,
      "num_tokens": 6879996.0,
      "step": 5720
    },
    {
      "entropy": 1.9824547916650772,
      "epoch": 0.017762498653475504,
      "grad_norm": 16.377723693847656,
      "learning_rate": 2.7748716458393883e-06,
      "loss": 0.6998,
      "mean_token_accuracy": 0.7981508508324623,
      "num_tokens": 6890686.0,
      "step": 5730
    },
    {
      "entropy": 1.8871808886528014,
      "epoch": 0.0177934977785252,
      "grad_norm": 8.088064193725586,
      "learning_rate": 2.779715199070038e-06,
      "loss": 0.6382,
      "mean_token_accuracy": 0.8009686887264251,
      "num_tokens": 6902728.0,
      "step": 5740
    },
    {
      "entropy": 1.9506133437156676,
      "epoch": 0.017824496903574898,
      "grad_norm": 13.706978797912598,
      "learning_rate": 2.7845587523006877e-06,
      "loss": 0.7109,
      "mean_token_accuracy": 0.7973517820239067,
      "num_tokens": 6913584.0,
      "step": 5750
    },
    {
      "entropy": 1.8800716385245324,
      "epoch": 0.01785549602862459,
      "grad_norm": 14.467167854309082,
      "learning_rate": 2.789402305531338e-06,
      "loss": 0.6481,
      "mean_token_accuracy": 0.8017850533127785,
      "num_tokens": 6925591.0,
      "step": 5760
    },
    {
      "entropy": 1.9344962298870088,
      "epoch": 0.017886495153674287,
      "grad_norm": 13.922574996948242,
      "learning_rate": 2.794245858761988e-06,
      "loss": 0.7058,
      "mean_token_accuracy": 0.8056110054254532,
      "num_tokens": 6936542.0,
      "step": 5770
    },
    {
      "entropy": 1.8384886384010315,
      "epoch": 0.017917494278723984,
      "grad_norm": 13.464262962341309,
      "learning_rate": 2.799089411992638e-06,
      "loss": 0.6779,
      "mean_token_accuracy": 0.8026763945817947,
      "num_tokens": 6948233.0,
      "step": 5780
    },
    {
      "entropy": 1.8690043538808823,
      "epoch": 0.017948493403773677,
      "grad_norm": 19.54326820373535,
      "learning_rate": 2.8039329652232885e-06,
      "loss": 0.6308,
      "mean_token_accuracy": 0.8030871346592903,
      "num_tokens": 6959330.0,
      "step": 5790
    },
    {
      "entropy": 1.8759321600198746,
      "epoch": 0.017979492528823374,
      "grad_norm": 12.744894981384277,
      "learning_rate": 2.808776518453938e-06,
      "loss": 0.6481,
      "mean_token_accuracy": 0.8097668170928956,
      "num_tokens": 6970815.0,
      "step": 5800
    },
    {
      "entropy": 1.828912016749382,
      "epoch": 0.01801049165387307,
      "grad_norm": 12.839016914367676,
      "learning_rate": 2.813620071684588e-06,
      "loss": 0.629,
      "mean_token_accuracy": 0.8034252509474754,
      "num_tokens": 6982527.0,
      "step": 5810
    },
    {
      "entropy": 1.8393566399812697,
      "epoch": 0.018041490778922763,
      "grad_norm": 12.746121406555176,
      "learning_rate": 2.818463624915238e-06,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8172145113348961,
      "num_tokens": 6993866.0,
      "step": 5820
    },
    {
      "entropy": 1.9092845901846887,
      "epoch": 0.01807248990397246,
      "grad_norm": 6.71993350982666,
      "learning_rate": 2.823307178145888e-06,
      "loss": 0.6816,
      "mean_token_accuracy": 0.7965900272130966,
      "num_tokens": 7005083.0,
      "step": 5830
    },
    {
      "entropy": 1.8039664879441262,
      "epoch": 0.018103489029022157,
      "grad_norm": 14.516589164733887,
      "learning_rate": 2.8281507313765383e-06,
      "loss": 0.6241,
      "mean_token_accuracy": 0.8058669179677963,
      "num_tokens": 7017669.0,
      "step": 5840
    },
    {
      "entropy": 1.8136644974350928,
      "epoch": 0.01813448815407185,
      "grad_norm": 5.473270893096924,
      "learning_rate": 2.8329942846071878e-06,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8144378513097763,
      "num_tokens": 7030554.0,
      "step": 5850
    },
    {
      "entropy": 1.8616077184677124,
      "epoch": 0.018165487279121546,
      "grad_norm": 6.733494281768799,
      "learning_rate": 2.837837837837838e-06,
      "loss": 0.6268,
      "mean_token_accuracy": 0.8107643172144889,
      "num_tokens": 7043459.0,
      "step": 5860
    },
    {
      "entropy": 1.8695634379982948,
      "epoch": 0.018196486404171243,
      "grad_norm": 13.21927547454834,
      "learning_rate": 2.842681391068488e-06,
      "loss": 0.66,
      "mean_token_accuracy": 0.7932876944541931,
      "num_tokens": 7055971.0,
      "step": 5870
    },
    {
      "entropy": 1.9111657977104186,
      "epoch": 0.018227485529220936,
      "grad_norm": 12.689513206481934,
      "learning_rate": 2.8475249442991383e-06,
      "loss": 0.6847,
      "mean_token_accuracy": 0.8026935487985611,
      "num_tokens": 7067129.0,
      "step": 5880
    },
    {
      "entropy": 1.9150221094489097,
      "epoch": 0.018258484654270633,
      "grad_norm": 15.985901832580566,
      "learning_rate": 2.852368497529788e-06,
      "loss": 0.636,
      "mean_token_accuracy": 0.8080303505063057,
      "num_tokens": 7078292.0,
      "step": 5890
    },
    {
      "entropy": 1.9083669915795327,
      "epoch": 0.01828948377932033,
      "grad_norm": 6.71138334274292,
      "learning_rate": 2.857212050760438e-06,
      "loss": 0.625,
      "mean_token_accuracy": 0.8103194802999496,
      "num_tokens": 7089362.0,
      "step": 5900
    },
    {
      "entropy": 1.9217876300215722,
      "epoch": 0.018320482904370026,
      "grad_norm": 15.180530548095703,
      "learning_rate": 2.862055603991088e-06,
      "loss": 0.7363,
      "mean_token_accuracy": 0.8016302824020386,
      "num_tokens": 7100640.0,
      "step": 5910
    },
    {
      "entropy": 1.8832744032144546,
      "epoch": 0.01835148202941972,
      "grad_norm": 7.438052654266357,
      "learning_rate": 2.8668991572217382e-06,
      "loss": 0.6582,
      "mean_token_accuracy": 0.7962846517562866,
      "num_tokens": 7112772.0,
      "step": 5920
    },
    {
      "entropy": 1.855891165137291,
      "epoch": 0.018382481154469416,
      "grad_norm": 14.392753601074219,
      "learning_rate": 2.871742710452388e-06,
      "loss": 0.6342,
      "mean_token_accuracy": 0.7954742982983589,
      "num_tokens": 7124974.0,
      "step": 5930
    },
    {
      "entropy": 1.906690326333046,
      "epoch": 0.018413480279519112,
      "grad_norm": 5.785147666931152,
      "learning_rate": 2.8765862636830384e-06,
      "loss": 0.6857,
      "mean_token_accuracy": 0.7974092051386833,
      "num_tokens": 7136071.0,
      "step": 5940
    },
    {
      "entropy": 1.8553122743964194,
      "epoch": 0.018444479404568805,
      "grad_norm": 12.362131118774414,
      "learning_rate": 2.8814298169136883e-06,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8133484557271004,
      "num_tokens": 7148231.0,
      "step": 5950
    },
    {
      "entropy": 1.8504623636603355,
      "epoch": 0.018475478529618502,
      "grad_norm": 12.675224304199219,
      "learning_rate": 2.8862733701443378e-06,
      "loss": 0.6486,
      "mean_token_accuracy": 0.8114531069993973,
      "num_tokens": 7159463.0,
      "step": 5960
    },
    {
      "entropy": 1.7733413144946097,
      "epoch": 0.0185064776546682,
      "grad_norm": 14.104755401611328,
      "learning_rate": 2.891116923374988e-06,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8127880603075027,
      "num_tokens": 7172067.0,
      "step": 5970
    },
    {
      "entropy": 1.9293433710932733,
      "epoch": 0.01853747677971789,
      "grad_norm": 13.121195793151855,
      "learning_rate": 2.895960476605638e-06,
      "loss": 0.6808,
      "mean_token_accuracy": 0.7942067563533783,
      "num_tokens": 7183215.0,
      "step": 5980
    },
    {
      "entropy": 1.9015644788742065,
      "epoch": 0.018568475904767588,
      "grad_norm": 14.695234298706055,
      "learning_rate": 2.9008040298362883e-06,
      "loss": 0.6561,
      "mean_token_accuracy": 0.807443767786026,
      "num_tokens": 7193804.0,
      "step": 5990
    },
    {
      "entropy": 1.8584507659077645,
      "epoch": 0.018599475029817285,
      "grad_norm": 16.059717178344727,
      "learning_rate": 2.9056475830669386e-06,
      "loss": 0.6235,
      "mean_token_accuracy": 0.808501772582531,
      "num_tokens": 7205778.0,
      "step": 6000
    },
    {
      "entropy": 1.8808137744665145,
      "epoch": 0.018630474154866978,
      "grad_norm": 13.323634147644043,
      "learning_rate": 2.910491136297588e-06,
      "loss": 0.6352,
      "mean_token_accuracy": 0.8102394938468933,
      "num_tokens": 7217406.0,
      "step": 6010
    },
    {
      "entropy": 1.9097854852676392,
      "epoch": 0.018661473279916675,
      "grad_norm": 12.783294677734375,
      "learning_rate": 2.915334689528238e-06,
      "loss": 0.6669,
      "mean_token_accuracy": 0.8101324290037155,
      "num_tokens": 7228513.0,
      "step": 6020
    },
    {
      "entropy": 1.949710837006569,
      "epoch": 0.01869247240496637,
      "grad_norm": 8.33492660522461,
      "learning_rate": 2.9201782427588882e-06,
      "loss": 0.7241,
      "mean_token_accuracy": 0.7899691253900528,
      "num_tokens": 7240091.0,
      "step": 6030
    },
    {
      "entropy": 1.7995387852191924,
      "epoch": 0.018723471530016064,
      "grad_norm": 14.658129692077637,
      "learning_rate": 2.925021795989538e-06,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8123287737369538,
      "num_tokens": 7252717.0,
      "step": 6040
    },
    {
      "entropy": 1.8480943590402603,
      "epoch": 0.01875447065506576,
      "grad_norm": 12.393167495727539,
      "learning_rate": 2.9298653492201884e-06,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8123827025294303,
      "num_tokens": 7265132.0,
      "step": 6050
    },
    {
      "entropy": 1.9156224384903908,
      "epoch": 0.018785469780115457,
      "grad_norm": 17.64678382873535,
      "learning_rate": 2.934708902450838e-06,
      "loss": 0.637,
      "mean_token_accuracy": 0.7986405953764916,
      "num_tokens": 7277392.0,
      "step": 6060
    },
    {
      "entropy": 1.8689854100346566,
      "epoch": 0.01881646890516515,
      "grad_norm": 13.045455932617188,
      "learning_rate": 2.939552455681488e-06,
      "loss": 0.605,
      "mean_token_accuracy": 0.8187311470508576,
      "num_tokens": 7289307.0,
      "step": 6070
    },
    {
      "entropy": 1.9299559414386749,
      "epoch": 0.018847468030214847,
      "grad_norm": 14.602327346801758,
      "learning_rate": 2.944396008912138e-06,
      "loss": 0.6499,
      "mean_token_accuracy": 0.8123447954654693,
      "num_tokens": 7300976.0,
      "step": 6080
    },
    {
      "entropy": 1.8863015726208687,
      "epoch": 0.018878467155264544,
      "grad_norm": 16.277345657348633,
      "learning_rate": 2.9492395621427884e-06,
      "loss": 0.6285,
      "mean_token_accuracy": 0.805467925965786,
      "num_tokens": 7312816.0,
      "step": 6090
    },
    {
      "entropy": 1.88625720590353,
      "epoch": 0.018909466280314237,
      "grad_norm": 13.414533615112305,
      "learning_rate": 2.9540831153734383e-06,
      "loss": 0.61,
      "mean_token_accuracy": 0.8089847132563591,
      "num_tokens": 7325107.0,
      "step": 6100
    },
    {
      "entropy": 1.9170991733670235,
      "epoch": 0.018940465405363933,
      "grad_norm": 12.490900039672852,
      "learning_rate": 2.958926668604088e-06,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8148081079125404,
      "num_tokens": 7337186.0,
      "step": 6110
    },
    {
      "entropy": 1.8110515862703322,
      "epoch": 0.01897146453041363,
      "grad_norm": 6.398469924926758,
      "learning_rate": 2.963770221834738e-06,
      "loss": 0.5864,
      "mean_token_accuracy": 0.8025536656379699,
      "num_tokens": 7351631.0,
      "step": 6120
    },
    {
      "entropy": 1.925946244597435,
      "epoch": 0.019002463655463323,
      "grad_norm": 12.947382926940918,
      "learning_rate": 2.9686137750653883e-06,
      "loss": 0.6428,
      "mean_token_accuracy": 0.8030333101749421,
      "num_tokens": 7363334.0,
      "step": 6130
    },
    {
      "entropy": 1.84049913585186,
      "epoch": 0.01903346278051302,
      "grad_norm": 13.585901260375977,
      "learning_rate": 2.9734573282960382e-06,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8090210378170013,
      "num_tokens": 7376221.0,
      "step": 6140
    },
    {
      "entropy": 1.913094098865986,
      "epoch": 0.019064461905562716,
      "grad_norm": 13.463624954223633,
      "learning_rate": 2.9783008815266885e-06,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8032952442765235,
      "num_tokens": 7388430.0,
      "step": 6150
    },
    {
      "entropy": 1.9327594608068466,
      "epoch": 0.01909546103061241,
      "grad_norm": 13.593147277832031,
      "learning_rate": 2.9831444347573384e-06,
      "loss": 0.6631,
      "mean_token_accuracy": 0.8024881407618523,
      "num_tokens": 7400021.0,
      "step": 6160
    },
    {
      "entropy": 1.920905977487564,
      "epoch": 0.019126460155662106,
      "grad_norm": 12.407073020935059,
      "learning_rate": 2.987987987987988e-06,
      "loss": 0.6858,
      "mean_token_accuracy": 0.8053328812122345,
      "num_tokens": 7411092.0,
      "step": 6170
    },
    {
      "entropy": 1.9286428660154342,
      "epoch": 0.019157459280711803,
      "grad_norm": 11.15323543548584,
      "learning_rate": 2.992831541218638e-06,
      "loss": 0.6284,
      "mean_token_accuracy": 0.8029859021306038,
      "num_tokens": 7423534.0,
      "step": 6180
    },
    {
      "entropy": 1.91870975792408,
      "epoch": 0.019188458405761496,
      "grad_norm": 6.104159832000732,
      "learning_rate": 2.9976750944492885e-06,
      "loss": 0.6215,
      "mean_token_accuracy": 0.804423876106739,
      "num_tokens": 7434989.0,
      "step": 6190
    },
    {
      "entropy": 1.9800142824649811,
      "epoch": 0.019219457530811192,
      "grad_norm": 15.913412094116211,
      "learning_rate": 3.0025186476799384e-06,
      "loss": 0.6803,
      "mean_token_accuracy": 0.8035276308655739,
      "num_tokens": 7446518.0,
      "step": 6200
    },
    {
      "entropy": 1.925678089261055,
      "epoch": 0.01925045665586089,
      "grad_norm": 16.255130767822266,
      "learning_rate": 3.0073622009105887e-06,
      "loss": 0.6876,
      "mean_token_accuracy": 0.8033014148473739,
      "num_tokens": 7457926.0,
      "step": 6210
    },
    {
      "entropy": 1.951409675180912,
      "epoch": 0.019281455780910582,
      "grad_norm": 13.905477523803711,
      "learning_rate": 3.012205754141238e-06,
      "loss": 0.6945,
      "mean_token_accuracy": 0.793704767525196,
      "num_tokens": 7469282.0,
      "step": 6220
    },
    {
      "entropy": 1.8324480682611466,
      "epoch": 0.01931245490596028,
      "grad_norm": 6.767990589141846,
      "learning_rate": 3.017049307371888e-06,
      "loss": 0.5416,
      "mean_token_accuracy": 0.818550530076027,
      "num_tokens": 7481902.0,
      "step": 6230
    },
    {
      "entropy": 1.9617059826850891,
      "epoch": 0.019343454031009975,
      "grad_norm": 6.368531227111816,
      "learning_rate": 3.0218928606025383e-06,
      "loss": 0.6478,
      "mean_token_accuracy": 0.8018120333552361,
      "num_tokens": 7493336.0,
      "step": 6240
    },
    {
      "entropy": 1.8573770090937614,
      "epoch": 0.019374453156059672,
      "grad_norm": 6.3957295417785645,
      "learning_rate": 3.0267364138331882e-06,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8206924736499787,
      "num_tokens": 7506208.0,
      "step": 6250
    },
    {
      "entropy": 1.8330803513526917,
      "epoch": 0.019405452281109365,
      "grad_norm": 6.754683494567871,
      "learning_rate": 3.0315799670638385e-06,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8097323834896087,
      "num_tokens": 7519948.0,
      "step": 6260
    },
    {
      "entropy": 1.911997850239277,
      "epoch": 0.01943645140615906,
      "grad_norm": 14.22529125213623,
      "learning_rate": 3.036423520294488e-06,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8061494305729866,
      "num_tokens": 7531446.0,
      "step": 6270
    },
    {
      "entropy": 1.7933774992823601,
      "epoch": 0.019467450531208758,
      "grad_norm": 12.579453468322754,
      "learning_rate": 3.0412670735251383e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8117320775985718,
      "num_tokens": 7545248.0,
      "step": 6280
    },
    {
      "entropy": 2.001555660367012,
      "epoch": 0.01949844965625845,
      "grad_norm": 11.950443267822266,
      "learning_rate": 3.046110626755788e-06,
      "loss": 0.6827,
      "mean_token_accuracy": 0.8007282823324203,
      "num_tokens": 7555914.0,
      "step": 6290
    },
    {
      "entropy": 1.832805335521698,
      "epoch": 0.019529448781308148,
      "grad_norm": 16.177709579467773,
      "learning_rate": 3.0509541799864385e-06,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8259735867381096,
      "num_tokens": 7568442.0,
      "step": 6300
    },
    {
      "entropy": 1.9412631839513779,
      "epoch": 0.019560447906357845,
      "grad_norm": 13.161090850830078,
      "learning_rate": 3.0557977332170884e-06,
      "loss": 0.7107,
      "mean_token_accuracy": 0.7944168001413345,
      "num_tokens": 7579082.0,
      "step": 6310
    },
    {
      "entropy": 1.8757167264819146,
      "epoch": 0.019591447031407538,
      "grad_norm": 11.784358978271484,
      "learning_rate": 3.0606412864477382e-06,
      "loss": 0.5951,
      "mean_token_accuracy": 0.8066261202096939,
      "num_tokens": 7590950.0,
      "step": 6320
    },
    {
      "entropy": 1.8339521378278731,
      "epoch": 0.019622446156457234,
      "grad_norm": 14.665081977844238,
      "learning_rate": 3.065484839678388e-06,
      "loss": 0.5788,
      "mean_token_accuracy": 0.8134505435824394,
      "num_tokens": 7602873.0,
      "step": 6330
    },
    {
      "entropy": 1.853539614379406,
      "epoch": 0.01965344528150693,
      "grad_norm": 14.00977611541748,
      "learning_rate": 3.0703283929090384e-06,
      "loss": 0.6516,
      "mean_token_accuracy": 0.8087933987379075,
      "num_tokens": 7614636.0,
      "step": 6340
    },
    {
      "entropy": 1.8311856165528297,
      "epoch": 0.019684444406556624,
      "grad_norm": 13.730056762695312,
      "learning_rate": 3.0751719461396883e-06,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8031821623444557,
      "num_tokens": 7627435.0,
      "step": 6350
    },
    {
      "entropy": 1.9072323009371757,
      "epoch": 0.01971544353160632,
      "grad_norm": 5.6056036949157715,
      "learning_rate": 3.0800154993703386e-06,
      "loss": 0.6932,
      "mean_token_accuracy": 0.7961813092231751,
      "num_tokens": 7639023.0,
      "step": 6360
    },
    {
      "entropy": 1.8226707085967064,
      "epoch": 0.019746442656656017,
      "grad_norm": 15.677154541015625,
      "learning_rate": 3.0848590526009885e-06,
      "loss": 0.6193,
      "mean_token_accuracy": 0.8108833208680153,
      "num_tokens": 7651962.0,
      "step": 6370
    },
    {
      "entropy": 1.7706042945384979,
      "epoch": 0.01977744178170571,
      "grad_norm": 13.05368423461914,
      "learning_rate": 3.089702605831638e-06,
      "loss": 0.5883,
      "mean_token_accuracy": 0.8179239988327026,
      "num_tokens": 7664383.0,
      "step": 6380
    },
    {
      "entropy": 1.888214285671711,
      "epoch": 0.019808440906755407,
      "grad_norm": 12.008337020874023,
      "learning_rate": 3.0945461590622883e-06,
      "loss": 0.6435,
      "mean_token_accuracy": 0.8116688475012779,
      "num_tokens": 7675538.0,
      "step": 6390
    },
    {
      "entropy": 1.8811195820569993,
      "epoch": 0.019839440031805104,
      "grad_norm": 13.883013725280762,
      "learning_rate": 3.0993897122929386e-06,
      "loss": 0.6319,
      "mean_token_accuracy": 0.8048406511545181,
      "num_tokens": 7687099.0,
      "step": 6400
    },
    {
      "entropy": 1.8231300503015517,
      "epoch": 0.019870439156854797,
      "grad_norm": 11.85662841796875,
      "learning_rate": 3.1042332655235885e-06,
      "loss": 0.5582,
      "mean_token_accuracy": 0.816590279340744,
      "num_tokens": 7700987.0,
      "step": 6410
    },
    {
      "entropy": 1.8579674810171127,
      "epoch": 0.019901438281904493,
      "grad_norm": 15.262175559997559,
      "learning_rate": 3.1090768187542388e-06,
      "loss": 0.6259,
      "mean_token_accuracy": 0.8095797553658486,
      "num_tokens": 7713844.0,
      "step": 6420
    },
    {
      "entropy": 1.879908984899521,
      "epoch": 0.01993243740695419,
      "grad_norm": 13.756391525268555,
      "learning_rate": 3.1139203719848882e-06,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8058964297175407,
      "num_tokens": 7726962.0,
      "step": 6430
    },
    {
      "entropy": 1.9172916844487191,
      "epoch": 0.019963436532003883,
      "grad_norm": 6.88080358505249,
      "learning_rate": 3.118763925215538e-06,
      "loss": 0.6446,
      "mean_token_accuracy": 0.7972790181636811,
      "num_tokens": 7738773.0,
      "step": 6440
    },
    {
      "entropy": 1.9533959448337554,
      "epoch": 0.01999443565705358,
      "grad_norm": 14.810494422912598,
      "learning_rate": 3.1236074784461884e-06,
      "loss": 0.7092,
      "mean_token_accuracy": 0.797048932313919,
      "num_tokens": 7749295.0,
      "step": 6450
    },
    {
      "entropy": 1.904009547829628,
      "epoch": 0.020025434782103276,
      "grad_norm": 12.683305740356445,
      "learning_rate": 3.1284510316768383e-06,
      "loss": 0.6457,
      "mean_token_accuracy": 0.8019493162631989,
      "num_tokens": 7760568.0,
      "step": 6460
    },
    {
      "entropy": 1.8452860102057458,
      "epoch": 0.02005643390715297,
      "grad_norm": 6.6882171630859375,
      "learning_rate": 3.1332945849074886e-06,
      "loss": 0.591,
      "mean_token_accuracy": 0.8225623354315758,
      "num_tokens": 7773203.0,
      "step": 6470
    },
    {
      "entropy": 1.891256783902645,
      "epoch": 0.020087433032202666,
      "grad_norm": 15.097151756286621,
      "learning_rate": 3.138138138138138e-06,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8069896474480629,
      "num_tokens": 7785046.0,
      "step": 6480
    },
    {
      "entropy": 1.841528557240963,
      "epoch": 0.020118432157252363,
      "grad_norm": 4.251867771148682,
      "learning_rate": 3.1429816913687884e-06,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8183139115571976,
      "num_tokens": 7798180.0,
      "step": 6490
    },
    {
      "entropy": 1.8745207443833352,
      "epoch": 0.020149431282302056,
      "grad_norm": 13.895639419555664,
      "learning_rate": 3.1478252445994383e-06,
      "loss": 0.6282,
      "mean_token_accuracy": 0.8085477828979493,
      "num_tokens": 7810555.0,
      "step": 6500
    },
    {
      "entropy": 1.8891053274273872,
      "epoch": 0.020180430407351752,
      "grad_norm": 14.992795944213867,
      "learning_rate": 3.1526687978300886e-06,
      "loss": 0.6291,
      "mean_token_accuracy": 0.80573670566082,
      "num_tokens": 7821921.0,
      "step": 6510
    },
    {
      "entropy": 1.9124128386378287,
      "epoch": 0.02021142953240145,
      "grad_norm": 12.522514343261719,
      "learning_rate": 3.1575123510607385e-06,
      "loss": 0.6336,
      "mean_token_accuracy": 0.820842219889164,
      "num_tokens": 7832746.0,
      "step": 6520
    },
    {
      "entropy": 1.7800809681415557,
      "epoch": 0.020242428657451142,
      "grad_norm": 11.537628173828125,
      "learning_rate": 3.1623559042913884e-06,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8302456393837929,
      "num_tokens": 7845908.0,
      "step": 6530
    },
    {
      "entropy": 1.8512769252061845,
      "epoch": 0.02027342778250084,
      "grad_norm": 13.687551498413086,
      "learning_rate": 3.1671994575220382e-06,
      "loss": 0.594,
      "mean_token_accuracy": 0.8206526070833207,
      "num_tokens": 7857703.0,
      "step": 6540
    },
    {
      "entropy": 1.9494429141283036,
      "epoch": 0.020304426907550535,
      "grad_norm": 13.054556846618652,
      "learning_rate": 3.1720430107526885e-06,
      "loss": 0.6967,
      "mean_token_accuracy": 0.7968014314770698,
      "num_tokens": 7868457.0,
      "step": 6550
    },
    {
      "entropy": 1.872872059047222,
      "epoch": 0.02033542603260023,
      "grad_norm": 14.116209030151367,
      "learning_rate": 3.1768865639833384e-06,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8167121425271034,
      "num_tokens": 7880101.0,
      "step": 6560
    },
    {
      "entropy": 1.8299372598528862,
      "epoch": 0.020366425157649925,
      "grad_norm": 15.94222354888916,
      "learning_rate": 3.1817301172139887e-06,
      "loss": 0.6118,
      "mean_token_accuracy": 0.8142679139971734,
      "num_tokens": 7892509.0,
      "step": 6570
    },
    {
      "entropy": 1.8935213565826416,
      "epoch": 0.02039742428269962,
      "grad_norm": 15.16163158416748,
      "learning_rate": 3.1865736704446386e-06,
      "loss": 0.6278,
      "mean_token_accuracy": 0.8115665182471276,
      "num_tokens": 7903535.0,
      "step": 6580
    },
    {
      "entropy": 1.737954817712307,
      "epoch": 0.020428423407749315,
      "grad_norm": 13.716304779052734,
      "learning_rate": 3.191417223675288e-06,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8255019202828408,
      "num_tokens": 7917342.0,
      "step": 6590
    },
    {
      "entropy": 1.8174697816371919,
      "epoch": 0.02045942253279901,
      "grad_norm": 13.559125900268555,
      "learning_rate": 3.1962607769059384e-06,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8168426662683487,
      "num_tokens": 7929488.0,
      "step": 6600
    },
    {
      "entropy": 1.8946778982877732,
      "epoch": 0.020490421657848708,
      "grad_norm": 15.61142635345459,
      "learning_rate": 3.2011043301365887e-06,
      "loss": 0.7263,
      "mean_token_accuracy": 0.799173790216446,
      "num_tokens": 7940278.0,
      "step": 6610
    },
    {
      "entropy": 1.8623188078403472,
      "epoch": 0.020521420782898404,
      "grad_norm": 11.181509017944336,
      "learning_rate": 3.2059478833672386e-06,
      "loss": 0.6354,
      "mean_token_accuracy": 0.8031112432479859,
      "num_tokens": 7951379.0,
      "step": 6620
    },
    {
      "entropy": 1.814198338985443,
      "epoch": 0.020552419907948098,
      "grad_norm": 13.728723526000977,
      "learning_rate": 3.210791436597889e-06,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8157623007893562,
      "num_tokens": 7963714.0,
      "step": 6630
    },
    {
      "entropy": 1.7929986909031868,
      "epoch": 0.020583419032997794,
      "grad_norm": 15.344460487365723,
      "learning_rate": 3.2156349898285383e-06,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8098177522420883,
      "num_tokens": 7976870.0,
      "step": 6640
    },
    {
      "entropy": 1.947634482383728,
      "epoch": 0.02061441815804749,
      "grad_norm": 14.946516036987305,
      "learning_rate": 3.2204785430591882e-06,
      "loss": 0.7049,
      "mean_token_accuracy": 0.79284388422966,
      "num_tokens": 7987426.0,
      "step": 6650
    },
    {
      "entropy": 1.8572285056114197,
      "epoch": 0.020645417283097184,
      "grad_norm": 6.301254749298096,
      "learning_rate": 3.2253220962898385e-06,
      "loss": 0.6301,
      "mean_token_accuracy": 0.8085553154349328,
      "num_tokens": 8000549.0,
      "step": 6660
    },
    {
      "entropy": 1.803759203851223,
      "epoch": 0.02067641640814688,
      "grad_norm": 12.17286491394043,
      "learning_rate": 3.2301656495204884e-06,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8267651617527008,
      "num_tokens": 8015000.0,
      "step": 6670
    },
    {
      "entropy": 1.8756111353635787,
      "epoch": 0.020707415533196577,
      "grad_norm": 15.4856595993042,
      "learning_rate": 3.2350092027511387e-06,
      "loss": 0.5969,
      "mean_token_accuracy": 0.80719293653965,
      "num_tokens": 8026773.0,
      "step": 6680
    },
    {
      "entropy": 1.9130279645323753,
      "epoch": 0.02073841465824627,
      "grad_norm": 14.601099014282227,
      "learning_rate": 3.239852755981788e-06,
      "loss": 0.6407,
      "mean_token_accuracy": 0.8074927061796189,
      "num_tokens": 8038375.0,
      "step": 6690
    },
    {
      "entropy": 1.84910279661417,
      "epoch": 0.020769413783295967,
      "grad_norm": 13.639535903930664,
      "learning_rate": 3.2446963092124385e-06,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8115042328834534,
      "num_tokens": 8050746.0,
      "step": 6700
    },
    {
      "entropy": 1.9024951666593553,
      "epoch": 0.020800412908345663,
      "grad_norm": 11.334275245666504,
      "learning_rate": 3.2495398624430884e-06,
      "loss": 0.6458,
      "mean_token_accuracy": 0.8086831450462342,
      "num_tokens": 8062275.0,
      "step": 6710
    },
    {
      "entropy": 1.953022199869156,
      "epoch": 0.020831412033395356,
      "grad_norm": 12.958690643310547,
      "learning_rate": 3.2543834156737387e-06,
      "loss": 0.6934,
      "mean_token_accuracy": 0.8028324991464615,
      "num_tokens": 8073879.0,
      "step": 6720
    },
    {
      "entropy": 1.9028348535299302,
      "epoch": 0.020862411158445053,
      "grad_norm": 12.473337173461914,
      "learning_rate": 3.2592269689043886e-06,
      "loss": 0.6281,
      "mean_token_accuracy": 0.808925811946392,
      "num_tokens": 8086058.0,
      "step": 6730
    },
    {
      "entropy": 1.8883586995303632,
      "epoch": 0.02089341028349475,
      "grad_norm": 13.6912841796875,
      "learning_rate": 3.2640705221350385e-06,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8137388199567794,
      "num_tokens": 8097481.0,
      "step": 6740
    },
    {
      "entropy": 1.8357958167791366,
      "epoch": 0.020924409408544443,
      "grad_norm": 12.956893920898438,
      "learning_rate": 3.2689140753656883e-06,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8275958806276321,
      "num_tokens": 8109412.0,
      "step": 6750
    },
    {
      "entropy": 1.7908343866467475,
      "epoch": 0.02095540853359414,
      "grad_norm": 14.640457153320312,
      "learning_rate": 3.2737576285963386e-06,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8118628889322281,
      "num_tokens": 8122837.0,
      "step": 6760
    },
    {
      "entropy": 1.7841477155685426,
      "epoch": 0.020986407658643836,
      "grad_norm": 12.808713912963867,
      "learning_rate": 3.2786011818269885e-06,
      "loss": 0.6055,
      "mean_token_accuracy": 0.8140031367540359,
      "num_tokens": 8136163.0,
      "step": 6770
    },
    {
      "entropy": 1.8466138496994973,
      "epoch": 0.02101740678369353,
      "grad_norm": 14.429435729980469,
      "learning_rate": 3.283444735057639e-06,
      "loss": 0.6075,
      "mean_token_accuracy": 0.8045207351446152,
      "num_tokens": 8148613.0,
      "step": 6780
    },
    {
      "entropy": 1.7560095891356469,
      "epoch": 0.021048405908743226,
      "grad_norm": 7.405726909637451,
      "learning_rate": 3.2882882882882887e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8221228688955307,
      "num_tokens": 8162089.0,
      "step": 6790
    },
    {
      "entropy": 1.9308898389339446,
      "epoch": 0.021079405033792922,
      "grad_norm": 14.129651069641113,
      "learning_rate": 3.293131841518938e-06,
      "loss": 0.6899,
      "mean_token_accuracy": 0.8042490169405937,
      "num_tokens": 8173779.0,
      "step": 6800
    },
    {
      "entropy": 1.9463881149888038,
      "epoch": 0.021110404158842615,
      "grad_norm": 8.938533782958984,
      "learning_rate": 3.2979753947495885e-06,
      "loss": 0.6966,
      "mean_token_accuracy": 0.8073234111070633,
      "num_tokens": 8184942.0,
      "step": 6810
    },
    {
      "entropy": 1.8929473131895065,
      "epoch": 0.021141403283892312,
      "grad_norm": 17.423473358154297,
      "learning_rate": 3.302818947980239e-06,
      "loss": 0.7093,
      "mean_token_accuracy": 0.7885128021240234,
      "num_tokens": 8196565.0,
      "step": 6820
    },
    {
      "entropy": 1.8256116762757302,
      "epoch": 0.02117240240894201,
      "grad_norm": 11.359122276306152,
      "learning_rate": 3.3076625012108887e-06,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8059591919183731,
      "num_tokens": 8208383.0,
      "step": 6830
    },
    {
      "entropy": 1.8639884352684022,
      "epoch": 0.021203401533991702,
      "grad_norm": 14.2997407913208,
      "learning_rate": 3.312506054441539e-06,
      "loss": 0.6282,
      "mean_token_accuracy": 0.8068465679883957,
      "num_tokens": 8220104.0,
      "step": 6840
    },
    {
      "entropy": 1.9167477533221244,
      "epoch": 0.0212344006590414,
      "grad_norm": 6.671177387237549,
      "learning_rate": 3.3173496076721885e-06,
      "loss": 0.6529,
      "mean_token_accuracy": 0.8033589154481888,
      "num_tokens": 8231370.0,
      "step": 6850
    },
    {
      "entropy": 1.8746015638113023,
      "epoch": 0.021265399784091095,
      "grad_norm": 14.27453327178955,
      "learning_rate": 3.3221931609028383e-06,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8159545630216598,
      "num_tokens": 8243023.0,
      "step": 6860
    },
    {
      "entropy": 1.917281800508499,
      "epoch": 0.021296398909140788,
      "grad_norm": 11.446333885192871,
      "learning_rate": 3.3270367141334886e-06,
      "loss": 0.6558,
      "mean_token_accuracy": 0.7994436591863632,
      "num_tokens": 8254401.0,
      "step": 6870
    },
    {
      "entropy": 1.8957316786050797,
      "epoch": 0.021327398034190485,
      "grad_norm": 11.739078521728516,
      "learning_rate": 3.3318802673641385e-06,
      "loss": 0.7036,
      "mean_token_accuracy": 0.8000991299748421,
      "num_tokens": 8265419.0,
      "step": 6880
    },
    {
      "entropy": 1.7615302629768848,
      "epoch": 0.02135839715924018,
      "grad_norm": 5.259082317352295,
      "learning_rate": 3.336723820594789e-06,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8190023839473725,
      "num_tokens": 8278764.0,
      "step": 6890
    },
    {
      "entropy": 1.7910688430070878,
      "epoch": 0.021389396284289874,
      "grad_norm": 12.807435035705566,
      "learning_rate": 3.3415673738254383e-06,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8227502018213272,
      "num_tokens": 8291327.0,
      "step": 6900
    },
    {
      "entropy": 1.7641884714365006,
      "epoch": 0.02142039540933957,
      "grad_norm": 12.872736930847168,
      "learning_rate": 3.3464109270560886e-06,
      "loss": 0.5901,
      "mean_token_accuracy": 0.8182562112808227,
      "num_tokens": 8303703.0,
      "step": 6910
    },
    {
      "entropy": 1.8527298361063003,
      "epoch": 0.021451394534389268,
      "grad_norm": 14.935564041137695,
      "learning_rate": 3.3512544802867385e-06,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8045241579413414,
      "num_tokens": 8315526.0,
      "step": 6920
    },
    {
      "entropy": 1.8295287489891052,
      "epoch": 0.02148239365943896,
      "grad_norm": 13.789701461791992,
      "learning_rate": 3.356098033517389e-06,
      "loss": 0.6008,
      "mean_token_accuracy": 0.8134370803833008,
      "num_tokens": 8327672.0,
      "step": 6930
    },
    {
      "entropy": 1.896050798892975,
      "epoch": 0.021513392784488657,
      "grad_norm": 12.54214859008789,
      "learning_rate": 3.3609415867480387e-06,
      "loss": 0.6978,
      "mean_token_accuracy": 0.7970891669392586,
      "num_tokens": 8339084.0,
      "step": 6940
    },
    {
      "entropy": 1.8715087831020356,
      "epoch": 0.021544391909538354,
      "grad_norm": 13.210384368896484,
      "learning_rate": 3.3657851399786886e-06,
      "loss": 0.6879,
      "mean_token_accuracy": 0.7981110483407974,
      "num_tokens": 8350484.0,
      "step": 6950
    },
    {
      "entropy": 1.747877648472786,
      "epoch": 0.021575391034588047,
      "grad_norm": 13.573982238769531,
      "learning_rate": 3.3706286932093384e-06,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8265791207551956,
      "num_tokens": 8363033.0,
      "step": 6960
    },
    {
      "entropy": 1.7737156122922897,
      "epoch": 0.021606390159637744,
      "grad_norm": 16.360973358154297,
      "learning_rate": 3.3754722464399888e-06,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8136523142457008,
      "num_tokens": 8375536.0,
      "step": 6970
    },
    {
      "entropy": 1.7883441895246506,
      "epoch": 0.02163738928468744,
      "grad_norm": 13.998126029968262,
      "learning_rate": 3.3803157996706386e-06,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8114990651607513,
      "num_tokens": 8388141.0,
      "step": 6980
    },
    {
      "entropy": 1.9003767311573028,
      "epoch": 0.021668388409737137,
      "grad_norm": 17.087512969970703,
      "learning_rate": 3.385159352901289e-06,
      "loss": 0.677,
      "mean_token_accuracy": 0.8007842287421226,
      "num_tokens": 8399614.0,
      "step": 6990
    },
    {
      "entropy": 1.8496545001864433,
      "epoch": 0.02169938753478683,
      "grad_norm": 11.947819709777832,
      "learning_rate": 3.390002906131939e-06,
      "loss": 0.5943,
      "mean_token_accuracy": 0.8178739234805107,
      "num_tokens": 8410656.0,
      "step": 7000
    },
    {
      "entropy": 1.8931600719690322,
      "epoch": 0.021730386659836527,
      "grad_norm": 12.223316192626953,
      "learning_rate": 3.3948464593625883e-06,
      "loss": 0.6515,
      "mean_token_accuracy": 0.804112882912159,
      "num_tokens": 8422509.0,
      "step": 7010
    },
    {
      "entropy": 1.8370696052908897,
      "epoch": 0.021761385784886223,
      "grad_norm": 13.245317459106445,
      "learning_rate": 3.3996900125932386e-06,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8125099033117295,
      "num_tokens": 8434469.0,
      "step": 7020
    },
    {
      "entropy": 1.9446087509393692,
      "epoch": 0.021792384909935916,
      "grad_norm": 13.750452995300293,
      "learning_rate": 3.404533565823889e-06,
      "loss": 0.6823,
      "mean_token_accuracy": 0.793831068277359,
      "num_tokens": 8445989.0,
      "step": 7030
    },
    {
      "entropy": 1.8339137956500053,
      "epoch": 0.021823384034985613,
      "grad_norm": 12.146233558654785,
      "learning_rate": 3.409377119054539e-06,
      "loss": 0.585,
      "mean_token_accuracy": 0.8186416819691658,
      "num_tokens": 8459049.0,
      "step": 7040
    },
    {
      "entropy": 1.9154435515403747,
      "epoch": 0.02185438316003531,
      "grad_norm": 13.75799560546875,
      "learning_rate": 3.414220672285189e-06,
      "loss": 0.6339,
      "mean_token_accuracy": 0.8105358377099037,
      "num_tokens": 8470694.0,
      "step": 7050
    },
    {
      "entropy": 1.8546890705823897,
      "epoch": 0.021885382285085003,
      "grad_norm": 14.610628128051758,
      "learning_rate": 3.4190642255158386e-06,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8084279328584671,
      "num_tokens": 8483926.0,
      "step": 7060
    },
    {
      "entropy": 1.9373271316289902,
      "epoch": 0.0219163814101347,
      "grad_norm": 13.321712493896484,
      "learning_rate": 3.4239077787464884e-06,
      "loss": 0.635,
      "mean_token_accuracy": 0.7939963206648827,
      "num_tokens": 8495258.0,
      "step": 7070
    },
    {
      "entropy": 1.8608694806694985,
      "epoch": 0.021947380535184396,
      "grad_norm": 12.879969596862793,
      "learning_rate": 3.4287513319771387e-06,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8156390026211738,
      "num_tokens": 8507779.0,
      "step": 7080
    },
    {
      "entropy": 1.9322496995329856,
      "epoch": 0.02197837966023409,
      "grad_norm": 15.521547317504883,
      "learning_rate": 3.4335948852077886e-06,
      "loss": 0.65,
      "mean_token_accuracy": 0.8115120708942414,
      "num_tokens": 8519474.0,
      "step": 7090
    },
    {
      "entropy": 1.7319314986467362,
      "epoch": 0.022009378785283785,
      "grad_norm": 16.432004928588867,
      "learning_rate": 3.438438438438439e-06,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8291773125529289,
      "num_tokens": 8533400.0,
      "step": 7100
    },
    {
      "entropy": 1.8349164828658104,
      "epoch": 0.022040377910333482,
      "grad_norm": 4.809366703033447,
      "learning_rate": 3.4432819916690884e-06,
      "loss": 0.598,
      "mean_token_accuracy": 0.809609466791153,
      "num_tokens": 8547001.0,
      "step": 7110
    },
    {
      "entropy": 1.8549170672893525,
      "epoch": 0.022071377035383175,
      "grad_norm": 12.525724411010742,
      "learning_rate": 3.4481255448997387e-06,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8268691346049308,
      "num_tokens": 8559045.0,
      "step": 7120
    },
    {
      "entropy": 1.8226833492517471,
      "epoch": 0.022102376160432872,
      "grad_norm": 11.418012619018555,
      "learning_rate": 3.4529690981303886e-06,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8124491006135941,
      "num_tokens": 8572509.0,
      "step": 7130
    },
    {
      "entropy": 1.7938767224550247,
      "epoch": 0.02213337528548257,
      "grad_norm": 14.121147155761719,
      "learning_rate": 3.457812651361039e-06,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8254873216152191,
      "num_tokens": 8585878.0,
      "step": 7140
    },
    {
      "entropy": 1.7850069940090179,
      "epoch": 0.02216437441053226,
      "grad_norm": 3.6775126457214355,
      "learning_rate": 3.4626562045916888e-06,
      "loss": 0.563,
      "mean_token_accuracy": 0.8238705024123192,
      "num_tokens": 8599106.0,
      "step": 7150
    },
    {
      "entropy": 1.8556845039129257,
      "epoch": 0.022195373535581958,
      "grad_norm": 11.72307300567627,
      "learning_rate": 3.4674997578223387e-06,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8144358694553375,
      "num_tokens": 8611671.0,
      "step": 7160
    },
    {
      "entropy": 1.9032898783683776,
      "epoch": 0.022226372660631655,
      "grad_norm": 15.8334379196167,
      "learning_rate": 3.4723433110529886e-06,
      "loss": 0.6952,
      "mean_token_accuracy": 0.7943542674183846,
      "num_tokens": 8623666.0,
      "step": 7170
    },
    {
      "entropy": 1.8691613361239434,
      "epoch": 0.022257371785681348,
      "grad_norm": 11.972046852111816,
      "learning_rate": 3.477186864283639e-06,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8114256381988525,
      "num_tokens": 8635505.0,
      "step": 7180
    },
    {
      "entropy": 1.8365458711981772,
      "epoch": 0.022288370910731044,
      "grad_norm": 14.07669448852539,
      "learning_rate": 3.4820304175142887e-06,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8142189055681228,
      "num_tokens": 8648420.0,
      "step": 7190
    },
    {
      "entropy": 1.8941682323813438,
      "epoch": 0.02231937003578074,
      "grad_norm": 12.625129699707031,
      "learning_rate": 3.486873970744939e-06,
      "loss": 0.6311,
      "mean_token_accuracy": 0.8075986996293067,
      "num_tokens": 8661128.0,
      "step": 7200
    },
    {
      "entropy": 1.8504259541630745,
      "epoch": 0.022350369160830434,
      "grad_norm": 6.088210105895996,
      "learning_rate": 3.491717523975589e-06,
      "loss": 0.6102,
      "mean_token_accuracy": 0.8124124839901924,
      "num_tokens": 8672830.0,
      "step": 7210
    },
    {
      "entropy": 1.8653683230280875,
      "epoch": 0.02238136828588013,
      "grad_norm": 15.06690502166748,
      "learning_rate": 3.496561077206239e-06,
      "loss": 0.6134,
      "mean_token_accuracy": 0.8111754149198532,
      "num_tokens": 8684754.0,
      "step": 7220
    },
    {
      "entropy": 1.9144769728183746,
      "epoch": 0.022412367410929827,
      "grad_norm": 13.07332992553711,
      "learning_rate": 3.5014046304368887e-06,
      "loss": 0.7025,
      "mean_token_accuracy": 0.8014708399772644,
      "num_tokens": 8695969.0,
      "step": 7230
    },
    {
      "entropy": 1.8702037513256073,
      "epoch": 0.02244336653597952,
      "grad_norm": 13.671896934509277,
      "learning_rate": 3.506248183667539e-06,
      "loss": 0.657,
      "mean_token_accuracy": 0.809376485645771,
      "num_tokens": 8707677.0,
      "step": 7240
    },
    {
      "entropy": 1.9119601994752884,
      "epoch": 0.022474365661029217,
      "grad_norm": 16.43385124206543,
      "learning_rate": 3.511091736898189e-06,
      "loss": 0.6931,
      "mean_token_accuracy": 0.7956320688128471,
      "num_tokens": 8718648.0,
      "step": 7250
    },
    {
      "entropy": 1.8430123403668404,
      "epoch": 0.022505364786078914,
      "grad_norm": 5.884398937225342,
      "learning_rate": 3.515935290128839e-06,
      "loss": 0.6187,
      "mean_token_accuracy": 0.8040406808257103,
      "num_tokens": 8731028.0,
      "step": 7260
    },
    {
      "entropy": 1.7957608908414842,
      "epoch": 0.022536363911128607,
      "grad_norm": 6.280706882476807,
      "learning_rate": 3.5207788433594887e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8192771345376968,
      "num_tokens": 8743771.0,
      "step": 7270
    },
    {
      "entropy": 1.8952230989933014,
      "epoch": 0.022567363036178303,
      "grad_norm": 11.960763931274414,
      "learning_rate": 3.5256223965901385e-06,
      "loss": 0.6716,
      "mean_token_accuracy": 0.7986912429332733,
      "num_tokens": 8754987.0,
      "step": 7280
    },
    {
      "entropy": 1.7763853132724763,
      "epoch": 0.022598362161228,
      "grad_norm": 11.904229164123535,
      "learning_rate": 3.530465949820789e-06,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8182390749454498,
      "num_tokens": 8767593.0,
      "step": 7290
    },
    {
      "entropy": 1.8223341763019563,
      "epoch": 0.022629361286277693,
      "grad_norm": 13.150022506713867,
      "learning_rate": 3.5353095030514387e-06,
      "loss": 0.6274,
      "mean_token_accuracy": 0.8080811381340027,
      "num_tokens": 8779386.0,
      "step": 7300
    },
    {
      "entropy": 1.8453481420874596,
      "epoch": 0.02266036041132739,
      "grad_norm": 16.510391235351562,
      "learning_rate": 3.540153056282089e-06,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8153272330760956,
      "num_tokens": 8791800.0,
      "step": 7310
    },
    {
      "entropy": 1.8597090423107148,
      "epoch": 0.022691359536377086,
      "grad_norm": 11.312858581542969,
      "learning_rate": 3.5449966095127385e-06,
      "loss": 0.6451,
      "mean_token_accuracy": 0.8115374326705933,
      "num_tokens": 8802831.0,
      "step": 7320
    },
    {
      "entropy": 1.7629368484020234,
      "epoch": 0.022722358661426783,
      "grad_norm": 10.922659873962402,
      "learning_rate": 3.549840162743389e-06,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8109967395663261,
      "num_tokens": 8816085.0,
      "step": 7330
    },
    {
      "entropy": 1.8767753183841704,
      "epoch": 0.022753357786476476,
      "grad_norm": 12.062516212463379,
      "learning_rate": 3.5546837159740387e-06,
      "loss": 0.6765,
      "mean_token_accuracy": 0.7968204706907273,
      "num_tokens": 8828125.0,
      "step": 7340
    },
    {
      "entropy": 1.798590750992298,
      "epoch": 0.022784356911526173,
      "grad_norm": 6.616394996643066,
      "learning_rate": 3.559527269204689e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8197192326188087,
      "num_tokens": 8840873.0,
      "step": 7350
    },
    {
      "entropy": 1.8430965647101403,
      "epoch": 0.02281535603657587,
      "grad_norm": 12.304119110107422,
      "learning_rate": 3.564370822435339e-06,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8144068956375122,
      "num_tokens": 8853322.0,
      "step": 7360
    },
    {
      "entropy": 1.8360921517014503,
      "epoch": 0.022846355161625562,
      "grad_norm": 13.151806831359863,
      "learning_rate": 3.5692143756659888e-06,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8184613898396492,
      "num_tokens": 8866377.0,
      "step": 7370
    },
    {
      "entropy": 1.8777944207191468,
      "epoch": 0.02287735428667526,
      "grad_norm": 11.325216293334961,
      "learning_rate": 3.5740579288966387e-06,
      "loss": 0.6267,
      "mean_token_accuracy": 0.8124101653695106,
      "num_tokens": 8877572.0,
      "step": 7380
    },
    {
      "entropy": 1.8015395179390907,
      "epoch": 0.022908353411724956,
      "grad_norm": 13.632472038269043,
      "learning_rate": 3.578901482127289e-06,
      "loss": 0.5888,
      "mean_token_accuracy": 0.8304689675569534,
      "num_tokens": 8890279.0,
      "step": 7390
    },
    {
      "entropy": 1.9334001183509826,
      "epoch": 0.02293935253677465,
      "grad_norm": 18.318246841430664,
      "learning_rate": 3.583745035357939e-06,
      "loss": 0.6549,
      "mean_token_accuracy": 0.8056276828050614,
      "num_tokens": 8901650.0,
      "step": 7400
    },
    {
      "entropy": 1.8892749726772309,
      "epoch": 0.022970351661824345,
      "grad_norm": 13.172524452209473,
      "learning_rate": 3.588588588588589e-06,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8132000923156738,
      "num_tokens": 8912513.0,
      "step": 7410
    },
    {
      "entropy": 1.9355769097805022,
      "epoch": 0.023001350786874042,
      "grad_norm": 12.621254920959473,
      "learning_rate": 3.593432141819239e-06,
      "loss": 0.6664,
      "mean_token_accuracy": 0.8047688767313957,
      "num_tokens": 8923668.0,
      "step": 7420
    },
    {
      "entropy": 1.7977433398365974,
      "epoch": 0.023032349911923735,
      "grad_norm": 14.412604331970215,
      "learning_rate": 3.598275695049889e-06,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8217224717140198,
      "num_tokens": 8936743.0,
      "step": 7430
    },
    {
      "entropy": 1.8730682261288165,
      "epoch": 0.02306334903697343,
      "grad_norm": 18.06209373474121,
      "learning_rate": 3.603119248280539e-06,
      "loss": 0.687,
      "mean_token_accuracy": 0.7976897180080413,
      "num_tokens": 8948361.0,
      "step": 7440
    },
    {
      "entropy": 1.8379228845238686,
      "epoch": 0.023094348162023128,
      "grad_norm": 16.5709171295166,
      "learning_rate": 3.607962801511189e-06,
      "loss": 0.647,
      "mean_token_accuracy": 0.8052407920360565,
      "num_tokens": 8960081.0,
      "step": 7450
    },
    {
      "entropy": 1.8426879778504373,
      "epoch": 0.02312534728707282,
      "grad_norm": 6.525999546051025,
      "learning_rate": 3.612806354741839e-06,
      "loss": 0.6043,
      "mean_token_accuracy": 0.8133143231272697,
      "num_tokens": 8972165.0,
      "step": 7460
    },
    {
      "entropy": 1.8481390669941902,
      "epoch": 0.023156346412122518,
      "grad_norm": 11.893180847167969,
      "learning_rate": 3.6176499079724893e-06,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8177680626511574,
      "num_tokens": 8983683.0,
      "step": 7470
    },
    {
      "entropy": 1.8435255289077759,
      "epoch": 0.023187345537172215,
      "grad_norm": 13.133444786071777,
      "learning_rate": 3.6224934612031388e-06,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8149840265512467,
      "num_tokens": 8994970.0,
      "step": 7480
    },
    {
      "entropy": 1.8090941041707993,
      "epoch": 0.023218344662221908,
      "grad_norm": 5.264407157897949,
      "learning_rate": 3.6273370144337886e-06,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8030720934271812,
      "num_tokens": 9007512.0,
      "step": 7490
    },
    {
      "entropy": 1.8767491683363915,
      "epoch": 0.023249343787271604,
      "grad_norm": 12.091042518615723,
      "learning_rate": 3.632180567664439e-06,
      "loss": 0.6655,
      "mean_token_accuracy": 0.8082658454775811,
      "num_tokens": 9018984.0,
      "step": 7500
    },
    {
      "entropy": 1.8025119572877883,
      "epoch": 0.0232803429123213,
      "grad_norm": 14.580635070800781,
      "learning_rate": 3.637024120895089e-06,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8104995712637901,
      "num_tokens": 9032061.0,
      "step": 7510
    },
    {
      "entropy": 1.807465235888958,
      "epoch": 0.023311342037370994,
      "grad_norm": 12.528902053833008,
      "learning_rate": 3.641867674125739e-06,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8124234288930893,
      "num_tokens": 9044338.0,
      "step": 7520
    },
    {
      "entropy": 1.8326406210660935,
      "epoch": 0.02334234116242069,
      "grad_norm": 15.056619644165039,
      "learning_rate": 3.6467112273563886e-06,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8178173303604126,
      "num_tokens": 9056355.0,
      "step": 7530
    },
    {
      "entropy": 1.7988777324557303,
      "epoch": 0.023373340287470387,
      "grad_norm": 12.382633209228516,
      "learning_rate": 3.651554780587039e-06,
      "loss": 0.638,
      "mean_token_accuracy": 0.8013660088181496,
      "num_tokens": 9068937.0,
      "step": 7540
    },
    {
      "entropy": 1.8599283427000046,
      "epoch": 0.02340433941252008,
      "grad_norm": 11.771322250366211,
      "learning_rate": 3.656398333817689e-06,
      "loss": 0.644,
      "mean_token_accuracy": 0.8095093712210655,
      "num_tokens": 9080598.0,
      "step": 7550
    },
    {
      "entropy": 1.9167393915355206,
      "epoch": 0.023435338537569777,
      "grad_norm": 15.524273872375488,
      "learning_rate": 3.661241887048339e-06,
      "loss": 0.6336,
      "mean_token_accuracy": 0.807981975376606,
      "num_tokens": 9091965.0,
      "step": 7560
    },
    {
      "entropy": 1.9095857173204422,
      "epoch": 0.023466337662619473,
      "grad_norm": 11.96094799041748,
      "learning_rate": 3.666085440278989e-06,
      "loss": 0.6676,
      "mean_token_accuracy": 0.7986720770597457,
      "num_tokens": 9103815.0,
      "step": 7570
    },
    {
      "entropy": 1.8987512320280076,
      "epoch": 0.023497336787669167,
      "grad_norm": 13.277989387512207,
      "learning_rate": 3.670928993509639e-06,
      "loss": 0.6695,
      "mean_token_accuracy": 0.8065518125891685,
      "num_tokens": 9115160.0,
      "step": 7580
    },
    {
      "entropy": 1.8922671616077422,
      "epoch": 0.023528335912718863,
      "grad_norm": 13.145485877990723,
      "learning_rate": 3.6757725467402888e-06,
      "loss": 0.685,
      "mean_token_accuracy": 0.8018161416053772,
      "num_tokens": 9126074.0,
      "step": 7590
    },
    {
      "entropy": 1.8771102100610733,
      "epoch": 0.02355933503776856,
      "grad_norm": 10.288253784179688,
      "learning_rate": 3.680616099970939e-06,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8214558124542236,
      "num_tokens": 9138014.0,
      "step": 7600
    },
    {
      "entropy": 1.8878172472119332,
      "epoch": 0.023590334162818253,
      "grad_norm": 12.76760482788086,
      "learning_rate": 3.685459653201589e-06,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8086960777640343,
      "num_tokens": 9149156.0,
      "step": 7610
    },
    {
      "entropy": 1.8934050709009171,
      "epoch": 0.02362133328786795,
      "grad_norm": 12.631563186645508,
      "learning_rate": 3.6903032064322393e-06,
      "loss": 0.7241,
      "mean_token_accuracy": 0.7934870898723603,
      "num_tokens": 9161339.0,
      "step": 7620
    },
    {
      "entropy": 1.9421094968914985,
      "epoch": 0.023652332412917646,
      "grad_norm": 14.10696792602539,
      "learning_rate": 3.695146759662889e-06,
      "loss": 0.6605,
      "mean_token_accuracy": 0.7969638511538506,
      "num_tokens": 9172617.0,
      "step": 7630
    },
    {
      "entropy": 1.8514475598931313,
      "epoch": 0.02368333153796734,
      "grad_norm": 13.774374008178711,
      "learning_rate": 3.699990312893539e-06,
      "loss": 0.6989,
      "mean_token_accuracy": 0.7987961351871491,
      "num_tokens": 9184783.0,
      "step": 7640
    },
    {
      "entropy": 1.7891797423362732,
      "epoch": 0.023714330663017036,
      "grad_norm": 5.731255054473877,
      "learning_rate": 3.704833866124189e-06,
      "loss": 0.595,
      "mean_token_accuracy": 0.8199530601501465,
      "num_tokens": 9197781.0,
      "step": 7650
    },
    {
      "entropy": 1.8698611691594125,
      "epoch": 0.023745329788066732,
      "grad_norm": 12.50722599029541,
      "learning_rate": 3.7096774193548392e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.822203965485096,
      "num_tokens": 9210004.0,
      "step": 7660
    },
    {
      "entropy": 1.9275674849748612,
      "epoch": 0.023776328913116426,
      "grad_norm": 11.671341896057129,
      "learning_rate": 3.714520972585489e-06,
      "loss": 0.6877,
      "mean_token_accuracy": 0.7972215577960015,
      "num_tokens": 9221045.0,
      "step": 7670
    },
    {
      "entropy": 1.8518989741802216,
      "epoch": 0.023807328038166122,
      "grad_norm": 12.143589973449707,
      "learning_rate": 3.7193645258161394e-06,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8160637483000756,
      "num_tokens": 9233582.0,
      "step": 7680
    },
    {
      "entropy": 1.9745838671922684,
      "epoch": 0.02383832716321582,
      "grad_norm": 8.162403106689453,
      "learning_rate": 3.724208079046789e-06,
      "loss": 0.6552,
      "mean_token_accuracy": 0.8079164877533913,
      "num_tokens": 9244759.0,
      "step": 7690
    },
    {
      "entropy": 1.906966118514538,
      "epoch": 0.023869326288265515,
      "grad_norm": 5.672482490539551,
      "learning_rate": 3.7290516322774388e-06,
      "loss": 0.6534,
      "mean_token_accuracy": 0.8026331990957261,
      "num_tokens": 9256899.0,
      "step": 7700
    },
    {
      "entropy": 1.8956855058670044,
      "epoch": 0.02390032541331521,
      "grad_norm": 11.346663475036621,
      "learning_rate": 3.733895185508089e-06,
      "loss": 0.5923,
      "mean_token_accuracy": 0.8169514790177346,
      "num_tokens": 9268280.0,
      "step": 7710
    },
    {
      "entropy": 1.9229377016425133,
      "epoch": 0.023931324538364905,
      "grad_norm": 6.657054424285889,
      "learning_rate": 3.7387387387387394e-06,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8021419748663903,
      "num_tokens": 9280505.0,
      "step": 7720
    },
    {
      "entropy": 1.7850939154624939,
      "epoch": 0.0239623236634146,
      "grad_norm": 3.4274275302886963,
      "learning_rate": 3.7435822919693893e-06,
      "loss": 0.5326,
      "mean_token_accuracy": 0.829072143137455,
      "num_tokens": 9293099.0,
      "step": 7730
    },
    {
      "entropy": 1.9557977497577668,
      "epoch": 0.023993322788464295,
      "grad_norm": 13.326607704162598,
      "learning_rate": 3.7484258452000387e-06,
      "loss": 0.6884,
      "mean_token_accuracy": 0.798516571521759,
      "num_tokens": 9304408.0,
      "step": 7740
    },
    {
      "entropy": 1.8509003862738609,
      "epoch": 0.02402432191351399,
      "grad_norm": 13.444816589355469,
      "learning_rate": 3.753269398430689e-06,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8102645307779313,
      "num_tokens": 9316689.0,
      "step": 7750
    },
    {
      "entropy": 1.7711122930049896,
      "epoch": 0.024055321038563688,
      "grad_norm": 5.156175136566162,
      "learning_rate": 3.758112951661339e-06,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8136220246553421,
      "num_tokens": 9329593.0,
      "step": 7760
    },
    {
      "entropy": 1.8134596601128579,
      "epoch": 0.02408632016361338,
      "grad_norm": 12.387152671813965,
      "learning_rate": 3.7629565048919892e-06,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8308922097086906,
      "num_tokens": 9341486.0,
      "step": 7770
    },
    {
      "entropy": 1.8755627959966659,
      "epoch": 0.024117319288663078,
      "grad_norm": 12.338440895080566,
      "learning_rate": 3.767800058122639e-06,
      "loss": 0.6576,
      "mean_token_accuracy": 0.8139973327517509,
      "num_tokens": 9352093.0,
      "step": 7780
    },
    {
      "entropy": 1.7985035717487334,
      "epoch": 0.024148318413712774,
      "grad_norm": 15.153599739074707,
      "learning_rate": 3.772643611353289e-06,
      "loss": 0.5679,
      "mean_token_accuracy": 0.8182680875062942,
      "num_tokens": 9364468.0,
      "step": 7790
    },
    {
      "entropy": 1.841192065924406,
      "epoch": 0.024179317538762467,
      "grad_norm": 4.2616729736328125,
      "learning_rate": 3.777487164583939e-06,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8146574392914772,
      "num_tokens": 9377926.0,
      "step": 7800
    },
    {
      "entropy": 1.920288896560669,
      "epoch": 0.024210316663812164,
      "grad_norm": 14.35584545135498,
      "learning_rate": 3.782330717814589e-06,
      "loss": 0.7463,
      "mean_token_accuracy": 0.7909737974405289,
      "num_tokens": 9389642.0,
      "step": 7810
    },
    {
      "entropy": 1.8628991067409515,
      "epoch": 0.02424131578886186,
      "grad_norm": 15.304876327514648,
      "learning_rate": 3.787174271045239e-06,
      "loss": 0.586,
      "mean_token_accuracy": 0.8081997647881508,
      "num_tokens": 9402632.0,
      "step": 7820
    },
    {
      "entropy": 1.9279965221881867,
      "epoch": 0.024272314913911554,
      "grad_norm": 14.144513130187988,
      "learning_rate": 3.7920178242758894e-06,
      "loss": 0.6125,
      "mean_token_accuracy": 0.8075599849224091,
      "num_tokens": 9414752.0,
      "step": 7830
    },
    {
      "entropy": 1.8668213859200478,
      "epoch": 0.02430331403896125,
      "grad_norm": 13.267974853515625,
      "learning_rate": 3.7968613775065392e-06,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8120471283793449,
      "num_tokens": 9426149.0,
      "step": 7840
    },
    {
      "entropy": 1.9365358769893646,
      "epoch": 0.024334313164010947,
      "grad_norm": 13.412445068359375,
      "learning_rate": 3.801704930737189e-06,
      "loss": 0.6644,
      "mean_token_accuracy": 0.8024364963173867,
      "num_tokens": 9437306.0,
      "step": 7850
    },
    {
      "entropy": 1.9267346888780594,
      "epoch": 0.02436531228906064,
      "grad_norm": 13.718780517578125,
      "learning_rate": 3.806548483967839e-06,
      "loss": 0.6209,
      "mean_token_accuracy": 0.7963616490364075,
      "num_tokens": 9449401.0,
      "step": 7860
    },
    {
      "entropy": 1.8495199769735335,
      "epoch": 0.024396311414110337,
      "grad_norm": 11.75393009185791,
      "learning_rate": 3.8113920371984893e-06,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8257737800478935,
      "num_tokens": 9461930.0,
      "step": 7870
    },
    {
      "entropy": 1.8680783659219742,
      "epoch": 0.024427310539160033,
      "grad_norm": 10.541101455688477,
      "learning_rate": 3.81623559042914e-06,
      "loss": 0.5844,
      "mean_token_accuracy": 0.8095761522650718,
      "num_tokens": 9473753.0,
      "step": 7880
    },
    {
      "entropy": 1.874305109679699,
      "epoch": 0.024458309664209726,
      "grad_norm": 17.521041870117188,
      "learning_rate": 3.8210791436597895e-06,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8210688814520836,
      "num_tokens": 9486165.0,
      "step": 7890
    },
    {
      "entropy": 1.9577010348439217,
      "epoch": 0.024489308789259423,
      "grad_norm": 11.793618202209473,
      "learning_rate": 3.8259226968904386e-06,
      "loss": 0.656,
      "mean_token_accuracy": 0.8024436876177787,
      "num_tokens": 9497697.0,
      "step": 7900
    },
    {
      "entropy": 1.9006599888205529,
      "epoch": 0.02452030791430912,
      "grad_norm": 13.784941673278809,
      "learning_rate": 3.830766250121089e-06,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8112906947731972,
      "num_tokens": 9510027.0,
      "step": 7910
    },
    {
      "entropy": 1.943654653429985,
      "epoch": 0.024551307039358813,
      "grad_norm": 12.328388214111328,
      "learning_rate": 3.835609803351739e-06,
      "loss": 0.6456,
      "mean_token_accuracy": 0.8070406153798103,
      "num_tokens": 9521309.0,
      "step": 7920
    },
    {
      "entropy": 1.8668604627251626,
      "epoch": 0.02458230616440851,
      "grad_norm": 13.092666625976562,
      "learning_rate": 3.840453356582389e-06,
      "loss": 0.6008,
      "mean_token_accuracy": 0.8188393861055374,
      "num_tokens": 9533756.0,
      "step": 7930
    },
    {
      "entropy": 1.8963513985276221,
      "epoch": 0.024613305289458206,
      "grad_norm": 12.246744155883789,
      "learning_rate": 3.84529690981304e-06,
      "loss": 0.6602,
      "mean_token_accuracy": 0.8034892365336418,
      "num_tokens": 9545338.0,
      "step": 7940
    },
    {
      "entropy": 1.8683793991804123,
      "epoch": 0.0246443044145079,
      "grad_norm": 5.166250228881836,
      "learning_rate": 3.850140463043689e-06,
      "loss": 0.6427,
      "mean_token_accuracy": 0.8070285364985466,
      "num_tokens": 9558352.0,
      "step": 7950
    },
    {
      "entropy": 1.848561166226864,
      "epoch": 0.024675303539557596,
      "grad_norm": 12.724311828613281,
      "learning_rate": 3.854984016274339e-06,
      "loss": 0.603,
      "mean_token_accuracy": 0.8135858446359634,
      "num_tokens": 9570765.0,
      "step": 7960
    },
    {
      "entropy": 1.9112862929701806,
      "epoch": 0.024706302664607292,
      "grad_norm": 12.425032615661621,
      "learning_rate": 3.8598275695049894e-06,
      "loss": 0.6219,
      "mean_token_accuracy": 0.8120498448610306,
      "num_tokens": 9581809.0,
      "step": 7970
    },
    {
      "entropy": 1.9325566202402116,
      "epoch": 0.024737301789656985,
      "grad_norm": 12.291463851928711,
      "learning_rate": 3.864671122735639e-06,
      "loss": 0.7302,
      "mean_token_accuracy": 0.7900342762470245,
      "num_tokens": 9592614.0,
      "step": 7980
    },
    {
      "entropy": 1.8553043097257613,
      "epoch": 0.024768300914706682,
      "grad_norm": 14.992151260375977,
      "learning_rate": 3.869514675966289e-06,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8028819039463997,
      "num_tokens": 9604643.0,
      "step": 7990
    },
    {
      "entropy": 1.86794516146183,
      "epoch": 0.02479930003975638,
      "grad_norm": 11.671442985534668,
      "learning_rate": 3.874358229196939e-06,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8125581681728363,
      "num_tokens": 9617101.0,
      "step": 8000
    },
    {
      "entropy": 1.977939623594284,
      "epoch": 0.02483029916480607,
      "grad_norm": 14.295232772827148,
      "learning_rate": 3.879201782427589e-06,
      "loss": 0.7147,
      "mean_token_accuracy": 0.7930570155382156,
      "num_tokens": 9628082.0,
      "step": 8010
    },
    {
      "entropy": 1.9121051743626594,
      "epoch": 0.02486129828985577,
      "grad_norm": 13.757525444030762,
      "learning_rate": 3.884045335658239e-06,
      "loss": 0.6761,
      "mean_token_accuracy": 0.8087022632360459,
      "num_tokens": 9639932.0,
      "step": 8020
    },
    {
      "entropy": 1.841791082918644,
      "epoch": 0.024892297414905465,
      "grad_norm": 6.790825366973877,
      "learning_rate": 3.88888888888889e-06,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8306658193469048,
      "num_tokens": 9652182.0,
      "step": 8030
    },
    {
      "entropy": 1.937432810664177,
      "epoch": 0.02492329653995516,
      "grad_norm": 13.771592140197754,
      "learning_rate": 3.8937324421195395e-06,
      "loss": 0.635,
      "mean_token_accuracy": 0.8062109544873237,
      "num_tokens": 9664539.0,
      "step": 8040
    },
    {
      "entropy": 1.8707983300089837,
      "epoch": 0.024954295665004855,
      "grad_norm": 12.660676956176758,
      "learning_rate": 3.898575995350189e-06,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8023678585886955,
      "num_tokens": 9676330.0,
      "step": 8050
    },
    {
      "entropy": 1.9288925692439078,
      "epoch": 0.02498529479005455,
      "grad_norm": 14.32672119140625,
      "learning_rate": 3.903419548580839e-06,
      "loss": 0.65,
      "mean_token_accuracy": 0.8022472143173218,
      "num_tokens": 9687521.0,
      "step": 8060
    },
    {
      "entropy": 1.84913961738348,
      "epoch": 0.025016293915104248,
      "grad_norm": 6.136168003082275,
      "learning_rate": 3.908263101811489e-06,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8220839589834213,
      "num_tokens": 9699208.0,
      "step": 8070
    },
    {
      "entropy": 1.8712972477078438,
      "epoch": 0.02504729304015394,
      "grad_norm": 11.905717849731445,
      "learning_rate": 3.913106655042139e-06,
      "loss": 0.6119,
      "mean_token_accuracy": 0.8191757917404174,
      "num_tokens": 9711146.0,
      "step": 8080
    },
    {
      "entropy": 1.9155054926872253,
      "epoch": 0.025078292165203638,
      "grad_norm": 12.616907119750977,
      "learning_rate": 3.91795020827279e-06,
      "loss": 0.689,
      "mean_token_accuracy": 0.8038010418415069,
      "num_tokens": 9722222.0,
      "step": 8090
    },
    {
      "entropy": 1.9103090360760688,
      "epoch": 0.025109291290253334,
      "grad_norm": 12.909311294555664,
      "learning_rate": 3.92279376150344e-06,
      "loss": 0.6034,
      "mean_token_accuracy": 0.812662410736084,
      "num_tokens": 9733890.0,
      "step": 8100
    },
    {
      "entropy": 1.8312827795743942,
      "epoch": 0.025140290415303027,
      "grad_norm": 14.832688331604004,
      "learning_rate": 3.927637314734089e-06,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8249428883194924,
      "num_tokens": 9746809.0,
      "step": 8110
    },
    {
      "entropy": 1.8507552802562715,
      "epoch": 0.025171289540352724,
      "grad_norm": 13.549945831298828,
      "learning_rate": 3.932480867964739e-06,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8166124686598778,
      "num_tokens": 9759355.0,
      "step": 8120
    },
    {
      "entropy": 1.8369555801153183,
      "epoch": 0.02520228866540242,
      "grad_norm": 11.552787780761719,
      "learning_rate": 3.937324421195389e-06,
      "loss": 0.5874,
      "mean_token_accuracy": 0.825542688369751,
      "num_tokens": 9771999.0,
      "step": 8130
    },
    {
      "entropy": 1.8414116382598877,
      "epoch": 0.025233287790452114,
      "grad_norm": 12.822636604309082,
      "learning_rate": 3.942167974426039e-06,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8180017948150635,
      "num_tokens": 9784135.0,
      "step": 8140
    },
    {
      "entropy": 1.847972247004509,
      "epoch": 0.02526428691550181,
      "grad_norm": 6.315905570983887,
      "learning_rate": 3.94701152765669e-06,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8069576799869538,
      "num_tokens": 9795872.0,
      "step": 8150
    },
    {
      "entropy": 1.8612953543663024,
      "epoch": 0.025295286040551507,
      "grad_norm": 7.054393768310547,
      "learning_rate": 3.951855080887339e-06,
      "loss": 0.6359,
      "mean_token_accuracy": 0.808410094678402,
      "num_tokens": 9807148.0,
      "step": 8160
    },
    {
      "entropy": 1.745456214249134,
      "epoch": 0.0253262851656012,
      "grad_norm": 10.395822525024414,
      "learning_rate": 3.956698634117989e-06,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8252757102251053,
      "num_tokens": 9820692.0,
      "step": 8170
    },
    {
      "entropy": 1.933900985121727,
      "epoch": 0.025357284290650896,
      "grad_norm": 15.611356735229492,
      "learning_rate": 3.9615421873486395e-06,
      "loss": 0.8023,
      "mean_token_accuracy": 0.7968411594629288,
      "num_tokens": 9832230.0,
      "step": 8180
    },
    {
      "entropy": 1.8410508632659912,
      "epoch": 0.025388283415700593,
      "grad_norm": 15.906883239746094,
      "learning_rate": 3.966385740579289e-06,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8164623320102692,
      "num_tokens": 9844672.0,
      "step": 8190
    },
    {
      "entropy": 1.9910759955644608,
      "epoch": 0.025419282540750286,
      "grad_norm": 13.914506912231445,
      "learning_rate": 3.971229293809939e-06,
      "loss": 0.7414,
      "mean_token_accuracy": 0.7940586417913437,
      "num_tokens": 9856014.0,
      "step": 8200
    },
    {
      "entropy": 1.9779651939868927,
      "epoch": 0.025450281665799983,
      "grad_norm": 12.388845443725586,
      "learning_rate": 3.976072847040589e-06,
      "loss": 0.6406,
      "mean_token_accuracy": 0.810752010345459,
      "num_tokens": 9867585.0,
      "step": 8210
    },
    {
      "entropy": 1.950091141462326,
      "epoch": 0.02548128079084968,
      "grad_norm": 15.438441276550293,
      "learning_rate": 3.980916400271239e-06,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8093546569347382,
      "num_tokens": 9879228.0,
      "step": 8220
    },
    {
      "entropy": 1.8706778109073638,
      "epoch": 0.025512279915899373,
      "grad_norm": 10.18140983581543,
      "learning_rate": 3.985759953501889e-06,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8231084540486335,
      "num_tokens": 9892299.0,
      "step": 8230
    },
    {
      "entropy": 1.9114290565252303,
      "epoch": 0.02554327904094907,
      "grad_norm": 11.814135551452637,
      "learning_rate": 3.99060350673254e-06,
      "loss": 0.6377,
      "mean_token_accuracy": 0.8094885662198067,
      "num_tokens": 9905027.0,
      "step": 8240
    },
    {
      "entropy": 1.8559225603938103,
      "epoch": 0.025574278165998766,
      "grad_norm": 16.11547088623047,
      "learning_rate": 3.9954470599631896e-06,
      "loss": 0.6124,
      "mean_token_accuracy": 0.8048399239778519,
      "num_tokens": 9917498.0,
      "step": 8250
    },
    {
      "entropy": 1.9120782747864724,
      "epoch": 0.02560527729104846,
      "grad_norm": 6.817448616027832,
      "learning_rate": 4.0002906131938395e-06,
      "loss": 0.6091,
      "mean_token_accuracy": 0.8105715274810791,
      "num_tokens": 9929963.0,
      "step": 8260
    },
    {
      "entropy": 1.8872886016964912,
      "epoch": 0.025636276416098155,
      "grad_norm": 13.45474624633789,
      "learning_rate": 4.005134166424489e-06,
      "loss": 0.5923,
      "mean_token_accuracy": 0.8145119935274124,
      "num_tokens": 9942393.0,
      "step": 8270
    },
    {
      "entropy": 1.9791032537817954,
      "epoch": 0.025667275541147852,
      "grad_norm": 12.72803020477295,
      "learning_rate": 4.009977719655139e-06,
      "loss": 0.6922,
      "mean_token_accuracy": 0.7994261741638183,
      "num_tokens": 9953810.0,
      "step": 8280
    },
    {
      "entropy": 1.9359009936451912,
      "epoch": 0.025698274666197545,
      "grad_norm": 11.705611228942871,
      "learning_rate": 4.014821272885789e-06,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8066216111183167,
      "num_tokens": 9965045.0,
      "step": 8290
    },
    {
      "entropy": 1.8848050013184547,
      "epoch": 0.025729273791247242,
      "grad_norm": 14.14693832397461,
      "learning_rate": 4.01966482611644e-06,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8066284090280533,
      "num_tokens": 9977622.0,
      "step": 8300
    },
    {
      "entropy": 1.9196738287806512,
      "epoch": 0.02576027291629694,
      "grad_norm": 13.333892822265625,
      "learning_rate": 4.02450837934709e-06,
      "loss": 0.6378,
      "mean_token_accuracy": 0.8114329054951668,
      "num_tokens": 9989396.0,
      "step": 8310
    },
    {
      "entropy": 1.9220668867230415,
      "epoch": 0.02579127204134663,
      "grad_norm": 8.307788848876953,
      "learning_rate": 4.029351932577739e-06,
      "loss": 0.676,
      "mean_token_accuracy": 0.8107616409659386,
      "num_tokens": 10001039.0,
      "step": 8320
    },
    {
      "entropy": 1.924265030026436,
      "epoch": 0.025822271166396328,
      "grad_norm": 13.42192554473877,
      "learning_rate": 4.0341954858083895e-06,
      "loss": 0.6432,
      "mean_token_accuracy": 0.8023953422904014,
      "num_tokens": 10012862.0,
      "step": 8330
    },
    {
      "entropy": 1.8803733557462692,
      "epoch": 0.025853270291446025,
      "grad_norm": 14.513736724853516,
      "learning_rate": 4.039039039039039e-06,
      "loss": 0.614,
      "mean_token_accuracy": 0.8223489403724671,
      "num_tokens": 10024615.0,
      "step": 8340
    },
    {
      "entropy": 1.8715410023927688,
      "epoch": 0.025884269416495718,
      "grad_norm": 13.777809143066406,
      "learning_rate": 4.043882592269689e-06,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8061651542782784,
      "num_tokens": 10036943.0,
      "step": 8350
    },
    {
      "entropy": 1.8626436039805412,
      "epoch": 0.025915268541545414,
      "grad_norm": 5.88369607925415,
      "learning_rate": 4.04872614550034e-06,
      "loss": 0.5776,
      "mean_token_accuracy": 0.816595071554184,
      "num_tokens": 10049582.0,
      "step": 8360
    },
    {
      "entropy": 1.893427351117134,
      "epoch": 0.02594626766659511,
      "grad_norm": 5.693576335906982,
      "learning_rate": 4.053569698730989e-06,
      "loss": 0.5993,
      "mean_token_accuracy": 0.8074675589799881,
      "num_tokens": 10061474.0,
      "step": 8370
    },
    {
      "entropy": 1.9071956276893616,
      "epoch": 0.025977266791644804,
      "grad_norm": 6.2122063636779785,
      "learning_rate": 4.058413251961639e-06,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8195055171847343,
      "num_tokens": 10073377.0,
      "step": 8380
    },
    {
      "entropy": 1.9347770288586617,
      "epoch": 0.0260082659166945,
      "grad_norm": 6.3852434158325195,
      "learning_rate": 4.06325680519229e-06,
      "loss": 0.6468,
      "mean_token_accuracy": 0.7985034555196762,
      "num_tokens": 10085414.0,
      "step": 8390
    },
    {
      "entropy": 1.879788914322853,
      "epoch": 0.026039265041744197,
      "grad_norm": 5.960961818695068,
      "learning_rate": 4.0681003584229395e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8187567561864852,
      "num_tokens": 10098381.0,
      "step": 8400
    },
    {
      "entropy": 1.915135058760643,
      "epoch": 0.026070264166793894,
      "grad_norm": 5.453474521636963,
      "learning_rate": 4.072943911653589e-06,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8210947424173355,
      "num_tokens": 10110037.0,
      "step": 8410
    },
    {
      "entropy": 1.9362727150321006,
      "epoch": 0.026101263291843587,
      "grad_norm": 14.594948768615723,
      "learning_rate": 4.077787464884239e-06,
      "loss": 0.6608,
      "mean_token_accuracy": 0.7989747643470764,
      "num_tokens": 10121901.0,
      "step": 8420
    },
    {
      "entropy": 1.8305003002285958,
      "epoch": 0.026132262416893284,
      "grad_norm": 14.329607963562012,
      "learning_rate": 4.082631018114889e-06,
      "loss": 0.5911,
      "mean_token_accuracy": 0.8075525343418122,
      "num_tokens": 10134905.0,
      "step": 8430
    },
    {
      "entropy": 1.8578375533223153,
      "epoch": 0.02616326154194298,
      "grad_norm": 13.74600887298584,
      "learning_rate": 4.087474571345539e-06,
      "loss": 0.5891,
      "mean_token_accuracy": 0.8238616168498993,
      "num_tokens": 10146939.0,
      "step": 8440
    },
    {
      "entropy": 1.9704759448766709,
      "epoch": 0.026194260666992673,
      "grad_norm": 12.829072952270508,
      "learning_rate": 4.09231812457619e-06,
      "loss": 0.6982,
      "mean_token_accuracy": 0.7840036064386368,
      "num_tokens": 10158552.0,
      "step": 8450
    },
    {
      "entropy": 1.8708418890833856,
      "epoch": 0.02622525979204237,
      "grad_norm": 12.569872856140137,
      "learning_rate": 4.09716167780684e-06,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8157704427838326,
      "num_tokens": 10170966.0,
      "step": 8460
    },
    {
      "entropy": 1.9303714036941528,
      "epoch": 0.026256258917092067,
      "grad_norm": 14.426962852478027,
      "learning_rate": 4.1020052310374896e-06,
      "loss": 0.6081,
      "mean_token_accuracy": 0.8103082060813904,
      "num_tokens": 10183633.0,
      "step": 8470
    },
    {
      "entropy": 1.862780438363552,
      "epoch": 0.02628725804214176,
      "grad_norm": 5.966731548309326,
      "learning_rate": 4.1068487842681394e-06,
      "loss": 0.593,
      "mean_token_accuracy": 0.804168826341629,
      "num_tokens": 10195713.0,
      "step": 8480
    },
    {
      "entropy": 1.9560957163572312,
      "epoch": 0.026318257167191456,
      "grad_norm": 13.00171947479248,
      "learning_rate": 4.111692337498789e-06,
      "loss": 0.6648,
      "mean_token_accuracy": 0.799228847026825,
      "num_tokens": 10207536.0,
      "step": 8490
    },
    {
      "entropy": 1.889628753066063,
      "epoch": 0.026349256292241153,
      "grad_norm": 6.381255626678467,
      "learning_rate": 4.116535890729439e-06,
      "loss": 0.5822,
      "mean_token_accuracy": 0.8094806551933289,
      "num_tokens": 10220149.0,
      "step": 8500
    },
    {
      "entropy": 1.9717933177947997,
      "epoch": 0.026380255417290846,
      "grad_norm": 12.86796760559082,
      "learning_rate": 4.12137944396009e-06,
      "loss": 0.6727,
      "mean_token_accuracy": 0.8050152316689492,
      "num_tokens": 10231949.0,
      "step": 8510
    },
    {
      "entropy": 1.934307949244976,
      "epoch": 0.026411254542340543,
      "grad_norm": 11.483423233032227,
      "learning_rate": 4.12622299719074e-06,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8233723506331444,
      "num_tokens": 10243617.0,
      "step": 8520
    },
    {
      "entropy": 1.9711205691099167,
      "epoch": 0.02644225366739024,
      "grad_norm": 11.086894035339355,
      "learning_rate": 4.131066550421389e-06,
      "loss": 0.67,
      "mean_token_accuracy": 0.7918197363615036,
      "num_tokens": 10254929.0,
      "step": 8530
    },
    {
      "entropy": 1.8315666139125824,
      "epoch": 0.026473252792439932,
      "grad_norm": 12.072243690490723,
      "learning_rate": 4.13591010365204e-06,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8164573073387146,
      "num_tokens": 10268348.0,
      "step": 8540
    },
    {
      "entropy": 1.8700304999947548,
      "epoch": 0.02650425191748963,
      "grad_norm": 7.212819576263428,
      "learning_rate": 4.1407536568826895e-06,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8133175373077393,
      "num_tokens": 10281420.0,
      "step": 8550
    },
    {
      "entropy": 1.8715410739183427,
      "epoch": 0.026535251042539326,
      "grad_norm": 15.102907180786133,
      "learning_rate": 4.145597210113339e-06,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8137237802147865,
      "num_tokens": 10293312.0,
      "step": 8560
    },
    {
      "entropy": 1.8536837711930274,
      "epoch": 0.02656625016758902,
      "grad_norm": 18.23057746887207,
      "learning_rate": 4.15044076334399e-06,
      "loss": 0.5679,
      "mean_token_accuracy": 0.819478040933609,
      "num_tokens": 10305555.0,
      "step": 8570
    },
    {
      "entropy": 1.8068823255598545,
      "epoch": 0.026597249292638715,
      "grad_norm": 6.016721725463867,
      "learning_rate": 4.155284316574639e-06,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8214111477136612,
      "num_tokens": 10318500.0,
      "step": 8580
    },
    {
      "entropy": 1.927972738444805,
      "epoch": 0.026628248417688412,
      "grad_norm": 13.958089828491211,
      "learning_rate": 4.160127869805289e-06,
      "loss": 0.6243,
      "mean_token_accuracy": 0.8016948789358139,
      "num_tokens": 10330081.0,
      "step": 8590
    },
    {
      "entropy": 1.8768570333719254,
      "epoch": 0.026659247542738105,
      "grad_norm": 11.411067962646484,
      "learning_rate": 4.16497142303594e-06,
      "loss": 0.5993,
      "mean_token_accuracy": 0.8204882755875588,
      "num_tokens": 10341630.0,
      "step": 8600
    },
    {
      "entropy": 1.9243871226906777,
      "epoch": 0.0266902466677878,
      "grad_norm": 12.71778678894043,
      "learning_rate": 4.16981497626659e-06,
      "loss": 0.6498,
      "mean_token_accuracy": 0.8025202408432961,
      "num_tokens": 10353521.0,
      "step": 8610
    },
    {
      "entropy": 1.8957347482442857,
      "epoch": 0.026721245792837498,
      "grad_norm": 13.246805191040039,
      "learning_rate": 4.1746585294972395e-06,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8230790719389915,
      "num_tokens": 10365655.0,
      "step": 8620
    },
    {
      "entropy": 1.9469288036227226,
      "epoch": 0.02675224491788719,
      "grad_norm": 13.388124465942383,
      "learning_rate": 4.179502082727889e-06,
      "loss": 0.6473,
      "mean_token_accuracy": 0.8096310868859291,
      "num_tokens": 10377607.0,
      "step": 8630
    },
    {
      "entropy": 1.8346548154950142,
      "epoch": 0.026783244042936888,
      "grad_norm": 13.726068496704102,
      "learning_rate": 4.184345635958539e-06,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8302373275160789,
      "num_tokens": 10390062.0,
      "step": 8640
    },
    {
      "entropy": 1.892690037190914,
      "epoch": 0.026814243167986584,
      "grad_norm": 15.530143737792969,
      "learning_rate": 4.189189189189189e-06,
      "loss": 0.6238,
      "mean_token_accuracy": 0.8069882616400719,
      "num_tokens": 10402361.0,
      "step": 8650
    },
    {
      "entropy": 1.8609767884016037,
      "epoch": 0.026845242293036278,
      "grad_norm": 6.528801918029785,
      "learning_rate": 4.19403274241984e-06,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8153262332081794,
      "num_tokens": 10415461.0,
      "step": 8660
    },
    {
      "entropy": 1.869553703069687,
      "epoch": 0.026876241418085974,
      "grad_norm": 12.446064949035645,
      "learning_rate": 4.19887629565049e-06,
      "loss": 0.5657,
      "mean_token_accuracy": 0.8269910797476768,
      "num_tokens": 10427235.0,
      "step": 8670
    },
    {
      "entropy": 1.8075580313801765,
      "epoch": 0.02690724054313567,
      "grad_norm": 5.994472980499268,
      "learning_rate": 4.20371984888114e-06,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8210108175873756,
      "num_tokens": 10440500.0,
      "step": 8680
    },
    {
      "entropy": 1.857536444067955,
      "epoch": 0.026938239668185364,
      "grad_norm": 7.034928321838379,
      "learning_rate": 4.2085634021117895e-06,
      "loss": 0.6801,
      "mean_token_accuracy": 0.8013882115483284,
      "num_tokens": 10452545.0,
      "step": 8690
    },
    {
      "entropy": 1.8659397497773171,
      "epoch": 0.02696923879323506,
      "grad_norm": 12.33353042602539,
      "learning_rate": 4.2134069553424394e-06,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8184954881668091,
      "num_tokens": 10464085.0,
      "step": 8700
    },
    {
      "entropy": 1.9467614516615868,
      "epoch": 0.027000237918284757,
      "grad_norm": 14.808907508850098,
      "learning_rate": 4.218250508573089e-06,
      "loss": 0.6854,
      "mean_token_accuracy": 0.8033217743039132,
      "num_tokens": 10475471.0,
      "step": 8710
    },
    {
      "entropy": 1.9029730170965196,
      "epoch": 0.02703123704333445,
      "grad_norm": 6.386728763580322,
      "learning_rate": 4.22309406180374e-06,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8169241547584534,
      "num_tokens": 10487494.0,
      "step": 8720
    },
    {
      "entropy": 1.8430425137281419,
      "epoch": 0.027062236168384147,
      "grad_norm": 5.942121982574463,
      "learning_rate": 4.22793761503439e-06,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8265012905001641,
      "num_tokens": 10499362.0,
      "step": 8730
    },
    {
      "entropy": 1.8841292724013328,
      "epoch": 0.027093235293433843,
      "grad_norm": 5.7246198654174805,
      "learning_rate": 4.232781168265039e-06,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8051643744111061,
      "num_tokens": 10510802.0,
      "step": 8740
    },
    {
      "entropy": 1.8404579356312751,
      "epoch": 0.02712423441848354,
      "grad_norm": 14.219062805175781,
      "learning_rate": 4.23762472149569e-06,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8213804766535759,
      "num_tokens": 10523317.0,
      "step": 8750
    },
    {
      "entropy": 1.8721755370497704,
      "epoch": 0.027155233543533233,
      "grad_norm": 16.284940719604492,
      "learning_rate": 4.24246827472634e-06,
      "loss": 0.6095,
      "mean_token_accuracy": 0.807612107694149,
      "num_tokens": 10534703.0,
      "step": 8760
    },
    {
      "entropy": 1.8348284885287285,
      "epoch": 0.02718623266858293,
      "grad_norm": 11.95847225189209,
      "learning_rate": 4.2473118279569895e-06,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8146825641393661,
      "num_tokens": 10547080.0,
      "step": 8770
    },
    {
      "entropy": 1.8284624338150024,
      "epoch": 0.027217231793632626,
      "grad_norm": 14.658815383911133,
      "learning_rate": 4.25215538118764e-06,
      "loss": 0.6493,
      "mean_token_accuracy": 0.8004456847906113,
      "num_tokens": 10559321.0,
      "step": 8780
    },
    {
      "entropy": 1.7974061518907547,
      "epoch": 0.02724823091868232,
      "grad_norm": 8.202540397644043,
      "learning_rate": 4.256998934418289e-06,
      "loss": 0.6258,
      "mean_token_accuracy": 0.8043118000030518,
      "num_tokens": 10572053.0,
      "step": 8790
    },
    {
      "entropy": 1.8207244381308556,
      "epoch": 0.027279230043732016,
      "grad_norm": 14.289605140686035,
      "learning_rate": 4.261842487648939e-06,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8130275592207908,
      "num_tokens": 10584178.0,
      "step": 8800
    },
    {
      "entropy": 1.7965356424450873,
      "epoch": 0.027310229168781713,
      "grad_norm": 5.546839714050293,
      "learning_rate": 4.26668604087959e-06,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8202337950468064,
      "num_tokens": 10597231.0,
      "step": 8810
    },
    {
      "entropy": 1.849162083864212,
      "epoch": 0.027341228293831406,
      "grad_norm": 13.331665992736816,
      "learning_rate": 4.27152959411024e-06,
      "loss": 0.5852,
      "mean_token_accuracy": 0.8217923492193222,
      "num_tokens": 10609555.0,
      "step": 8820
    },
    {
      "entropy": 1.8186723545193673,
      "epoch": 0.027372227418881102,
      "grad_norm": 13.591072082519531,
      "learning_rate": 4.27637314734089e-06,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8041837394237519,
      "num_tokens": 10621885.0,
      "step": 8830
    },
    {
      "entropy": 1.8741968870162964,
      "epoch": 0.0274032265439308,
      "grad_norm": 12.443244934082031,
      "learning_rate": 4.2812167005715395e-06,
      "loss": 0.598,
      "mean_token_accuracy": 0.807262459397316,
      "num_tokens": 10633736.0,
      "step": 8840
    },
    {
      "entropy": 1.9037314355373383,
      "epoch": 0.027434225668980492,
      "grad_norm": 13.946099281311035,
      "learning_rate": 4.286060253802189e-06,
      "loss": 0.6851,
      "mean_token_accuracy": 0.808636249601841,
      "num_tokens": 10645193.0,
      "step": 8850
    },
    {
      "entropy": 1.8897378742694855,
      "epoch": 0.02746522479403019,
      "grad_norm": 13.437854766845703,
      "learning_rate": 4.290903807032839e-06,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8217552006244659,
      "num_tokens": 10656076.0,
      "step": 8860
    },
    {
      "entropy": 1.9088585555553437,
      "epoch": 0.027496223919079885,
      "grad_norm": 15.542739868164062,
      "learning_rate": 4.29574736026349e-06,
      "loss": 0.6387,
      "mean_token_accuracy": 0.8007620498538017,
      "num_tokens": 10666538.0,
      "step": 8870
    },
    {
      "entropy": 1.8244119063019753,
      "epoch": 0.02752722304412958,
      "grad_norm": 13.687409400939941,
      "learning_rate": 4.30059091349414e-06,
      "loss": 0.5834,
      "mean_token_accuracy": 0.8106234610080719,
      "num_tokens": 10679216.0,
      "step": 8880
    },
    {
      "entropy": 1.8295306876301765,
      "epoch": 0.027558222169179275,
      "grad_norm": 13.513102531433105,
      "learning_rate": 4.30543446672479e-06,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8163102254271507,
      "num_tokens": 10690908.0,
      "step": 8890
    },
    {
      "entropy": 1.8037668392062187,
      "epoch": 0.02758922129422897,
      "grad_norm": 10.39285659790039,
      "learning_rate": 4.31027801995544e-06,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8148684576153755,
      "num_tokens": 10702625.0,
      "step": 8900
    },
    {
      "entropy": 1.826628988981247,
      "epoch": 0.027620220419278665,
      "grad_norm": 10.972448348999023,
      "learning_rate": 4.3151215731860895e-06,
      "loss": 0.6049,
      "mean_token_accuracy": 0.8112099289894104,
      "num_tokens": 10715241.0,
      "step": 8910
    },
    {
      "entropy": 1.9132762208580971,
      "epoch": 0.02765121954432836,
      "grad_norm": 11.588542938232422,
      "learning_rate": 4.319965126416739e-06,
      "loss": 0.6643,
      "mean_token_accuracy": 0.800639072060585,
      "num_tokens": 10726279.0,
      "step": 8920
    },
    {
      "entropy": 1.8298836424946785,
      "epoch": 0.027682218669378058,
      "grad_norm": 11.887832641601562,
      "learning_rate": 4.32480867964739e-06,
      "loss": 0.5828,
      "mean_token_accuracy": 0.81552804261446,
      "num_tokens": 10738806.0,
      "step": 8930
    },
    {
      "entropy": 1.8580557018518449,
      "epoch": 0.02771321779442775,
      "grad_norm": 10.222613334655762,
      "learning_rate": 4.32965223287804e-06,
      "loss": 0.6352,
      "mean_token_accuracy": 0.8103110626339912,
      "num_tokens": 10750764.0,
      "step": 8940
    },
    {
      "entropy": 1.7699718743562698,
      "epoch": 0.027744216919477448,
      "grad_norm": 5.6745147705078125,
      "learning_rate": 4.334495786108689e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8255440399050713,
      "num_tokens": 10764128.0,
      "step": 8950
    },
    {
      "entropy": 1.8009968280792237,
      "epoch": 0.027775216044527144,
      "grad_norm": 11.968152046203613,
      "learning_rate": 4.33933933933934e-06,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8253274574875832,
      "num_tokens": 10776273.0,
      "step": 8960
    },
    {
      "entropy": 1.8581290900707246,
      "epoch": 0.027806215169576837,
      "grad_norm": 11.328075408935547,
      "learning_rate": 4.34418289256999e-06,
      "loss": 0.6036,
      "mean_token_accuracy": 0.7987002015113831,
      "num_tokens": 10787895.0,
      "step": 8970
    },
    {
      "entropy": 1.835921722650528,
      "epoch": 0.027837214294626534,
      "grad_norm": 15.295876502990723,
      "learning_rate": 4.3490264458006396e-06,
      "loss": 0.5987,
      "mean_token_accuracy": 0.8129387706518173,
      "num_tokens": 10800249.0,
      "step": 8980
    },
    {
      "entropy": 1.8953902840614318,
      "epoch": 0.02786821341967623,
      "grad_norm": 11.311132431030273,
      "learning_rate": 4.35386999903129e-06,
      "loss": 0.6602,
      "mean_token_accuracy": 0.7996082842350006,
      "num_tokens": 10811728.0,
      "step": 8990
    },
    {
      "entropy": 1.869747567176819,
      "epoch": 0.027899212544725924,
      "grad_norm": 13.892035484313965,
      "learning_rate": 4.358713552261939e-06,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8065097257494926,
      "num_tokens": 10822949.0,
      "step": 9000
    },
    {
      "entropy": 1.8128714829683303,
      "epoch": 0.02793021166977562,
      "grad_norm": 10.744576454162598,
      "learning_rate": 4.363557105492589e-06,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8185910269618034,
      "num_tokens": 10835661.0,
      "step": 9010
    },
    {
      "entropy": 1.8096760362386703,
      "epoch": 0.027961210794825317,
      "grad_norm": 5.07841682434082,
      "learning_rate": 4.36840065872324e-06,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8252274811267852,
      "num_tokens": 10848267.0,
      "step": 9020
    },
    {
      "entropy": 1.9240417778491974,
      "epoch": 0.02799220991987501,
      "grad_norm": 13.373735427856445,
      "learning_rate": 4.37324421195389e-06,
      "loss": 0.6678,
      "mean_token_accuracy": 0.7926280677318573,
      "num_tokens": 10859705.0,
      "step": 9030
    },
    {
      "entropy": 1.9146719083189965,
      "epoch": 0.028023209044924707,
      "grad_norm": 11.327962875366211,
      "learning_rate": 4.37808776518454e-06,
      "loss": 0.6178,
      "mean_token_accuracy": 0.826309834420681,
      "num_tokens": 10871134.0,
      "step": 9040
    },
    {
      "entropy": 1.8291181206703186,
      "epoch": 0.028054208169974403,
      "grad_norm": 14.080740928649902,
      "learning_rate": 4.38293131841519e-06,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8222734659910202,
      "num_tokens": 10883664.0,
      "step": 9050
    },
    {
      "entropy": 1.931309811770916,
      "epoch": 0.028085207295024096,
      "grad_norm": 14.17933177947998,
      "learning_rate": 4.3877748716458395e-06,
      "loss": 0.671,
      "mean_token_accuracy": 0.7961855351924896,
      "num_tokens": 10894837.0,
      "step": 9060
    },
    {
      "entropy": 1.9364981949329376,
      "epoch": 0.028116206420073793,
      "grad_norm": 16.13866424560547,
      "learning_rate": 4.392618424876489e-06,
      "loss": 0.6832,
      "mean_token_accuracy": 0.8080704003572464,
      "num_tokens": 10905463.0,
      "step": 9070
    },
    {
      "entropy": 1.8846597149968147,
      "epoch": 0.02814720554512349,
      "grad_norm": 12.987224578857422,
      "learning_rate": 4.39746197810714e-06,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8108007118105889,
      "num_tokens": 10917119.0,
      "step": 9080
    },
    {
      "entropy": 1.9063777923583984,
      "epoch": 0.028178204670173183,
      "grad_norm": 14.922365188598633,
      "learning_rate": 4.40230553133779e-06,
      "loss": 0.6725,
      "mean_token_accuracy": 0.8062027513980865,
      "num_tokens": 10927669.0,
      "step": 9090
    },
    {
      "entropy": 1.8114924758672715,
      "epoch": 0.02820920379522288,
      "grad_norm": 10.971781730651855,
      "learning_rate": 4.40714908456844e-06,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8184123501181603,
      "num_tokens": 10939764.0,
      "step": 9100
    },
    {
      "entropy": 1.9051185458898545,
      "epoch": 0.028240202920272576,
      "grad_norm": 5.854284763336182,
      "learning_rate": 4.41199263779909e-06,
      "loss": 0.6057,
      "mean_token_accuracy": 0.8078436881303788,
      "num_tokens": 10951263.0,
      "step": 9110
    },
    {
      "entropy": 2.0060752868652343,
      "epoch": 0.028271202045322272,
      "grad_norm": 13.479512214660645,
      "learning_rate": 4.41683619102974e-06,
      "loss": 0.7195,
      "mean_token_accuracy": 0.7970252841711044,
      "num_tokens": 10962301.0,
      "step": 9120
    },
    {
      "entropy": 1.9399765685200692,
      "epoch": 0.028302201170371966,
      "grad_norm": 13.337047576904297,
      "learning_rate": 4.4216797442603895e-06,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8143787398934365,
      "num_tokens": 10973469.0,
      "step": 9130
    },
    {
      "entropy": 1.8391277641057968,
      "epoch": 0.028333200295421662,
      "grad_norm": 6.049315452575684,
      "learning_rate": 4.42652329749104e-06,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8160114631056785,
      "num_tokens": 10986193.0,
      "step": 9140
    },
    {
      "entropy": 1.8236407771706582,
      "epoch": 0.02836419942047136,
      "grad_norm": 12.77992057800293,
      "learning_rate": 4.43136685072169e-06,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8185855254530907,
      "num_tokens": 10998743.0,
      "step": 9150
    },
    {
      "entropy": 1.9133103117346764,
      "epoch": 0.028395198545521052,
      "grad_norm": 11.950321197509766,
      "learning_rate": 4.436210403952339e-06,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8190466240048409,
      "num_tokens": 11010167.0,
      "step": 9160
    },
    {
      "entropy": 1.8094365164637565,
      "epoch": 0.02842619767057075,
      "grad_norm": 15.926004409790039,
      "learning_rate": 4.44105395718299e-06,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8278137043118476,
      "num_tokens": 11023270.0,
      "step": 9170
    },
    {
      "entropy": 1.9534984081983566,
      "epoch": 0.028457196795620445,
      "grad_norm": 13.150014877319336,
      "learning_rate": 4.44589751041364e-06,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8099380254745483,
      "num_tokens": 11034084.0,
      "step": 9180
    },
    {
      "entropy": 1.9308686077594757,
      "epoch": 0.028488195920670138,
      "grad_norm": 13.831839561462402,
      "learning_rate": 4.45074106364429e-06,
      "loss": 0.6109,
      "mean_token_accuracy": 0.8114223212003708,
      "num_tokens": 11045463.0,
      "step": 9190
    },
    {
      "entropy": 1.922127565741539,
      "epoch": 0.028519195045719835,
      "grad_norm": 14.833643913269043,
      "learning_rate": 4.45558461687494e-06,
      "loss": 0.6848,
      "mean_token_accuracy": 0.8016584351658821,
      "num_tokens": 11057209.0,
      "step": 9200
    },
    {
      "entropy": 1.9555729806423188,
      "epoch": 0.02855019417076953,
      "grad_norm": 12.715826988220215,
      "learning_rate": 4.4604281701055894e-06,
      "loss": 0.6802,
      "mean_token_accuracy": 0.805061075091362,
      "num_tokens": 11068291.0,
      "step": 9210
    },
    {
      "entropy": 1.942775259912014,
      "epoch": 0.028581193295819225,
      "grad_norm": 7.836392879486084,
      "learning_rate": 4.465271723336239e-06,
      "loss": 0.6386,
      "mean_token_accuracy": 0.8060253381729126,
      "num_tokens": 11079075.0,
      "step": 9220
    },
    {
      "entropy": 1.9078367695212364,
      "epoch": 0.02861219242086892,
      "grad_norm": 13.057887077331543,
      "learning_rate": 4.47011527656689e-06,
      "loss": 0.6303,
      "mean_token_accuracy": 0.806829422712326,
      "num_tokens": 11090899.0,
      "step": 9230
    },
    {
      "entropy": 1.9537988051772117,
      "epoch": 0.028643191545918618,
      "grad_norm": 13.034477233886719,
      "learning_rate": 4.47495882979754e-06,
      "loss": 0.643,
      "mean_token_accuracy": 0.8042517855763436,
      "num_tokens": 11102296.0,
      "step": 9240
    },
    {
      "entropy": 1.8851880788803101,
      "epoch": 0.02867419067096831,
      "grad_norm": 15.656882286071777,
      "learning_rate": 4.47980238302819e-06,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8147235870361328,
      "num_tokens": 11113748.0,
      "step": 9250
    },
    {
      "entropy": 1.770183938741684,
      "epoch": 0.028705189796018007,
      "grad_norm": 12.287094116210938,
      "learning_rate": 4.48464593625884e-06,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8376102089881897,
      "num_tokens": 11126535.0,
      "step": 9260
    },
    {
      "entropy": 1.8187786787748337,
      "epoch": 0.028736188921067704,
      "grad_norm": 12.566884994506836,
      "learning_rate": 4.48948948948949e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8142774716019631,
      "num_tokens": 11139226.0,
      "step": 9270
    },
    {
      "entropy": 1.9643137067556382,
      "epoch": 0.028767188046117397,
      "grad_norm": 13.199538230895996,
      "learning_rate": 4.4943330427201395e-06,
      "loss": 0.7005,
      "mean_token_accuracy": 0.7958864018321037,
      "num_tokens": 11150412.0,
      "step": 9280
    },
    {
      "entropy": 1.7707602977752686,
      "epoch": 0.028798187171167094,
      "grad_norm": 11.749283790588379,
      "learning_rate": 4.49917659595079e-06,
      "loss": 0.4821,
      "mean_token_accuracy": 0.826721802353859,
      "num_tokens": 11164245.0,
      "step": 9290
    },
    {
      "entropy": 1.8696741789579392,
      "epoch": 0.02882918629621679,
      "grad_norm": 13.251580238342285,
      "learning_rate": 4.50402014918144e-06,
      "loss": 0.5609,
      "mean_token_accuracy": 0.819263182580471,
      "num_tokens": 11176738.0,
      "step": 9300
    },
    {
      "entropy": 1.9785585284233094,
      "epoch": 0.028860185421266484,
      "grad_norm": 14.014039039611816,
      "learning_rate": 4.50886370241209e-06,
      "loss": 0.629,
      "mean_token_accuracy": 0.8204264283180237,
      "num_tokens": 11187416.0,
      "step": 9310
    },
    {
      "entropy": 1.9258405417203903,
      "epoch": 0.02889118454631618,
      "grad_norm": 12.051921844482422,
      "learning_rate": 4.51370725564274e-06,
      "loss": 0.6911,
      "mean_token_accuracy": 0.8011237904429436,
      "num_tokens": 11198731.0,
      "step": 9320
    },
    {
      "entropy": 1.7640791162848473,
      "epoch": 0.028922183671365877,
      "grad_norm": 13.392481803894043,
      "learning_rate": 4.51855080887339e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8237902507185936,
      "num_tokens": 11212726.0,
      "step": 9330
    },
    {
      "entropy": 1.9153805747628212,
      "epoch": 0.02895318279641557,
      "grad_norm": 12.259576797485352,
      "learning_rate": 4.52339436210404e-06,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8111542999744416,
      "num_tokens": 11225651.0,
      "step": 9340
    },
    {
      "entropy": 2.023898732662201,
      "epoch": 0.028984181921465266,
      "grad_norm": 15.207432746887207,
      "learning_rate": 4.52823791533469e-06,
      "loss": 0.7204,
      "mean_token_accuracy": 0.7969193458557129,
      "num_tokens": 11236751.0,
      "step": 9350
    },
    {
      "entropy": 1.9569285050034524,
      "epoch": 0.029015181046514963,
      "grad_norm": 15.066774368286133,
      "learning_rate": 4.53308146856534e-06,
      "loss": 0.6648,
      "mean_token_accuracy": 0.8044527977705002,
      "num_tokens": 11248128.0,
      "step": 9360
    },
    {
      "entropy": 1.8909387812018394,
      "epoch": 0.029046180171564656,
      "grad_norm": 12.793632507324219,
      "learning_rate": 4.537925021795989e-06,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8120290517807007,
      "num_tokens": 11260048.0,
      "step": 9370
    },
    {
      "entropy": 1.9061389192938805,
      "epoch": 0.029077179296614353,
      "grad_norm": 10.562239646911621,
      "learning_rate": 4.54276857502664e-06,
      "loss": 0.6658,
      "mean_token_accuracy": 0.7987107247114181,
      "num_tokens": 11271960.0,
      "step": 9380
    },
    {
      "entropy": 1.9409636542201043,
      "epoch": 0.02910817842166405,
      "grad_norm": 12.370716094970703,
      "learning_rate": 4.54761212825729e-06,
      "loss": 0.641,
      "mean_token_accuracy": 0.8062451392412185,
      "num_tokens": 11283082.0,
      "step": 9390
    },
    {
      "entropy": 1.9235450312495233,
      "epoch": 0.029139177546713742,
      "grad_norm": 11.286612510681152,
      "learning_rate": 4.55245568148794e-06,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8160214677453042,
      "num_tokens": 11294418.0,
      "step": 9400
    },
    {
      "entropy": 1.9037536427378654,
      "epoch": 0.02917017667176344,
      "grad_norm": 12.946687698364258,
      "learning_rate": 4.5572992347185905e-06,
      "loss": 0.6663,
      "mean_token_accuracy": 0.7993274956941605,
      "num_tokens": 11306264.0,
      "step": 9410
    },
    {
      "entropy": 1.9348094776272773,
      "epoch": 0.029201175796813136,
      "grad_norm": 11.020223617553711,
      "learning_rate": 4.5621427879492395e-06,
      "loss": 0.647,
      "mean_token_accuracy": 0.8057587504386902,
      "num_tokens": 11316953.0,
      "step": 9420
    },
    {
      "entropy": 1.8925198674201966,
      "epoch": 0.02923217492186283,
      "grad_norm": 12.39678955078125,
      "learning_rate": 4.5669863411798894e-06,
      "loss": 0.6654,
      "mean_token_accuracy": 0.8044493556022644,
      "num_tokens": 11328541.0,
      "step": 9430
    },
    {
      "entropy": 1.871733333170414,
      "epoch": 0.029263174046912525,
      "grad_norm": 4.86548376083374,
      "learning_rate": 4.57182989441054e-06,
      "loss": 0.6326,
      "mean_token_accuracy": 0.806520962715149,
      "num_tokens": 11341221.0,
      "step": 9440
    },
    {
      "entropy": 1.9219319432973863,
      "epoch": 0.029294173171962222,
      "grad_norm": 13.70630168914795,
      "learning_rate": 4.57667344764119e-06,
      "loss": 0.6243,
      "mean_token_accuracy": 0.8117819979786873,
      "num_tokens": 11352400.0,
      "step": 9450
    },
    {
      "entropy": 1.9954150676727296,
      "epoch": 0.029325172297011915,
      "grad_norm": 16.177236557006836,
      "learning_rate": 4.58151700087184e-06,
      "loss": 0.6861,
      "mean_token_accuracy": 0.7988776013255119,
      "num_tokens": 11363461.0,
      "step": 9460
    },
    {
      "entropy": 1.960495764017105,
      "epoch": 0.02935617142206161,
      "grad_norm": 13.248772621154785,
      "learning_rate": 4.58636055410249e-06,
      "loss": 0.6945,
      "mean_token_accuracy": 0.799818865954876,
      "num_tokens": 11375111.0,
      "step": 9470
    },
    {
      "entropy": 1.9463176384568215,
      "epoch": 0.02938717054711131,
      "grad_norm": 13.98116683959961,
      "learning_rate": 4.59120410733314e-06,
      "loss": 0.6175,
      "mean_token_accuracy": 0.814936289191246,
      "num_tokens": 11386133.0,
      "step": 9480
    },
    {
      "entropy": 1.8891835004091262,
      "epoch": 0.029418169672161005,
      "grad_norm": 3.9220001697540283,
      "learning_rate": 4.59604766056379e-06,
      "loss": 0.58,
      "mean_token_accuracy": 0.8183272242546081,
      "num_tokens": 11399042.0,
      "step": 9490
    },
    {
      "entropy": 1.9922261223196984,
      "epoch": 0.029449168797210698,
      "grad_norm": 10.860396385192871,
      "learning_rate": 4.60089121379444e-06,
      "loss": 0.6858,
      "mean_token_accuracy": 0.8004097312688827,
      "num_tokens": 11410000.0,
      "step": 9500
    },
    {
      "entropy": 1.9224519044160844,
      "epoch": 0.029480167922260395,
      "grad_norm": 11.63853645324707,
      "learning_rate": 4.60573476702509e-06,
      "loss": 0.5717,
      "mean_token_accuracy": 0.816666342318058,
      "num_tokens": 11422038.0,
      "step": 9510
    },
    {
      "entropy": 1.8657364815473556,
      "epoch": 0.02951116704731009,
      "grad_norm": 13.743812561035156,
      "learning_rate": 4.61057832025574e-06,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8122471138834954,
      "num_tokens": 11435434.0,
      "step": 9520
    },
    {
      "entropy": 1.925044772028923,
      "epoch": 0.029542166172359784,
      "grad_norm": 7.6943230628967285,
      "learning_rate": 4.61542187348639e-06,
      "loss": 0.6255,
      "mean_token_accuracy": 0.8038001671433449,
      "num_tokens": 11446785.0,
      "step": 9530
    },
    {
      "entropy": 1.9045713931322097,
      "epoch": 0.02957316529740948,
      "grad_norm": 14.207799911499023,
      "learning_rate": 4.62026542671704e-06,
      "loss": 0.6089,
      "mean_token_accuracy": 0.8201724767684937,
      "num_tokens": 11458736.0,
      "step": 9540
    },
    {
      "entropy": 1.9084357440471649,
      "epoch": 0.029604164422459178,
      "grad_norm": 13.012181282043457,
      "learning_rate": 4.62510897994769e-06,
      "loss": 0.6914,
      "mean_token_accuracy": 0.8042386144399643,
      "num_tokens": 11470670.0,
      "step": 9550
    },
    {
      "entropy": 1.8864588722586633,
      "epoch": 0.02963516354750887,
      "grad_norm": 11.62660026550293,
      "learning_rate": 4.6299525331783405e-06,
      "loss": 0.6257,
      "mean_token_accuracy": 0.8103240177035331,
      "num_tokens": 11482128.0,
      "step": 9560
    },
    {
      "entropy": 1.8707260951399802,
      "epoch": 0.029666162672558567,
      "grad_norm": 12.582703590393066,
      "learning_rate": 4.63479608640899e-06,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8164230138063431,
      "num_tokens": 11494873.0,
      "step": 9570
    },
    {
      "entropy": 1.927544179558754,
      "epoch": 0.029697161797608264,
      "grad_norm": 14.461875915527344,
      "learning_rate": 4.63963963963964e-06,
      "loss": 0.624,
      "mean_token_accuracy": 0.8096611618995666,
      "num_tokens": 11506364.0,
      "step": 9580
    },
    {
      "entropy": 1.8048232197761536,
      "epoch": 0.029728160922657957,
      "grad_norm": 7.559422016143799,
      "learning_rate": 4.64448319287029e-06,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8254976123571396,
      "num_tokens": 11519319.0,
      "step": 9590
    },
    {
      "entropy": 1.868215946853161,
      "epoch": 0.029759160047707654,
      "grad_norm": 7.954585075378418,
      "learning_rate": 4.64932674610094e-06,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8161717414855957,
      "num_tokens": 11531172.0,
      "step": 9600
    },
    {
      "entropy": 1.9226632490754128,
      "epoch": 0.02979015917275735,
      "grad_norm": 12.952104568481445,
      "learning_rate": 4.65417029933159e-06,
      "loss": 0.6432,
      "mean_token_accuracy": 0.7985331505537033,
      "num_tokens": 11542291.0,
      "step": 9610
    },
    {
      "entropy": 1.763420394062996,
      "epoch": 0.029821158297807043,
      "grad_norm": 4.180752277374268,
      "learning_rate": 4.659013852562241e-06,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8286846920847892,
      "num_tokens": 11555662.0,
      "step": 9620
    },
    {
      "entropy": 1.9089765697717667,
      "epoch": 0.02985215742285674,
      "grad_norm": 5.77734899520874,
      "learning_rate": 4.66385740579289e-06,
      "loss": 0.6166,
      "mean_token_accuracy": 0.8161043807864189,
      "num_tokens": 11566884.0,
      "step": 9630
    },
    {
      "entropy": 1.7613512337207795,
      "epoch": 0.029883156547906436,
      "grad_norm": 11.983185768127441,
      "learning_rate": 4.6687009590235395e-06,
      "loss": 0.535,
      "mean_token_accuracy": 0.8203961864113808,
      "num_tokens": 11580772.0,
      "step": 9640
    },
    {
      "entropy": 1.8508950725197792,
      "epoch": 0.02991415567295613,
      "grad_norm": 4.04287576675415,
      "learning_rate": 4.67354451225419e-06,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8185210764408112,
      "num_tokens": 11593156.0,
      "step": 9650
    },
    {
      "entropy": 1.9007043689489365,
      "epoch": 0.029945154798005826,
      "grad_norm": 12.041136741638184,
      "learning_rate": 4.67838806548484e-06,
      "loss": 0.687,
      "mean_token_accuracy": 0.7983652919530868,
      "num_tokens": 11604534.0,
      "step": 9660
    },
    {
      "entropy": 1.8662898167967796,
      "epoch": 0.029976153923055523,
      "grad_norm": 5.588154315948486,
      "learning_rate": 4.68323161871549e-06,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8151004850864411,
      "num_tokens": 11616960.0,
      "step": 9670
    },
    {
      "entropy": 1.8811151057481765,
      "epoch": 0.030007153048105216,
      "grad_norm": 15.325428009033203,
      "learning_rate": 4.68807517194614e-06,
      "loss": 0.6616,
      "mean_token_accuracy": 0.7986567124724389,
      "num_tokens": 11628059.0,
      "step": 9680
    },
    {
      "entropy": 1.8047630965709687,
      "epoch": 0.030038152173154913,
      "grad_norm": 5.080140590667725,
      "learning_rate": 4.69291872517679e-06,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8145760789513588,
      "num_tokens": 11640534.0,
      "step": 9690
    },
    {
      "entropy": 1.9280565515160561,
      "epoch": 0.03006915129820461,
      "grad_norm": 6.16276216506958,
      "learning_rate": 4.69776227840744e-06,
      "loss": 0.6692,
      "mean_token_accuracy": 0.8054024249315261,
      "num_tokens": 11651985.0,
      "step": 9700
    },
    {
      "entropy": 1.9135492697358132,
      "epoch": 0.030100150423254302,
      "grad_norm": 13.727761268615723,
      "learning_rate": 4.70260583163809e-06,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8213163688778877,
      "num_tokens": 11664292.0,
      "step": 9710
    },
    {
      "entropy": 1.8022937417030334,
      "epoch": 0.030131149548304,
      "grad_norm": 11.88475513458252,
      "learning_rate": 4.70744938486874e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8253460243344307,
      "num_tokens": 11677213.0,
      "step": 9720
    },
    {
      "entropy": 1.9173430383205414,
      "epoch": 0.030162148673353695,
      "grad_norm": 10.792257308959961,
      "learning_rate": 4.71229293809939e-06,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8223668470978737,
      "num_tokens": 11689083.0,
      "step": 9730
    },
    {
      "entropy": 1.8793220937252044,
      "epoch": 0.03019314779840339,
      "grad_norm": 13.49764347076416,
      "learning_rate": 4.71713649133004e-06,
      "loss": 0.6808,
      "mean_token_accuracy": 0.7959660813212395,
      "num_tokens": 11701156.0,
      "step": 9740
    },
    {
      "entropy": 1.7766539767384528,
      "epoch": 0.030224146923453085,
      "grad_norm": 12.855195045471191,
      "learning_rate": 4.72198004456069e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.8286942020058632,
      "num_tokens": 11715335.0,
      "step": 9750
    },
    {
      "entropy": 1.914339354634285,
      "epoch": 0.030255146048502782,
      "grad_norm": 12.800863265991211,
      "learning_rate": 4.72682359779134e-06,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8148805812001229,
      "num_tokens": 11727922.0,
      "step": 9760
    },
    {
      "entropy": 1.910454373061657,
      "epoch": 0.030286145173552475,
      "grad_norm": 13.581644058227539,
      "learning_rate": 4.7316671510219906e-06,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8178346559405327,
      "num_tokens": 11739422.0,
      "step": 9770
    },
    {
      "entropy": 1.8079843878746034,
      "epoch": 0.03031714429860217,
      "grad_norm": 14.549524307250977,
      "learning_rate": 4.7365107042526405e-06,
      "loss": 0.543,
      "mean_token_accuracy": 0.8197317391633987,
      "num_tokens": 11752317.0,
      "step": 9780
    },
    {
      "entropy": 1.9487209469079971,
      "epoch": 0.030348143423651868,
      "grad_norm": 12.98496150970459,
      "learning_rate": 4.74135425748329e-06,
      "loss": 0.6664,
      "mean_token_accuracy": 0.8045255482196808,
      "num_tokens": 11763631.0,
      "step": 9790
    },
    {
      "entropy": 1.9213872998952866,
      "epoch": 0.03037914254870156,
      "grad_norm": 14.27520751953125,
      "learning_rate": 4.74619781071394e-06,
      "loss": 0.6433,
      "mean_token_accuracy": 0.8067422285676003,
      "num_tokens": 11775491.0,
      "step": 9800
    },
    {
      "entropy": 1.8178927034139634,
      "epoch": 0.030410141673751258,
      "grad_norm": 14.970010757446289,
      "learning_rate": 4.75104136394459e-06,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8228003144264221,
      "num_tokens": 11788717.0,
      "step": 9810
    },
    {
      "entropy": 1.9698729425668717,
      "epoch": 0.030441140798800954,
      "grad_norm": 12.28023624420166,
      "learning_rate": 4.75588491717524e-06,
      "loss": 0.6749,
      "mean_token_accuracy": 0.8033818736672401,
      "num_tokens": 11799616.0,
      "step": 9820
    },
    {
      "entropy": 1.8569468915462495,
      "epoch": 0.03047213992385065,
      "grad_norm": 12.524866104125977,
      "learning_rate": 4.760728470405891e-06,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8196278423070907,
      "num_tokens": 11812567.0,
      "step": 9830
    },
    {
      "entropy": 1.9362822517752647,
      "epoch": 0.030503139048900344,
      "grad_norm": 12.16551685333252,
      "learning_rate": 4.76557202363654e-06,
      "loss": 0.6641,
      "mean_token_accuracy": 0.802090086042881,
      "num_tokens": 11824072.0,
      "step": 9840
    },
    {
      "entropy": 1.9034365549683572,
      "epoch": 0.03053413817395004,
      "grad_norm": 11.310152053833008,
      "learning_rate": 4.77041557686719e-06,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8146296158432961,
      "num_tokens": 11835828.0,
      "step": 9850
    },
    {
      "entropy": 1.949262234568596,
      "epoch": 0.030565137298999737,
      "grad_norm": 13.651056289672852,
      "learning_rate": 4.77525913009784e-06,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8138030260801316,
      "num_tokens": 11847123.0,
      "step": 9860
    },
    {
      "entropy": 1.8827594295144081,
      "epoch": 0.03059613642404943,
      "grad_norm": 11.080418586730957,
      "learning_rate": 4.78010268332849e-06,
      "loss": 0.597,
      "mean_token_accuracy": 0.8088963508605957,
      "num_tokens": 11859028.0,
      "step": 9870
    },
    {
      "entropy": 1.8874688804149629,
      "epoch": 0.030627135549099127,
      "grad_norm": 12.232693672180176,
      "learning_rate": 4.78494623655914e-06,
      "loss": 0.6113,
      "mean_token_accuracy": 0.814998921751976,
      "num_tokens": 11869876.0,
      "step": 9880
    },
    {
      "entropy": 1.8576917335391046,
      "epoch": 0.030658134674148824,
      "grad_norm": 11.815441131591797,
      "learning_rate": 4.78978978978979e-06,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8156881853938103,
      "num_tokens": 11881144.0,
      "step": 9890
    },
    {
      "entropy": 1.8440957143902779,
      "epoch": 0.030689133799198517,
      "grad_norm": 12.98918628692627,
      "learning_rate": 4.79463334302044e-06,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8034987449645996,
      "num_tokens": 11893147.0,
      "step": 9900
    },
    {
      "entropy": 1.88355852663517,
      "epoch": 0.030720132924248213,
      "grad_norm": 6.641689300537109,
      "learning_rate": 4.79947689625109e-06,
      "loss": 0.6617,
      "mean_token_accuracy": 0.7953455910086632,
      "num_tokens": 11905450.0,
      "step": 9910
    },
    {
      "entropy": 1.9107140555977822,
      "epoch": 0.03075113204929791,
      "grad_norm": 5.667116641998291,
      "learning_rate": 4.8043204494817405e-06,
      "loss": 0.6608,
      "mean_token_accuracy": 0.8010135576128959,
      "num_tokens": 11916859.0,
      "step": 9920
    },
    {
      "entropy": 1.9044810444116593,
      "epoch": 0.030782131174347603,
      "grad_norm": 13.490184783935547,
      "learning_rate": 4.80916400271239e-06,
      "loss": 0.6331,
      "mean_token_accuracy": 0.8088266268372536,
      "num_tokens": 11927564.0,
      "step": 9930
    },
    {
      "entropy": 1.899574799835682,
      "epoch": 0.0308131302993973,
      "grad_norm": 13.318133354187012,
      "learning_rate": 4.81400755594304e-06,
      "loss": 0.6828,
      "mean_token_accuracy": 0.8059917777776718,
      "num_tokens": 11938823.0,
      "step": 9940
    },
    {
      "entropy": 1.8431586802005768,
      "epoch": 0.030844129424446996,
      "grad_norm": 11.968719482421875,
      "learning_rate": 4.81885110917369e-06,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8303678393363952,
      "num_tokens": 11950208.0,
      "step": 9950
    },
    {
      "entropy": 1.8151010930538178,
      "epoch": 0.03087512854949669,
      "grad_norm": 13.34570026397705,
      "learning_rate": 4.82369466240434e-06,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8233583375811577,
      "num_tokens": 11963211.0,
      "step": 9960
    },
    {
      "entropy": 1.9202764973044395,
      "epoch": 0.030906127674546386,
      "grad_norm": 12.33815860748291,
      "learning_rate": 4.82853821563499e-06,
      "loss": 0.6453,
      "mean_token_accuracy": 0.7984903916716576,
      "num_tokens": 11974413.0,
      "step": 9970
    },
    {
      "entropy": 1.7811052426695824,
      "epoch": 0.030937126799596083,
      "grad_norm": 13.612791061401367,
      "learning_rate": 4.833381768865641e-06,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8154791623353959,
      "num_tokens": 11986759.0,
      "step": 9980
    },
    {
      "entropy": 1.8961547821760179,
      "epoch": 0.030968125924645776,
      "grad_norm": 13.582847595214844,
      "learning_rate": 4.8382253220962906e-06,
      "loss": 0.6424,
      "mean_token_accuracy": 0.8010322540998459,
      "num_tokens": 11997882.0,
      "step": 9990
    },
    {
      "entropy": 1.7699449375271796,
      "epoch": 0.030999125049695472,
      "grad_norm": 12.8268461227417,
      "learning_rate": 4.8430688753269404e-06,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8104767709970474,
      "num_tokens": 12011562.0,
      "step": 10000
    },
    {
      "entropy": 1.7704377323389053,
      "epoch": 0.03103012417474517,
      "grad_norm": 7.003698348999023,
      "learning_rate": 4.84791242855759e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.817914342880249,
      "num_tokens": 12024642.0,
      "step": 10010
    },
    {
      "entropy": 1.8417405053973197,
      "epoch": 0.031061123299794862,
      "grad_norm": 12.85013198852539,
      "learning_rate": 4.85275598178824e-06,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8156400442123413,
      "num_tokens": 12035688.0,
      "step": 10020
    },
    {
      "entropy": 1.871736840903759,
      "epoch": 0.03109212242484456,
      "grad_norm": 10.676023483276367,
      "learning_rate": 4.85759953501889e-06,
      "loss": 0.6359,
      "mean_token_accuracy": 0.8091729044914245,
      "num_tokens": 12046810.0,
      "step": 10030
    },
    {
      "entropy": 1.8386752873659133,
      "epoch": 0.031123121549894255,
      "grad_norm": 5.60599422454834,
      "learning_rate": 4.862443088249541e-06,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8189519822597504,
      "num_tokens": 12059567.0,
      "step": 10040
    },
    {
      "entropy": 1.9133656844496727,
      "epoch": 0.03115412067494395,
      "grad_norm": 13.941995620727539,
      "learning_rate": 4.86728664148019e-06,
      "loss": 0.6437,
      "mean_token_accuracy": 0.8070616811513901,
      "num_tokens": 12070777.0,
      "step": 10050
    },
    {
      "entropy": 1.859761357307434,
      "epoch": 0.031185119799993645,
      "grad_norm": 11.398521423339844,
      "learning_rate": 4.87213019471084e-06,
      "loss": 0.557,
      "mean_token_accuracy": 0.8148921579122543,
      "num_tokens": 12082916.0,
      "step": 10060
    },
    {
      "entropy": 1.8685668006539344,
      "epoch": 0.03121611892504334,
      "grad_norm": 11.567840576171875,
      "learning_rate": 4.8769737479414905e-06,
      "loss": 0.6081,
      "mean_token_accuracy": 0.812491662800312,
      "num_tokens": 12094481.0,
      "step": 10070
    },
    {
      "entropy": 1.7921167001128198,
      "epoch": 0.031247118050093035,
      "grad_norm": 11.81413745880127,
      "learning_rate": 4.88181730117214e-06,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8256913438439369,
      "num_tokens": 12107579.0,
      "step": 10080
    },
    {
      "entropy": 1.9026924163103103,
      "epoch": 0.031278117175142735,
      "grad_norm": 5.831927299499512,
      "learning_rate": 4.88666085440279e-06,
      "loss": 0.5999,
      "mean_token_accuracy": 0.8175748229026795,
      "num_tokens": 12119504.0,
      "step": 10090
    },
    {
      "entropy": 1.8218435019254684,
      "epoch": 0.031309116300192424,
      "grad_norm": 6.674549579620361,
      "learning_rate": 4.89150440763344e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8210835456848145,
      "num_tokens": 12132997.0,
      "step": 10100
    },
    {
      "entropy": 1.7946079865098,
      "epoch": 0.03134011542524212,
      "grad_norm": 6.814689636230469,
      "learning_rate": 4.89634796086409e-06,
      "loss": 0.5653,
      "mean_token_accuracy": 0.8189253076910973,
      "num_tokens": 12145732.0,
      "step": 10110
    },
    {
      "entropy": 1.8335962176322937,
      "epoch": 0.03137111455029182,
      "grad_norm": 12.774432182312012,
      "learning_rate": 4.90119151409474e-06,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8219740748405456,
      "num_tokens": 12157527.0,
      "step": 10120
    },
    {
      "entropy": 1.9156937181949616,
      "epoch": 0.031402113675341514,
      "grad_norm": 13.017987251281738,
      "learning_rate": 4.906035067325391e-06,
      "loss": 0.6375,
      "mean_token_accuracy": 0.8114208281040192,
      "num_tokens": 12168529.0,
      "step": 10130
    },
    {
      "entropy": 1.929344841837883,
      "epoch": 0.03143311280039121,
      "grad_norm": 12.184325218200684,
      "learning_rate": 4.9108786205560405e-06,
      "loss": 0.6351,
      "mean_token_accuracy": 0.816659900546074,
      "num_tokens": 12179791.0,
      "step": 10140
    },
    {
      "entropy": 1.8231835559010505,
      "epoch": 0.03146411192544091,
      "grad_norm": 13.120546340942383,
      "learning_rate": 4.91572217378669e-06,
      "loss": 0.5967,
      "mean_token_accuracy": 0.8101472899317741,
      "num_tokens": 12192369.0,
      "step": 10150
    },
    {
      "entropy": 1.7879482999444007,
      "epoch": 0.0314951110504906,
      "grad_norm": 14.349749565124512,
      "learning_rate": 4.92056572701734e-06,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8103778302669525,
      "num_tokens": 12205341.0,
      "step": 10160
    },
    {
      "entropy": 1.9244187206029892,
      "epoch": 0.031526110175540294,
      "grad_norm": 11.694113731384277,
      "learning_rate": 4.92540928024799e-06,
      "loss": 0.643,
      "mean_token_accuracy": 0.8135433197021484,
      "num_tokens": 12217222.0,
      "step": 10170
    },
    {
      "entropy": 1.9573982939124108,
      "epoch": 0.03155710930058999,
      "grad_norm": 12.177897453308105,
      "learning_rate": 4.93025283347864e-06,
      "loss": 0.6126,
      "mean_token_accuracy": 0.809203888475895,
      "num_tokens": 12228264.0,
      "step": 10180
    },
    {
      "entropy": 1.9127047300338744,
      "epoch": 0.03158810842563969,
      "grad_norm": 13.4846830368042,
      "learning_rate": 4.935096386709291e-06,
      "loss": 0.6306,
      "mean_token_accuracy": 0.8137389361858368,
      "num_tokens": 12240050.0,
      "step": 10190
    },
    {
      "entropy": 1.910162153840065,
      "epoch": 0.03161910755068938,
      "grad_norm": 14.54655647277832,
      "learning_rate": 4.939939939939941e-06,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8089611247181893,
      "num_tokens": 12251372.0,
      "step": 10200
    },
    {
      "entropy": 1.9011714354157447,
      "epoch": 0.03165010667573908,
      "grad_norm": 12.256708145141602,
      "learning_rate": 4.9447834931705905e-06,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8112087488174439,
      "num_tokens": 12263039.0,
      "step": 10210
    },
    {
      "entropy": 1.8994778007268907,
      "epoch": 0.03168110580078877,
      "grad_norm": 14.00627613067627,
      "learning_rate": 4.9496270464012404e-06,
      "loss": 0.732,
      "mean_token_accuracy": 0.7959692940115929,
      "num_tokens": 12275633.0,
      "step": 10220
    },
    {
      "entropy": 1.7719287261366845,
      "epoch": 0.031712104925838466,
      "grad_norm": 10.754083633422852,
      "learning_rate": 4.95447059963189e-06,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8201853647828102,
      "num_tokens": 12288123.0,
      "step": 10230
    },
    {
      "entropy": 1.9072040289640426,
      "epoch": 0.03174310405088816,
      "grad_norm": 10.166359901428223,
      "learning_rate": 4.95931415286254e-06,
      "loss": 0.6276,
      "mean_token_accuracy": 0.8175885871052742,
      "num_tokens": 12299187.0,
      "step": 10240
    },
    {
      "entropy": 1.897184392809868,
      "epoch": 0.03177410317593786,
      "grad_norm": 10.87751293182373,
      "learning_rate": 4.964157706093191e-06,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8149495676159859,
      "num_tokens": 12309767.0,
      "step": 10250
    },
    {
      "entropy": 1.8759140372276306,
      "epoch": 0.031805102300987556,
      "grad_norm": 10.203163146972656,
      "learning_rate": 4.96900125932384e-06,
      "loss": 0.634,
      "mean_token_accuracy": 0.8065581321716309,
      "num_tokens": 12320842.0,
      "step": 10260
    },
    {
      "entropy": 1.8192242681980133,
      "epoch": 0.03183610142603725,
      "grad_norm": 13.203461647033691,
      "learning_rate": 4.97384481255449e-06,
      "loss": 0.5818,
      "mean_token_accuracy": 0.8157040163874626,
      "num_tokens": 12332641.0,
      "step": 10270
    },
    {
      "entropy": 1.9448095709085464,
      "epoch": 0.03186710055108694,
      "grad_norm": 16.432714462280273,
      "learning_rate": 4.978688365785141e-06,
      "loss": 0.7231,
      "mean_token_accuracy": 0.7967922583222389,
      "num_tokens": 12343283.0,
      "step": 10280
    },
    {
      "entropy": 1.8554027765989303,
      "epoch": 0.03189809967613664,
      "grad_norm": 10.973164558410645,
      "learning_rate": 4.9835319190157905e-06,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8122664794325829,
      "num_tokens": 12355277.0,
      "step": 10290
    },
    {
      "entropy": 1.8099492847919465,
      "epoch": 0.031929098801186336,
      "grad_norm": 13.487873077392578,
      "learning_rate": 4.98837547224644e-06,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8163850530982018,
      "num_tokens": 12367374.0,
      "step": 10300
    },
    {
      "entropy": 1.907564203441143,
      "epoch": 0.03196009792623603,
      "grad_norm": 14.73357105255127,
      "learning_rate": 4.99321902547709e-06,
      "loss": 0.6864,
      "mean_token_accuracy": 0.804863877594471,
      "num_tokens": 12379185.0,
      "step": 10310
    },
    {
      "entropy": 1.9018947511911393,
      "epoch": 0.03199109705128573,
      "grad_norm": 12.761754035949707,
      "learning_rate": 4.99806257870774e-06,
      "loss": 0.6285,
      "mean_token_accuracy": 0.7974288746714592,
      "num_tokens": 12390639.0,
      "step": 10320
    },
    {
      "entropy": 1.8945544630289077,
      "epoch": 0.032022096176335425,
      "grad_norm": 12.64550495147705,
      "learning_rate": 5.002906131938391e-06,
      "loss": 0.696,
      "mean_token_accuracy": 0.7960980877280235,
      "num_tokens": 12401445.0,
      "step": 10330
    },
    {
      "entropy": 1.8769317850470544,
      "epoch": 0.032053095301385115,
      "grad_norm": 10.872472763061523,
      "learning_rate": 5.00774968516904e-06,
      "loss": 0.6736,
      "mean_token_accuracy": 0.8056196987628936,
      "num_tokens": 12412842.0,
      "step": 10340
    },
    {
      "entropy": 1.874910195171833,
      "epoch": 0.03208409442643481,
      "grad_norm": 11.449685096740723,
      "learning_rate": 5.01259323839969e-06,
      "loss": 0.6081,
      "mean_token_accuracy": 0.8147278919816017,
      "num_tokens": 12424094.0,
      "step": 10350
    },
    {
      "entropy": 1.8590925335884094,
      "epoch": 0.03211509355148451,
      "grad_norm": 13.95566177368164,
      "learning_rate": 5.0174367916303405e-06,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8060554280877114,
      "num_tokens": 12436102.0,
      "step": 10360
    },
    {
      "entropy": 1.9232334434986114,
      "epoch": 0.032146092676534205,
      "grad_norm": 13.31452751159668,
      "learning_rate": 5.02228034486099e-06,
      "loss": 0.6143,
      "mean_token_accuracy": 0.819976469874382,
      "num_tokens": 12447424.0,
      "step": 10370
    },
    {
      "entropy": 1.860893575847149,
      "epoch": 0.0321770918015839,
      "grad_norm": 11.070972442626953,
      "learning_rate": 5.02712389809164e-06,
      "loss": 0.6446,
      "mean_token_accuracy": 0.8143087536096573,
      "num_tokens": 12459620.0,
      "step": 10380
    },
    {
      "entropy": 1.8405973672866822,
      "epoch": 0.0322080909266336,
      "grad_norm": 11.35168743133545,
      "learning_rate": 5.03196745132229e-06,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8185856312513351,
      "num_tokens": 12471581.0,
      "step": 10390
    },
    {
      "entropy": 1.8092712104320525,
      "epoch": 0.032239090051683295,
      "grad_norm": 12.700512886047363,
      "learning_rate": 5.036811004552941e-06,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8279015332460403,
      "num_tokens": 12483574.0,
      "step": 10400
    },
    {
      "entropy": 1.772294245660305,
      "epoch": 0.032270089176732984,
      "grad_norm": 11.940170288085938,
      "learning_rate": 5.041654557783591e-06,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8129552945494651,
      "num_tokens": 12496082.0,
      "step": 10410
    },
    {
      "entropy": 1.7716655775904655,
      "epoch": 0.03230108830178268,
      "grad_norm": 5.408534049987793,
      "learning_rate": 5.046498111014241e-06,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8172632664442062,
      "num_tokens": 12508722.0,
      "step": 10420
    },
    {
      "entropy": 1.7848970398306847,
      "epoch": 0.03233208742683238,
      "grad_norm": 10.916991233825684,
      "learning_rate": 5.0513416642448905e-06,
      "loss": 0.5381,
      "mean_token_accuracy": 0.822851151227951,
      "num_tokens": 12521638.0,
      "step": 10430
    },
    {
      "entropy": 1.88555389046669,
      "epoch": 0.032363086551882074,
      "grad_norm": 16.887577056884766,
      "learning_rate": 5.0561852174755396e-06,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7913921490311623,
      "num_tokens": 12533306.0,
      "step": 10440
    },
    {
      "entropy": 1.8448584645986557,
      "epoch": 0.03239408567693177,
      "grad_norm": 5.770571708679199,
      "learning_rate": 5.06102877070619e-06,
      "loss": 0.545,
      "mean_token_accuracy": 0.8273654997348785,
      "num_tokens": 12544900.0,
      "step": 10450
    },
    {
      "entropy": 1.8120036974549294,
      "epoch": 0.03242508480198147,
      "grad_norm": 12.20251178741455,
      "learning_rate": 5.06587232393684e-06,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8255019530653953,
      "num_tokens": 12557961.0,
      "step": 10460
    },
    {
      "entropy": 1.8894508898258209,
      "epoch": 0.03245608392703116,
      "grad_norm": 12.401219367980957,
      "learning_rate": 5.07071587716749e-06,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8066725865006447,
      "num_tokens": 12569409.0,
      "step": 10470
    },
    {
      "entropy": 1.8568513855338096,
      "epoch": 0.03248708305208085,
      "grad_norm": 15.682320594787598,
      "learning_rate": 5.07555943039814e-06,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8160646855831146,
      "num_tokens": 12581146.0,
      "step": 10480
    },
    {
      "entropy": 1.806802648305893,
      "epoch": 0.03251808217713055,
      "grad_norm": 13.31312370300293,
      "learning_rate": 5.080402983628791e-06,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8170752301812172,
      "num_tokens": 12593979.0,
      "step": 10490
    },
    {
      "entropy": 1.8914892196655273,
      "epoch": 0.03254908130218025,
      "grad_norm": 13.41655445098877,
      "learning_rate": 5.0852465368594406e-06,
      "loss": 0.6523,
      "mean_token_accuracy": 0.8074697732925415,
      "num_tokens": 12604872.0,
      "step": 10500
    },
    {
      "entropy": 1.8397974416613578,
      "epoch": 0.03258008042722994,
      "grad_norm": 13.245908737182617,
      "learning_rate": 5.0900900900900905e-06,
      "loss": 0.6411,
      "mean_token_accuracy": 0.8051371321082115,
      "num_tokens": 12616464.0,
      "step": 10510
    },
    {
      "entropy": 1.7762023329734802,
      "epoch": 0.03261107955227964,
      "grad_norm": 12.972115516662598,
      "learning_rate": 5.094933643320741e-06,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8134799718856811,
      "num_tokens": 12628431.0,
      "step": 10520
    },
    {
      "entropy": 1.7372042164206505,
      "epoch": 0.03264207867732933,
      "grad_norm": 14.975220680236816,
      "learning_rate": 5.099777196551391e-06,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8230763703584671,
      "num_tokens": 12641420.0,
      "step": 10530
    },
    {
      "entropy": 1.8806333974003793,
      "epoch": 0.032673077802379026,
      "grad_norm": 10.917582511901855,
      "learning_rate": 5.104620749782041e-06,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8124522119760513,
      "num_tokens": 12652755.0,
      "step": 10540
    },
    {
      "entropy": 1.8468591958284377,
      "epoch": 0.03270407692742872,
      "grad_norm": 15.210384368896484,
      "learning_rate": 5.10946430301269e-06,
      "loss": 0.6178,
      "mean_token_accuracy": 0.8065985247492791,
      "num_tokens": 12664093.0,
      "step": 10550
    },
    {
      "entropy": 1.7937588766217232,
      "epoch": 0.03273507605247842,
      "grad_norm": 11.96823787689209,
      "learning_rate": 5.11430785624334e-06,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8129746183753014,
      "num_tokens": 12676810.0,
      "step": 10560
    },
    {
      "entropy": 1.8590254932641983,
      "epoch": 0.032766075177528116,
      "grad_norm": 13.634770393371582,
      "learning_rate": 5.119151409473991e-06,
      "loss": 0.622,
      "mean_token_accuracy": 0.812176737189293,
      "num_tokens": 12688390.0,
      "step": 10570
    },
    {
      "entropy": 1.8288069173693657,
      "epoch": 0.03279707430257781,
      "grad_norm": 19.997926712036133,
      "learning_rate": 5.1239949627046405e-06,
      "loss": 0.6365,
      "mean_token_accuracy": 0.8077499896287919,
      "num_tokens": 12700682.0,
      "step": 10580
    },
    {
      "entropy": 1.9130954205989839,
      "epoch": 0.0328280734276275,
      "grad_norm": 13.037125587463379,
      "learning_rate": 5.12883851593529e-06,
      "loss": 0.7036,
      "mean_token_accuracy": 0.8007219597697258,
      "num_tokens": 12711160.0,
      "step": 10590
    },
    {
      "entropy": 1.908418272435665,
      "epoch": 0.0328590725526772,
      "grad_norm": 9.9063720703125,
      "learning_rate": 5.13368206916594e-06,
      "loss": 0.6536,
      "mean_token_accuracy": 0.8085066750645638,
      "num_tokens": 12722404.0,
      "step": 10600
    },
    {
      "entropy": 1.88576088398695,
      "epoch": 0.032890071677726895,
      "grad_norm": 15.735233306884766,
      "learning_rate": 5.138525622396591e-06,
      "loss": 0.6415,
      "mean_token_accuracy": 0.8067146822810173,
      "num_tokens": 12735219.0,
      "step": 10610
    },
    {
      "entropy": 1.8483037024736404,
      "epoch": 0.03292107080277659,
      "grad_norm": 16.17683982849121,
      "learning_rate": 5.143369175627241e-06,
      "loss": 0.5999,
      "mean_token_accuracy": 0.809206846356392,
      "num_tokens": 12747804.0,
      "step": 10620
    },
    {
      "entropy": 1.7985512629151343,
      "epoch": 0.03295206992782629,
      "grad_norm": 5.797553062438965,
      "learning_rate": 5.148212728857891e-06,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8201219871640205,
      "num_tokens": 12760529.0,
      "step": 10630
    },
    {
      "entropy": 1.829072842001915,
      "epoch": 0.032983069052875985,
      "grad_norm": 11.334760665893555,
      "learning_rate": 5.153056282088541e-06,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8162604227662087,
      "num_tokens": 12772598.0,
      "step": 10640
    },
    {
      "entropy": 1.8073836967349053,
      "epoch": 0.033014068177925675,
      "grad_norm": 11.55426025390625,
      "learning_rate": 5.15789983531919e-06,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8164139956235885,
      "num_tokens": 12784907.0,
      "step": 10650
    },
    {
      "entropy": 1.8018400803208352,
      "epoch": 0.03304506730297537,
      "grad_norm": 11.91412353515625,
      "learning_rate": 5.16274338854984e-06,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8153271213173866,
      "num_tokens": 12797883.0,
      "step": 10660
    },
    {
      "entropy": 1.7438832193613052,
      "epoch": 0.03307606642802507,
      "grad_norm": 13.903759956359863,
      "learning_rate": 5.16758694178049e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8333475485444068,
      "num_tokens": 12810429.0,
      "step": 10670
    },
    {
      "entropy": 1.8345250859856606,
      "epoch": 0.033107065553074765,
      "grad_norm": 12.650176048278809,
      "learning_rate": 5.17243049501114e-06,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8180657878518105,
      "num_tokens": 12822129.0,
      "step": 10680
    },
    {
      "entropy": 1.8904076486825943,
      "epoch": 0.03313806467812446,
      "grad_norm": 10.247991561889648,
      "learning_rate": 5.17727404824179e-06,
      "loss": 0.6336,
      "mean_token_accuracy": 0.8089143499732018,
      "num_tokens": 12833756.0,
      "step": 10690
    },
    {
      "entropy": 1.8453685097396373,
      "epoch": 0.03316906380317416,
      "grad_norm": 6.841571807861328,
      "learning_rate": 5.182117601472441e-06,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8071049571037292,
      "num_tokens": 12846112.0,
      "step": 10700
    },
    {
      "entropy": 1.8263072147965431,
      "epoch": 0.03320006292822385,
      "grad_norm": 10.83479118347168,
      "learning_rate": 5.186961154703091e-06,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8183331623673439,
      "num_tokens": 12858793.0,
      "step": 10710
    },
    {
      "entropy": 1.9237968116998672,
      "epoch": 0.033231062053273544,
      "grad_norm": 14.300333023071289,
      "learning_rate": 5.1918047079337406e-06,
      "loss": 0.6285,
      "mean_token_accuracy": 0.8162373065948486,
      "num_tokens": 12869520.0,
      "step": 10720
    },
    {
      "entropy": 1.804058338701725,
      "epoch": 0.03326206117832324,
      "grad_norm": 12.697318077087402,
      "learning_rate": 5.196648261164391e-06,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8053817078471184,
      "num_tokens": 12883158.0,
      "step": 10730
    },
    {
      "entropy": 1.858067548274994,
      "epoch": 0.03329306030337294,
      "grad_norm": 12.211063385009766,
      "learning_rate": 5.201491814395041e-06,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8255058348178863,
      "num_tokens": 12894190.0,
      "step": 10740
    },
    {
      "entropy": 1.8079882711172104,
      "epoch": 0.033324059428422634,
      "grad_norm": 6.42866325378418,
      "learning_rate": 5.206335367625691e-06,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8178721934556961,
      "num_tokens": 12907324.0,
      "step": 10750
    },
    {
      "entropy": 1.851043240725994,
      "epoch": 0.03335505855347233,
      "grad_norm": 13.750490188598633,
      "learning_rate": 5.21117892085634e-06,
      "loss": 0.6343,
      "mean_token_accuracy": 0.8060193166136742,
      "num_tokens": 12918683.0,
      "step": 10760
    },
    {
      "entropy": 1.8697371006011962,
      "epoch": 0.03338605767852203,
      "grad_norm": 6.069192409515381,
      "learning_rate": 5.21602247408699e-06,
      "loss": 0.6548,
      "mean_token_accuracy": 0.8067969933152199,
      "num_tokens": 12930503.0,
      "step": 10770
    },
    {
      "entropy": 1.833288662135601,
      "epoch": 0.03341705680357172,
      "grad_norm": 11.35355281829834,
      "learning_rate": 5.220866027317641e-06,
      "loss": 0.575,
      "mean_token_accuracy": 0.8082702040672303,
      "num_tokens": 12943156.0,
      "step": 10780
    },
    {
      "entropy": 1.895593549311161,
      "epoch": 0.03344805592862141,
      "grad_norm": 6.248627662658691,
      "learning_rate": 5.225709580548291e-06,
      "loss": 0.629,
      "mean_token_accuracy": 0.8071183800697327,
      "num_tokens": 12955988.0,
      "step": 10790
    },
    {
      "entropy": 1.8191024243831635,
      "epoch": 0.03347905505367111,
      "grad_norm": 4.646948337554932,
      "learning_rate": 5.2305531337789405e-06,
      "loss": 0.6002,
      "mean_token_accuracy": 0.81056018024683,
      "num_tokens": 12969403.0,
      "step": 10800
    },
    {
      "entropy": 1.8694421991705894,
      "epoch": 0.033510054178720806,
      "grad_norm": 6.7921977043151855,
      "learning_rate": 5.23539668700959e-06,
      "loss": 0.5993,
      "mean_token_accuracy": 0.8087976023554801,
      "num_tokens": 12981414.0,
      "step": 10810
    },
    {
      "entropy": 1.8624958485364913,
      "epoch": 0.0335410533037705,
      "grad_norm": 14.213687896728516,
      "learning_rate": 5.240240240240241e-06,
      "loss": 0.6583,
      "mean_token_accuracy": 0.8021715626120567,
      "num_tokens": 12992353.0,
      "step": 10820
    },
    {
      "entropy": 1.7845357745885848,
      "epoch": 0.0335720524288202,
      "grad_norm": 5.307274341583252,
      "learning_rate": 5.245083793470891e-06,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8313144221901894,
      "num_tokens": 13004955.0,
      "step": 10830
    },
    {
      "entropy": 1.8418287009000778,
      "epoch": 0.03360305155386989,
      "grad_norm": 8.079380989074707,
      "learning_rate": 5.249927346701541e-06,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8063233241438865,
      "num_tokens": 13017444.0,
      "step": 10840
    },
    {
      "entropy": 1.8194942593574523,
      "epoch": 0.033634050678919586,
      "grad_norm": 10.57718563079834,
      "learning_rate": 5.254770899932191e-06,
      "loss": 0.5911,
      "mean_token_accuracy": 0.8096415162086487,
      "num_tokens": 13029772.0,
      "step": 10850
    },
    {
      "entropy": 1.910832443833351,
      "epoch": 0.03366504980396928,
      "grad_norm": 11.854331970214844,
      "learning_rate": 5.25961445316284e-06,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8215459808707237,
      "num_tokens": 13041096.0,
      "step": 10860
    },
    {
      "entropy": 1.8814880549907684,
      "epoch": 0.03369604892901898,
      "grad_norm": 12.540098190307617,
      "learning_rate": 5.2644580063934905e-06,
      "loss": 0.6347,
      "mean_token_accuracy": 0.805778457224369,
      "num_tokens": 13052804.0,
      "step": 10870
    },
    {
      "entropy": 1.8503670692443848,
      "epoch": 0.033727048054068676,
      "grad_norm": 5.048069477081299,
      "learning_rate": 5.26930155962414e-06,
      "loss": 0.5956,
      "mean_token_accuracy": 0.8164945662021637,
      "num_tokens": 13064995.0,
      "step": 10880
    },
    {
      "entropy": 1.871761327981949,
      "epoch": 0.03375804717911837,
      "grad_norm": 17.457483291625977,
      "learning_rate": 5.27414511285479e-06,
      "loss": 0.5965,
      "mean_token_accuracy": 0.8171331033110618,
      "num_tokens": 13076407.0,
      "step": 10890
    },
    {
      "entropy": 1.8655700251460074,
      "epoch": 0.03378904630416806,
      "grad_norm": 5.408127307891846,
      "learning_rate": 5.27898866608544e-06,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8107570946216583,
      "num_tokens": 13088254.0,
      "step": 10900
    },
    {
      "entropy": 1.9251561522483827,
      "epoch": 0.03382004542921776,
      "grad_norm": 11.156283378601074,
      "learning_rate": 5.283832219316091e-06,
      "loss": 0.708,
      "mean_token_accuracy": 0.7975843846797943,
      "num_tokens": 13099187.0,
      "step": 10910
    },
    {
      "entropy": 1.9327981039881705,
      "epoch": 0.033851044554267455,
      "grad_norm": 15.034012794494629,
      "learning_rate": 5.288675772546741e-06,
      "loss": 0.6567,
      "mean_token_accuracy": 0.8078183338046074,
      "num_tokens": 13110633.0,
      "step": 10920
    },
    {
      "entropy": 1.8207261368632317,
      "epoch": 0.03388204367931715,
      "grad_norm": 13.088176727294922,
      "learning_rate": 5.293519325777391e-06,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8240138441324234,
      "num_tokens": 13122812.0,
      "step": 10930
    },
    {
      "entropy": 1.914224511384964,
      "epoch": 0.03391304280436685,
      "grad_norm": 12.61460018157959,
      "learning_rate": 5.298362879008041e-06,
      "loss": 0.6495,
      "mean_token_accuracy": 0.8125144988298416,
      "num_tokens": 13133966.0,
      "step": 10940
    },
    {
      "entropy": 1.8054762348532676,
      "epoch": 0.033944041929416545,
      "grad_norm": 12.354558944702148,
      "learning_rate": 5.303206432238691e-06,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8258527040481567,
      "num_tokens": 13146315.0,
      "step": 10950
    },
    {
      "entropy": 1.7843435242772103,
      "epoch": 0.033975041054466235,
      "grad_norm": 11.364967346191406,
      "learning_rate": 5.308049985469341e-06,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8167797103524208,
      "num_tokens": 13158286.0,
      "step": 10960
    },
    {
      "entropy": 1.8914837822318078,
      "epoch": 0.03400604017951593,
      "grad_norm": 12.683460235595703,
      "learning_rate": 5.31289353869999e-06,
      "loss": 0.6436,
      "mean_token_accuracy": 0.8010558471083641,
      "num_tokens": 13170443.0,
      "step": 10970
    },
    {
      "entropy": 1.8247480258345603,
      "epoch": 0.03403703930456563,
      "grad_norm": 5.69301176071167,
      "learning_rate": 5.31773709193064e-06,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8227360025048256,
      "num_tokens": 13181845.0,
      "step": 10980
    },
    {
      "entropy": 1.8641392514109612,
      "epoch": 0.034068038429615324,
      "grad_norm": 12.605916023254395,
      "learning_rate": 5.322580645161291e-06,
      "loss": 0.6679,
      "mean_token_accuracy": 0.80641338378191,
      "num_tokens": 13193899.0,
      "step": 10990
    },
    {
      "entropy": 1.8444667972624302,
      "epoch": 0.03409903755466502,
      "grad_norm": 5.727068901062012,
      "learning_rate": 5.327424198391941e-06,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8202215030789375,
      "num_tokens": 13206043.0,
      "step": 11000
    },
    {
      "entropy": 1.773200060427189,
      "epoch": 0.03413003667971472,
      "grad_norm": 13.433945655822754,
      "learning_rate": 5.332267751622591e-06,
      "loss": 0.5159,
      "mean_token_accuracy": 0.827538113296032,
      "num_tokens": 13218691.0,
      "step": 11010
    },
    {
      "entropy": 1.9272786289453507,
      "epoch": 0.03416103580476441,
      "grad_norm": 11.8448486328125,
      "learning_rate": 5.3371113048532405e-06,
      "loss": 0.6449,
      "mean_token_accuracy": 0.8075118377804756,
      "num_tokens": 13229756.0,
      "step": 11020
    },
    {
      "entropy": 1.8923597291111947,
      "epoch": 0.034192034929814104,
      "grad_norm": 5.769096374511719,
      "learning_rate": 5.341954858083891e-06,
      "loss": 0.5911,
      "mean_token_accuracy": 0.8067701622843743,
      "num_tokens": 13241705.0,
      "step": 11030
    },
    {
      "entropy": 1.8403789684176446,
      "epoch": 0.0342230340548638,
      "grad_norm": 12.353031158447266,
      "learning_rate": 5.346798411314541e-06,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8218411967158318,
      "num_tokens": 13254778.0,
      "step": 11040
    },
    {
      "entropy": 1.8932681947946548,
      "epoch": 0.0342540331799135,
      "grad_norm": 5.458434581756592,
      "learning_rate": 5.351641964545191e-06,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8183223947882652,
      "num_tokens": 13266049.0,
      "step": 11050
    },
    {
      "entropy": 1.86180839240551,
      "epoch": 0.034285032304963194,
      "grad_norm": 8.72637939453125,
      "learning_rate": 5.356485517775841e-06,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8129496097564697,
      "num_tokens": 13277888.0,
      "step": 11060
    },
    {
      "entropy": 1.9068016976118087,
      "epoch": 0.03431603143001289,
      "grad_norm": 12.052847862243652,
      "learning_rate": 5.36132907100649e-06,
      "loss": 0.6476,
      "mean_token_accuracy": 0.8061621204018593,
      "num_tokens": 13289432.0,
      "step": 11070
    },
    {
      "entropy": 1.8573195546865464,
      "epoch": 0.03434703055506259,
      "grad_norm": 10.996872901916504,
      "learning_rate": 5.366172624237141e-06,
      "loss": 0.5506,
      "mean_token_accuracy": 0.818072022497654,
      "num_tokens": 13301147.0,
      "step": 11080
    },
    {
      "entropy": 1.8704863846302033,
      "epoch": 0.034378029680112276,
      "grad_norm": 13.163795471191406,
      "learning_rate": 5.3710161774677905e-06,
      "loss": 0.5488,
      "mean_token_accuracy": 0.816085159778595,
      "num_tokens": 13312829.0,
      "step": 11090
    },
    {
      "entropy": 1.8130246475338936,
      "epoch": 0.03440902880516197,
      "grad_norm": 14.593989372253418,
      "learning_rate": 5.37585973069844e-06,
      "loss": 0.585,
      "mean_token_accuracy": 0.8232684478163719,
      "num_tokens": 13325336.0,
      "step": 11100
    },
    {
      "entropy": 1.835708625614643,
      "epoch": 0.03444002793021167,
      "grad_norm": 12.702786445617676,
      "learning_rate": 5.380703283929091e-06,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8147074475884437,
      "num_tokens": 13337824.0,
      "step": 11110
    },
    {
      "entropy": 1.8307800009846686,
      "epoch": 0.034471027055261366,
      "grad_norm": 11.551653861999512,
      "learning_rate": 5.385546837159741e-06,
      "loss": 0.594,
      "mean_token_accuracy": 0.811168585717678,
      "num_tokens": 13349534.0,
      "step": 11120
    },
    {
      "entropy": 1.8517664805054665,
      "epoch": 0.03450202618031106,
      "grad_norm": 2.7256433963775635,
      "learning_rate": 5.390390390390391e-06,
      "loss": 0.602,
      "mean_token_accuracy": 0.8006679996848106,
      "num_tokens": 13363408.0,
      "step": 11130
    },
    {
      "entropy": 1.8399043202400207,
      "epoch": 0.03453302530536076,
      "grad_norm": 7.206852912902832,
      "learning_rate": 5.395233943621041e-06,
      "loss": 0.6157,
      "mean_token_accuracy": 0.8117770627140999,
      "num_tokens": 13376213.0,
      "step": 11140
    },
    {
      "entropy": 1.9139886111021043,
      "epoch": 0.03456402443041045,
      "grad_norm": 12.817607879638672,
      "learning_rate": 5.4000774968516915e-06,
      "loss": 0.6346,
      "mean_token_accuracy": 0.8084447503089904,
      "num_tokens": 13387413.0,
      "step": 11150
    },
    {
      "entropy": 1.8213655844330787,
      "epoch": 0.034595023555460146,
      "grad_norm": 11.663803100585938,
      "learning_rate": 5.404921050082341e-06,
      "loss": 0.6088,
      "mean_token_accuracy": 0.8164772510528564,
      "num_tokens": 13399017.0,
      "step": 11160
    },
    {
      "entropy": 1.8936782956123352,
      "epoch": 0.03462602268050984,
      "grad_norm": 10.823102951049805,
      "learning_rate": 5.409764603312991e-06,
      "loss": 0.6693,
      "mean_token_accuracy": 0.810118442773819,
      "num_tokens": 13410180.0,
      "step": 11170
    },
    {
      "entropy": 1.8289520889520645,
      "epoch": 0.03465702180555954,
      "grad_norm": 12.268133163452148,
      "learning_rate": 5.41460815654364e-06,
      "loss": 0.5937,
      "mean_token_accuracy": 0.812929916381836,
      "num_tokens": 13422880.0,
      "step": 11180
    },
    {
      "entropy": 1.7611708968877793,
      "epoch": 0.034688020930609235,
      "grad_norm": 5.681921482086182,
      "learning_rate": 5.41945170977429e-06,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8295134902000427,
      "num_tokens": 13436525.0,
      "step": 11190
    },
    {
      "entropy": 1.894238282740116,
      "epoch": 0.03471902005565893,
      "grad_norm": 9.887272834777832,
      "learning_rate": 5.424295263004941e-06,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8203635275363922,
      "num_tokens": 13447769.0,
      "step": 11200
    },
    {
      "entropy": 1.8931616693735123,
      "epoch": 0.03475001918070862,
      "grad_norm": 10.567079544067383,
      "learning_rate": 5.429138816235591e-06,
      "loss": 0.6151,
      "mean_token_accuracy": 0.813125379383564,
      "num_tokens": 13459142.0,
      "step": 11210
    },
    {
      "entropy": 1.8201383396983146,
      "epoch": 0.03478101830575832,
      "grad_norm": 6.484344482421875,
      "learning_rate": 5.433982369466241e-06,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8103004142642021,
      "num_tokens": 13471313.0,
      "step": 11220
    },
    {
      "entropy": 1.8642778664827346,
      "epoch": 0.034812017430808015,
      "grad_norm": 13.053911209106445,
      "learning_rate": 5.4388259226968906e-06,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8092061296105385,
      "num_tokens": 13483091.0,
      "step": 11230
    },
    {
      "entropy": 1.8669605866074561,
      "epoch": 0.03484301655585771,
      "grad_norm": 11.9703369140625,
      "learning_rate": 5.443669475927541e-06,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8229537546634674,
      "num_tokens": 13494177.0,
      "step": 11240
    },
    {
      "entropy": 1.85545664280653,
      "epoch": 0.03487401568090741,
      "grad_norm": 6.333286762237549,
      "learning_rate": 5.448513029158191e-06,
      "loss": 0.6025,
      "mean_token_accuracy": 0.810517068207264,
      "num_tokens": 13507247.0,
      "step": 11250
    },
    {
      "entropy": 1.7715984418988229,
      "epoch": 0.034905014805957105,
      "grad_norm": 10.145940780639648,
      "learning_rate": 5.453356582388841e-06,
      "loss": 0.5342,
      "mean_token_accuracy": 0.825646486878395,
      "num_tokens": 13520781.0,
      "step": 11260
    },
    {
      "entropy": 1.9231941044330596,
      "epoch": 0.034936013931006794,
      "grad_norm": 12.940445899963379,
      "learning_rate": 5.458200135619491e-06,
      "loss": 0.6732,
      "mean_token_accuracy": 0.799634762108326,
      "num_tokens": 13531942.0,
      "step": 11270
    },
    {
      "entropy": 1.8580936834216117,
      "epoch": 0.03496701305605649,
      "grad_norm": 13.450532913208008,
      "learning_rate": 5.46304368885014e-06,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8047621414065361,
      "num_tokens": 13543831.0,
      "step": 11280
    },
    {
      "entropy": 1.8055393621325493,
      "epoch": 0.03499801218110619,
      "grad_norm": 12.781180381774902,
      "learning_rate": 5.467887242080791e-06,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8207426607608795,
      "num_tokens": 13555967.0,
      "step": 11290
    },
    {
      "entropy": 1.804980905354023,
      "epoch": 0.035029011306155884,
      "grad_norm": 12.926373481750488,
      "learning_rate": 5.472730795311441e-06,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8110515549778938,
      "num_tokens": 13569161.0,
      "step": 11300
    },
    {
      "entropy": 1.9503383368253708,
      "epoch": 0.03506001043120558,
      "grad_norm": 12.764983177185059,
      "learning_rate": 5.4775743485420905e-06,
      "loss": 0.6721,
      "mean_token_accuracy": 0.7947013214230537,
      "num_tokens": 13580275.0,
      "step": 11310
    },
    {
      "entropy": 1.88107870221138,
      "epoch": 0.03509100955625528,
      "grad_norm": 14.808971405029297,
      "learning_rate": 5.482417901772741e-06,
      "loss": 0.6075,
      "mean_token_accuracy": 0.8077093288302422,
      "num_tokens": 13592745.0,
      "step": 11320
    },
    {
      "entropy": 1.8953585937619208,
      "epoch": 0.03512200868130497,
      "grad_norm": 12.039188385009766,
      "learning_rate": 5.487261455003391e-06,
      "loss": 0.6186,
      "mean_token_accuracy": 0.8103330656886101,
      "num_tokens": 13604350.0,
      "step": 11330
    },
    {
      "entropy": 1.8628807738423347,
      "epoch": 0.035153007806354664,
      "grad_norm": 11.431326866149902,
      "learning_rate": 5.492105008234041e-06,
      "loss": 0.5886,
      "mean_token_accuracy": 0.8245170086622238,
      "num_tokens": 13616428.0,
      "step": 11340
    },
    {
      "entropy": 1.9492103517055512,
      "epoch": 0.03518400693140436,
      "grad_norm": 13.414464950561523,
      "learning_rate": 5.496948561464691e-06,
      "loss": 0.6753,
      "mean_token_accuracy": 0.7961807489395142,
      "num_tokens": 13627865.0,
      "step": 11350
    },
    {
      "entropy": 1.9296086087822915,
      "epoch": 0.03521500605645406,
      "grad_norm": 12.078662872314453,
      "learning_rate": 5.501792114695342e-06,
      "loss": 0.7047,
      "mean_token_accuracy": 0.7942573204636574,
      "num_tokens": 13639469.0,
      "step": 11360
    },
    {
      "entropy": 1.9266355335712433,
      "epoch": 0.03524600518150375,
      "grad_norm": 10.940107345581055,
      "learning_rate": 5.5066356679259915e-06,
      "loss": 0.6479,
      "mean_token_accuracy": 0.8066198214888572,
      "num_tokens": 13650098.0,
      "step": 11370
    },
    {
      "entropy": 1.7104891866445542,
      "epoch": 0.03527700430655345,
      "grad_norm": 13.101654052734375,
      "learning_rate": 5.511479221156641e-06,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8279589235782623,
      "num_tokens": 13664250.0,
      "step": 11380
    },
    {
      "entropy": 1.8383988574147225,
      "epoch": 0.03530800343160314,
      "grad_norm": 12.127103805541992,
      "learning_rate": 5.51632277438729e-06,
      "loss": 0.6044,
      "mean_token_accuracy": 0.8270195707678795,
      "num_tokens": 13675591.0,
      "step": 11390
    },
    {
      "entropy": 1.8982412829995154,
      "epoch": 0.035339002556652836,
      "grad_norm": 13.012894630432129,
      "learning_rate": 5.52116632761794e-06,
      "loss": 0.6433,
      "mean_token_accuracy": 0.8037685632705689,
      "num_tokens": 13687364.0,
      "step": 11400
    },
    {
      "entropy": 1.85855975151062,
      "epoch": 0.03537000168170253,
      "grad_norm": 5.7059454917907715,
      "learning_rate": 5.526009880848591e-06,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8210862413048744,
      "num_tokens": 13699442.0,
      "step": 11410
    },
    {
      "entropy": 1.8508766368031502,
      "epoch": 0.03540100080675223,
      "grad_norm": 12.565545082092285,
      "learning_rate": 5.530853434079241e-06,
      "loss": 0.6132,
      "mean_token_accuracy": 0.8048885554075241,
      "num_tokens": 13711862.0,
      "step": 11420
    },
    {
      "entropy": 1.8528811484575272,
      "epoch": 0.035431999931801926,
      "grad_norm": 10.851259231567383,
      "learning_rate": 5.535696987309891e-06,
      "loss": 0.6248,
      "mean_token_accuracy": 0.8083219036459923,
      "num_tokens": 13723049.0,
      "step": 11430
    },
    {
      "entropy": 1.7467531949281692,
      "epoch": 0.03546299905685162,
      "grad_norm": 12.76328182220459,
      "learning_rate": 5.540540540540541e-06,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8341274619102478,
      "num_tokens": 13736433.0,
      "step": 11440
    },
    {
      "entropy": 1.9269695818424224,
      "epoch": 0.03549399818190132,
      "grad_norm": 13.179561614990234,
      "learning_rate": 5.545384093771191e-06,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8090446889400482,
      "num_tokens": 13748433.0,
      "step": 11450
    },
    {
      "entropy": 1.871768619120121,
      "epoch": 0.03552499730695101,
      "grad_norm": 11.79694652557373,
      "learning_rate": 5.550227647001841e-06,
      "loss": 0.6131,
      "mean_token_accuracy": 0.8084805279970169,
      "num_tokens": 13760554.0,
      "step": 11460
    },
    {
      "entropy": 1.899412962794304,
      "epoch": 0.035555996432000705,
      "grad_norm": 10.14828872680664,
      "learning_rate": 5.555071200232491e-06,
      "loss": 0.6452,
      "mean_token_accuracy": 0.8113683596253395,
      "num_tokens": 13771849.0,
      "step": 11470
    },
    {
      "entropy": 1.8810555890202523,
      "epoch": 0.0355869955570504,
      "grad_norm": 13.629297256469727,
      "learning_rate": 5.559914753463141e-06,
      "loss": 0.6262,
      "mean_token_accuracy": 0.8057652726769448,
      "num_tokens": 13783349.0,
      "step": 11480
    },
    {
      "entropy": 1.8832147046923637,
      "epoch": 0.0356179946821001,
      "grad_norm": 11.237863540649414,
      "learning_rate": 5.56475830669379e-06,
      "loss": 0.6097,
      "mean_token_accuracy": 0.80062695145607,
      "num_tokens": 13794988.0,
      "step": 11490
    },
    {
      "entropy": 1.734711329638958,
      "epoch": 0.035648993807149795,
      "grad_norm": 15.71545696258545,
      "learning_rate": 5.569601859924441e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.8244102329015732,
      "num_tokens": 13808196.0,
      "step": 11500
    },
    {
      "entropy": 1.8736824676394463,
      "epoch": 0.03567999293219949,
      "grad_norm": 10.090490341186523,
      "learning_rate": 5.574445413155091e-06,
      "loss": 0.5995,
      "mean_token_accuracy": 0.8152802541851998,
      "num_tokens": 13820833.0,
      "step": 11510
    },
    {
      "entropy": 1.7825218871235848,
      "epoch": 0.03571099205724918,
      "grad_norm": 10.388526916503906,
      "learning_rate": 5.579288966385741e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8239463314414024,
      "num_tokens": 13834118.0,
      "step": 11520
    },
    {
      "entropy": 1.9106709718704225,
      "epoch": 0.03574199118229888,
      "grad_norm": 5.574368000030518,
      "learning_rate": 5.584132519616391e-06,
      "loss": 0.6794,
      "mean_token_accuracy": 0.8046201914548874,
      "num_tokens": 13845774.0,
      "step": 11530
    },
    {
      "entropy": 1.9362187922000884,
      "epoch": 0.035772990307348575,
      "grad_norm": 11.167245864868164,
      "learning_rate": 5.588976072847041e-06,
      "loss": 0.6901,
      "mean_token_accuracy": 0.7971135929226876,
      "num_tokens": 13856205.0,
      "step": 11540
    },
    {
      "entropy": 1.792908415198326,
      "epoch": 0.03580398943239827,
      "grad_norm": 5.944072723388672,
      "learning_rate": 5.593819626077691e-06,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8229343697428704,
      "num_tokens": 13868827.0,
      "step": 11550
    },
    {
      "entropy": 1.7896666795015335,
      "epoch": 0.03583498855744797,
      "grad_norm": 5.675596237182617,
      "learning_rate": 5.598663179308341e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8251826211810112,
      "num_tokens": 13881592.0,
      "step": 11560
    },
    {
      "entropy": 1.9190635159611702,
      "epoch": 0.035865987682497665,
      "grad_norm": 14.54418659210205,
      "learning_rate": 5.603506732538992e-06,
      "loss": 0.6535,
      "mean_token_accuracy": 0.8147982686758042,
      "num_tokens": 13892831.0,
      "step": 11570
    },
    {
      "entropy": 1.9398990795016289,
      "epoch": 0.035896986807547354,
      "grad_norm": 5.379489898681641,
      "learning_rate": 5.608350285769642e-06,
      "loss": 0.6346,
      "mean_token_accuracy": 0.807078929245472,
      "num_tokens": 13903824.0,
      "step": 11580
    },
    {
      "entropy": 1.8159923285245896,
      "epoch": 0.03592798593259705,
      "grad_norm": 12.649768829345703,
      "learning_rate": 5.6131938390002915e-06,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8225948542356492,
      "num_tokens": 13916211.0,
      "step": 11590
    },
    {
      "entropy": 1.8722815930843353,
      "epoch": 0.03595898505764675,
      "grad_norm": 11.376723289489746,
      "learning_rate": 5.6180373922309405e-06,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8152380838990212,
      "num_tokens": 13927975.0,
      "step": 11600
    },
    {
      "entropy": 1.8421933129429817,
      "epoch": 0.035989984182696444,
      "grad_norm": 11.949515342712402,
      "learning_rate": 5.62288094546159e-06,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8134393319487572,
      "num_tokens": 13939474.0,
      "step": 11610
    },
    {
      "entropy": 1.888566829264164,
      "epoch": 0.03602098330774614,
      "grad_norm": 17.51249885559082,
      "learning_rate": 5.627724498692241e-06,
      "loss": 0.624,
      "mean_token_accuracy": 0.817312179505825,
      "num_tokens": 13951148.0,
      "step": 11620
    },
    {
      "entropy": 1.8805680245161056,
      "epoch": 0.03605198243279584,
      "grad_norm": 11.506142616271973,
      "learning_rate": 5.632568051922891e-06,
      "loss": 0.6413,
      "mean_token_accuracy": 0.8139393076300621,
      "num_tokens": 13962535.0,
      "step": 11630
    },
    {
      "entropy": 1.939769622683525,
      "epoch": 0.03608298155784553,
      "grad_norm": 11.291607856750488,
      "learning_rate": 5.637411605153541e-06,
      "loss": 0.6347,
      "mean_token_accuracy": 0.81319679915905,
      "num_tokens": 13973297.0,
      "step": 11640
    },
    {
      "entropy": 1.871855989843607,
      "epoch": 0.03611398068289522,
      "grad_norm": 13.56961727142334,
      "learning_rate": 5.642255158384191e-06,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8270270243287087,
      "num_tokens": 13985337.0,
      "step": 11650
    },
    {
      "entropy": 1.807421001791954,
      "epoch": 0.03614497980794492,
      "grad_norm": 4.713283538818359,
      "learning_rate": 5.6470987116148415e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8386715143918991,
      "num_tokens": 13998425.0,
      "step": 11660
    },
    {
      "entropy": 1.784872618317604,
      "epoch": 0.03617597893299462,
      "grad_norm": 11.134720802307129,
      "learning_rate": 5.651942264845491e-06,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8182505384087563,
      "num_tokens": 14011957.0,
      "step": 11670
    },
    {
      "entropy": 1.842606595158577,
      "epoch": 0.03620697805804431,
      "grad_norm": 5.860245227813721,
      "learning_rate": 5.656785818076141e-06,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8185199961066246,
      "num_tokens": 14023989.0,
      "step": 11680
    },
    {
      "entropy": 1.7569019049406052,
      "epoch": 0.03623797718309401,
      "grad_norm": 11.391021728515625,
      "learning_rate": 5.661629371306791e-06,
      "loss": 0.4851,
      "mean_token_accuracy": 0.8376433789730072,
      "num_tokens": 14037138.0,
      "step": 11690
    },
    {
      "entropy": 1.8167656242847443,
      "epoch": 0.0362689763081437,
      "grad_norm": 14.077984809875488,
      "learning_rate": 5.66647292453744e-06,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8293104261159897,
      "num_tokens": 14049119.0,
      "step": 11700
    },
    {
      "entropy": 1.8543547958135604,
      "epoch": 0.036299975433193396,
      "grad_norm": 9.835662841796875,
      "learning_rate": 5.671316477768091e-06,
      "loss": 0.598,
      "mean_token_accuracy": 0.8196697190403939,
      "num_tokens": 14061251.0,
      "step": 11710
    },
    {
      "entropy": 1.8527500122785567,
      "epoch": 0.03633097455824309,
      "grad_norm": 13.09237289428711,
      "learning_rate": 5.676160030998741e-06,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8196158647537232,
      "num_tokens": 14073877.0,
      "step": 11720
    },
    {
      "entropy": 1.870137917995453,
      "epoch": 0.03636197368329279,
      "grad_norm": 12.378225326538086,
      "learning_rate": 5.681003584229391e-06,
      "loss": 0.6102,
      "mean_token_accuracy": 0.81648840457201,
      "num_tokens": 14086089.0,
      "step": 11730
    },
    {
      "entropy": 1.8904090464115142,
      "epoch": 0.036392972808342486,
      "grad_norm": 10.42041301727295,
      "learning_rate": 5.6858471374600414e-06,
      "loss": 0.5896,
      "mean_token_accuracy": 0.8135670512914658,
      "num_tokens": 14098409.0,
      "step": 11740
    },
    {
      "entropy": 1.9056991636753082,
      "epoch": 0.03642397193339218,
      "grad_norm": 5.637151718139648,
      "learning_rate": 5.690690690690691e-06,
      "loss": 0.6525,
      "mean_token_accuracy": 0.8007242888212204,
      "num_tokens": 14110078.0,
      "step": 11750
    },
    {
      "entropy": 1.7788226932287217,
      "epoch": 0.03645497105844187,
      "grad_norm": 6.211390018463135,
      "learning_rate": 5.695534243921341e-06,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8264580845832825,
      "num_tokens": 14122941.0,
      "step": 11760
    },
    {
      "entropy": 1.8122435554862022,
      "epoch": 0.03648597018349157,
      "grad_norm": 10.39504623413086,
      "learning_rate": 5.700377797151991e-06,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8231259554624557,
      "num_tokens": 14135164.0,
      "step": 11770
    },
    {
      "entropy": 1.9115518778562546,
      "epoch": 0.036516969308541265,
      "grad_norm": 11.225345611572266,
      "learning_rate": 5.705221350382642e-06,
      "loss": 0.6966,
      "mean_token_accuracy": 0.8027231857180596,
      "num_tokens": 14145993.0,
      "step": 11780
    },
    {
      "entropy": 1.833278726041317,
      "epoch": 0.03654796843359096,
      "grad_norm": 9.642135620117188,
      "learning_rate": 5.710064903613292e-06,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8039411261677742,
      "num_tokens": 14159775.0,
      "step": 11790
    },
    {
      "entropy": 1.7943380519747734,
      "epoch": 0.03657896755864066,
      "grad_norm": 11.112443923950195,
      "learning_rate": 5.714908456843942e-06,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8308914095163346,
      "num_tokens": 14172584.0,
      "step": 11800
    },
    {
      "entropy": 1.7970113635063172,
      "epoch": 0.036609966683690355,
      "grad_norm": 12.6159029006958,
      "learning_rate": 5.719752010074591e-06,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8075689300894737,
      "num_tokens": 14184681.0,
      "step": 11810
    },
    {
      "entropy": 1.7890537694096564,
      "epoch": 0.03664096580874005,
      "grad_norm": 4.384820461273193,
      "learning_rate": 5.7245955633052405e-06,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8110948413610458,
      "num_tokens": 14198022.0,
      "step": 11820
    },
    {
      "entropy": 1.814190225303173,
      "epoch": 0.03667196493378974,
      "grad_norm": 4.793267726898193,
      "learning_rate": 5.729439116535891e-06,
      "loss": 0.5878,
      "mean_token_accuracy": 0.8212701037526131,
      "num_tokens": 14209842.0,
      "step": 11830
    },
    {
      "entropy": 1.7815122097730636,
      "epoch": 0.03670296405883944,
      "grad_norm": 10.278016090393066,
      "learning_rate": 5.734282669766541e-06,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8179376661777497,
      "num_tokens": 14222224.0,
      "step": 11840
    },
    {
      "entropy": 1.8016631290316583,
      "epoch": 0.036733963183889135,
      "grad_norm": 11.094133377075195,
      "learning_rate": 5.739126222997191e-06,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8138196378946304,
      "num_tokens": 14234855.0,
      "step": 11850
    },
    {
      "entropy": 1.8083413437008857,
      "epoch": 0.03676496230893883,
      "grad_norm": 6.810488700866699,
      "learning_rate": 5.743969776227841e-06,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8194642826914788,
      "num_tokens": 14246248.0,
      "step": 11860
    },
    {
      "entropy": 1.754711863398552,
      "epoch": 0.03679596143398853,
      "grad_norm": 5.887208938598633,
      "learning_rate": 5.748813329458492e-06,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8068835064768791,
      "num_tokens": 14258816.0,
      "step": 11870
    },
    {
      "entropy": 1.7588348999619483,
      "epoch": 0.036826960559038224,
      "grad_norm": 12.953771591186523,
      "learning_rate": 5.7536568826891415e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8235802292823792,
      "num_tokens": 14271525.0,
      "step": 11880
    },
    {
      "entropy": 1.8742596834897995,
      "epoch": 0.036857959684087914,
      "grad_norm": 11.466156959533691,
      "learning_rate": 5.758500435919791e-06,
      "loss": 0.6855,
      "mean_token_accuracy": 0.8022027894854545,
      "num_tokens": 14282330.0,
      "step": 11890
    },
    {
      "entropy": 1.7715501874685287,
      "epoch": 0.03688895880913761,
      "grad_norm": 10.178691864013672,
      "learning_rate": 5.763343989150441e-06,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8171212136745453,
      "num_tokens": 14294446.0,
      "step": 11900
    },
    {
      "entropy": 1.8660178661346436,
      "epoch": 0.03691995793418731,
      "grad_norm": 11.348404884338379,
      "learning_rate": 5.76818754238109e-06,
      "loss": 0.6431,
      "mean_token_accuracy": 0.8038701593875885,
      "num_tokens": 14305947.0,
      "step": 11910
    },
    {
      "entropy": 1.7659137800335885,
      "epoch": 0.036950957059237004,
      "grad_norm": 11.47120475769043,
      "learning_rate": 5.773031095611741e-06,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8206716164946556,
      "num_tokens": 14318249.0,
      "step": 11920
    },
    {
      "entropy": 1.747839505970478,
      "epoch": 0.0369819561842867,
      "grad_norm": 5.5586018562316895,
      "learning_rate": 5.777874648842391e-06,
      "loss": 0.559,
      "mean_token_accuracy": 0.8257301524281502,
      "num_tokens": 14330993.0,
      "step": 11930
    },
    {
      "entropy": 1.8217350393533707,
      "epoch": 0.0370129553093364,
      "grad_norm": 10.629807472229004,
      "learning_rate": 5.782718202073041e-06,
      "loss": 0.6345,
      "mean_token_accuracy": 0.8065755069255829,
      "num_tokens": 14343198.0,
      "step": 11940
    },
    {
      "entropy": 1.8037606567144393,
      "epoch": 0.03704395443438609,
      "grad_norm": 8.647960662841797,
      "learning_rate": 5.7875617553036915e-06,
      "loss": 0.598,
      "mean_token_accuracy": 0.8171042606234551,
      "num_tokens": 14355609.0,
      "step": 11950
    },
    {
      "entropy": 1.8690994665026666,
      "epoch": 0.03707495355943578,
      "grad_norm": 13.143199920654297,
      "learning_rate": 5.792405308534341e-06,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8137925118207932,
      "num_tokens": 14367235.0,
      "step": 11960
    },
    {
      "entropy": 1.7714310929179191,
      "epoch": 0.03710595268448548,
      "grad_norm": 11.381326675415039,
      "learning_rate": 5.797248861764991e-06,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8231312274932862,
      "num_tokens": 14379885.0,
      "step": 11970
    },
    {
      "entropy": 1.8847584262490273,
      "epoch": 0.037136951809535176,
      "grad_norm": 12.403789520263672,
      "learning_rate": 5.802092414995641e-06,
      "loss": 0.6177,
      "mean_token_accuracy": 0.8122402995824813,
      "num_tokens": 14390884.0,
      "step": 11980
    },
    {
      "entropy": 1.7906603574752809,
      "epoch": 0.03716795093458487,
      "grad_norm": 12.341434478759766,
      "learning_rate": 5.806935968226292e-06,
      "loss": 0.6091,
      "mean_token_accuracy": 0.8056104972958564,
      "num_tokens": 14402659.0,
      "step": 11990
    },
    {
      "entropy": 1.8461511224508285,
      "epoch": 0.03719895005963457,
      "grad_norm": 11.438586235046387,
      "learning_rate": 5.811779521456942e-06,
      "loss": 0.6092,
      "mean_token_accuracy": 0.8133256167173386,
      "num_tokens": 14413793.0,
      "step": 12000
    },
    {
      "entropy": 1.797146451473236,
      "epoch": 0.03722994918468426,
      "grad_norm": 12.915786743164062,
      "learning_rate": 5.816623074687592e-06,
      "loss": 0.5399,
      "mean_token_accuracy": 0.826125793159008,
      "num_tokens": 14426225.0,
      "step": 12010
    },
    {
      "entropy": 1.7861742541193961,
      "epoch": 0.037260948309733956,
      "grad_norm": 11.406455993652344,
      "learning_rate": 5.821466627918241e-06,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8146935313940048,
      "num_tokens": 14438143.0,
      "step": 12020
    },
    {
      "entropy": 1.8457993239164352,
      "epoch": 0.03729194743478365,
      "grad_norm": 10.185896873474121,
      "learning_rate": 5.826310181148891e-06,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8021126002073288,
      "num_tokens": 14451306.0,
      "step": 12030
    },
    {
      "entropy": 1.8906755059957505,
      "epoch": 0.03732294655983335,
      "grad_norm": 12.862022399902344,
      "learning_rate": 5.831153734379541e-06,
      "loss": 0.722,
      "mean_token_accuracy": 0.7925963416695595,
      "num_tokens": 14463297.0,
      "step": 12040
    },
    {
      "entropy": 1.760866042971611,
      "epoch": 0.037353945684883046,
      "grad_norm": 12.60151195526123,
      "learning_rate": 5.835997287610191e-06,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8173602819442749,
      "num_tokens": 14476740.0,
      "step": 12050
    },
    {
      "entropy": 1.8725519806146622,
      "epoch": 0.03738494480993274,
      "grad_norm": 13.23414134979248,
      "learning_rate": 5.840840840840841e-06,
      "loss": 0.6585,
      "mean_token_accuracy": 0.804969422519207,
      "num_tokens": 14488274.0,
      "step": 12060
    },
    {
      "entropy": 1.821217942237854,
      "epoch": 0.03741594393498243,
      "grad_norm": 12.871102333068848,
      "learning_rate": 5.845684394071491e-06,
      "loss": 0.5877,
      "mean_token_accuracy": 0.8180323630571366,
      "num_tokens": 14499989.0,
      "step": 12070
    },
    {
      "entropy": 1.8127268552780151,
      "epoch": 0.03744694306003213,
      "grad_norm": 12.796869277954102,
      "learning_rate": 5.850527947302142e-06,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8186282083392143,
      "num_tokens": 14512825.0,
      "step": 12080
    },
    {
      "entropy": 1.8107111915946006,
      "epoch": 0.037477942185081825,
      "grad_norm": 5.6175360679626465,
      "learning_rate": 5.855371500532792e-06,
      "loss": 0.5838,
      "mean_token_accuracy": 0.809958079457283,
      "num_tokens": 14525109.0,
      "step": 12090
    },
    {
      "entropy": 1.8637285217642785,
      "epoch": 0.03750894131013152,
      "grad_norm": 12.704855918884277,
      "learning_rate": 5.8602150537634415e-06,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8088982105255127,
      "num_tokens": 14536754.0,
      "step": 12100
    },
    {
      "entropy": 1.7325578138232232,
      "epoch": 0.03753994043518122,
      "grad_norm": 6.377108573913574,
      "learning_rate": 5.865058606994092e-06,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8252245962619782,
      "num_tokens": 14551185.0,
      "step": 12110
    },
    {
      "entropy": 1.802834041416645,
      "epoch": 0.037570939560230915,
      "grad_norm": 12.303117752075195,
      "learning_rate": 5.86990216022474e-06,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8231405153870582,
      "num_tokens": 14564613.0,
      "step": 12120
    },
    {
      "entropy": 1.8981367230415345,
      "epoch": 0.037601938685280605,
      "grad_norm": 12.030719757080078,
      "learning_rate": 5.874745713455391e-06,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8138510972261429,
      "num_tokens": 14575459.0,
      "step": 12130
    },
    {
      "entropy": 1.8835298061370849,
      "epoch": 0.0376329378103303,
      "grad_norm": 11.876147270202637,
      "learning_rate": 5.879589266686041e-06,
      "loss": 0.624,
      "mean_token_accuracy": 0.8129880890250206,
      "num_tokens": 14587190.0,
      "step": 12140
    },
    {
      "entropy": 1.8323535963892936,
      "epoch": 0.03766393693538,
      "grad_norm": 13.480051040649414,
      "learning_rate": 5.884432819916691e-06,
      "loss": 0.5953,
      "mean_token_accuracy": 0.819337697327137,
      "num_tokens": 14598092.0,
      "step": 12150
    },
    {
      "entropy": 1.874643650650978,
      "epoch": 0.037694936060429694,
      "grad_norm": 10.773601531982422,
      "learning_rate": 5.889276373147342e-06,
      "loss": 0.6391,
      "mean_token_accuracy": 0.8008844360709191,
      "num_tokens": 14609253.0,
      "step": 12160
    },
    {
      "entropy": 1.864101167023182,
      "epoch": 0.03772593518547939,
      "grad_norm": 5.442148208618164,
      "learning_rate": 5.8941199263779915e-06,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8117442667484284,
      "num_tokens": 14620881.0,
      "step": 12170
    },
    {
      "entropy": 1.9181242629885673,
      "epoch": 0.03775693431052909,
      "grad_norm": 10.096672058105469,
      "learning_rate": 5.898963479608641e-06,
      "loss": 0.6121,
      "mean_token_accuracy": 0.8077910885214805,
      "num_tokens": 14632799.0,
      "step": 12180
    },
    {
      "entropy": 1.870473875105381,
      "epoch": 0.037787933435578784,
      "grad_norm": 12.328435897827148,
      "learning_rate": 5.903807032839291e-06,
      "loss": 0.5875,
      "mean_token_accuracy": 0.8157344311475754,
      "num_tokens": 14643832.0,
      "step": 12190
    },
    {
      "entropy": 1.8657541304826737,
      "epoch": 0.037818932560628474,
      "grad_norm": 14.332444190979004,
      "learning_rate": 5.908650586069942e-06,
      "loss": 0.62,
      "mean_token_accuracy": 0.8118143856525422,
      "num_tokens": 14655521.0,
      "step": 12200
    },
    {
      "entropy": 1.8177946463227272,
      "epoch": 0.03784993168567817,
      "grad_norm": 11.780436515808105,
      "learning_rate": 5.913494139300592e-06,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8122022807598114,
      "num_tokens": 14668492.0,
      "step": 12210
    },
    {
      "entropy": 1.8311690405011176,
      "epoch": 0.03788093081072787,
      "grad_norm": 10.684754371643066,
      "learning_rate": 5.918337692531242e-06,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8185856059193611,
      "num_tokens": 14681336.0,
      "step": 12220
    },
    {
      "entropy": 1.8790747031569481,
      "epoch": 0.037911929935777564,
      "grad_norm": 12.979804039001465,
      "learning_rate": 5.923181245761891e-06,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8123493686318397,
      "num_tokens": 14693888.0,
      "step": 12230
    },
    {
      "entropy": 1.8931312650442123,
      "epoch": 0.03794292906082726,
      "grad_norm": 12.229876518249512,
      "learning_rate": 5.928024798992541e-06,
      "loss": 0.625,
      "mean_token_accuracy": 0.8202015697956085,
      "num_tokens": 14704937.0,
      "step": 12240
    },
    {
      "entropy": 1.8407990396022798,
      "epoch": 0.03797392818587696,
      "grad_norm": 11.653843879699707,
      "learning_rate": 5.9328683522231914e-06,
      "loss": 0.5902,
      "mean_token_accuracy": 0.8176357612013817,
      "num_tokens": 14717674.0,
      "step": 12250
    },
    {
      "entropy": 1.8561912134289742,
      "epoch": 0.038004927310926646,
      "grad_norm": 13.023904800415039,
      "learning_rate": 5.937711905453841e-06,
      "loss": 0.5849,
      "mean_token_accuracy": 0.8190959095954895,
      "num_tokens": 14729749.0,
      "step": 12260
    },
    {
      "entropy": 1.9160822182893753,
      "epoch": 0.03803592643597634,
      "grad_norm": 12.721014976501465,
      "learning_rate": 5.942555458684491e-06,
      "loss": 0.6519,
      "mean_token_accuracy": 0.8113419517874718,
      "num_tokens": 14741605.0,
      "step": 12270
    },
    {
      "entropy": 1.8194750413298606,
      "epoch": 0.03806692556102604,
      "grad_norm": 11.955613136291504,
      "learning_rate": 5.947399011915141e-06,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8214196056127548,
      "num_tokens": 14754303.0,
      "step": 12280
    },
    {
      "entropy": 1.8162646040320396,
      "epoch": 0.038097924686075736,
      "grad_norm": 5.1032328605651855,
      "learning_rate": 5.952242565145792e-06,
      "loss": 0.5327,
      "mean_token_accuracy": 0.818191859126091,
      "num_tokens": 14766657.0,
      "step": 12290
    },
    {
      "entropy": 1.846658205986023,
      "epoch": 0.03812892381112543,
      "grad_norm": 5.808051109313965,
      "learning_rate": 5.957086118376442e-06,
      "loss": 0.63,
      "mean_token_accuracy": 0.8142410039901733,
      "num_tokens": 14778196.0,
      "step": 12300
    },
    {
      "entropy": 1.7994149655103684,
      "epoch": 0.03815992293617513,
      "grad_norm": 12.859893798828125,
      "learning_rate": 5.961929671607092e-06,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8262584999203682,
      "num_tokens": 14790293.0,
      "step": 12310
    },
    {
      "entropy": 1.951202955842018,
      "epoch": 0.03819092206122482,
      "grad_norm": 9.63222599029541,
      "learning_rate": 5.966773224837742e-06,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8155479997396469,
      "num_tokens": 14801257.0,
      "step": 12320
    },
    {
      "entropy": 1.8327611729502677,
      "epoch": 0.038221921186274516,
      "grad_norm": 12.394364356994629,
      "learning_rate": 5.9716167780683905e-06,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8244035348296166,
      "num_tokens": 14813680.0,
      "step": 12330
    },
    {
      "entropy": 1.8879570379853248,
      "epoch": 0.03825292031132421,
      "grad_norm": 9.891463279724121,
      "learning_rate": 5.976460331299041e-06,
      "loss": 0.5948,
      "mean_token_accuracy": 0.815042594075203,
      "num_tokens": 14825260.0,
      "step": 12340
    },
    {
      "entropy": 1.8341692447662354,
      "epoch": 0.03828391943637391,
      "grad_norm": 14.30816650390625,
      "learning_rate": 5.981303884529691e-06,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8117454037070274,
      "num_tokens": 14837969.0,
      "step": 12350
    },
    {
      "entropy": 1.8351563602685927,
      "epoch": 0.038314918561423605,
      "grad_norm": 12.036502838134766,
      "learning_rate": 5.986147437760341e-06,
      "loss": 0.5528,
      "mean_token_accuracy": 0.820985272526741,
      "num_tokens": 14850504.0,
      "step": 12360
    },
    {
      "entropy": 1.8652123495936395,
      "epoch": 0.0383459176864733,
      "grad_norm": 5.860367298126221,
      "learning_rate": 5.990990990990992e-06,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8303888604044914,
      "num_tokens": 14862430.0,
      "step": 12370
    },
    {
      "entropy": 1.8906607389450074,
      "epoch": 0.03837691681152299,
      "grad_norm": 11.164518356323242,
      "learning_rate": 5.995834544221642e-06,
      "loss": 0.5727,
      "mean_token_accuracy": 0.8233511716127395,
      "num_tokens": 14874608.0,
      "step": 12380
    },
    {
      "entropy": 1.9229727298021317,
      "epoch": 0.03840791593657269,
      "grad_norm": 10.484548568725586,
      "learning_rate": 6.0006780974522915e-06,
      "loss": 0.617,
      "mean_token_accuracy": 0.8203805893659591,
      "num_tokens": 14885555.0,
      "step": 12390
    },
    {
      "entropy": 1.858140294253826,
      "epoch": 0.038438915061622385,
      "grad_norm": 3.475583791732788,
      "learning_rate": 6.005521650682941e-06,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8278280302882195,
      "num_tokens": 14899067.0,
      "step": 12400
    },
    {
      "entropy": 1.9508850559592248,
      "epoch": 0.03846991418667208,
      "grad_norm": 14.349974632263184,
      "learning_rate": 6.010365203913592e-06,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8059279710054398,
      "num_tokens": 14910681.0,
      "step": 12410
    },
    {
      "entropy": 1.9063424944877625,
      "epoch": 0.03850091331172178,
      "grad_norm": 12.377765655517578,
      "learning_rate": 6.015208757144242e-06,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8251767575740814,
      "num_tokens": 14922446.0,
      "step": 12420
    },
    {
      "entropy": 1.8672158405184747,
      "epoch": 0.038531912436771475,
      "grad_norm": 14.53426456451416,
      "learning_rate": 6.020052310374892e-06,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8267401665449142,
      "num_tokens": 14934912.0,
      "step": 12430
    },
    {
      "entropy": 1.946044033765793,
      "epoch": 0.038562911561821164,
      "grad_norm": 12.349529266357422,
      "learning_rate": 6.024895863605541e-06,
      "loss": 0.7026,
      "mean_token_accuracy": 0.7958286583423615,
      "num_tokens": 14946270.0,
      "step": 12440
    },
    {
      "entropy": 1.9352963030338288,
      "epoch": 0.03859391068687086,
      "grad_norm": 15.071447372436523,
      "learning_rate": 6.029739416836191e-06,
      "loss": 0.6828,
      "mean_token_accuracy": 0.8007722824811936,
      "num_tokens": 14957285.0,
      "step": 12450
    },
    {
      "entropy": 1.9621846169233321,
      "epoch": 0.03862490981192056,
      "grad_norm": 11.935574531555176,
      "learning_rate": 6.0345829700668415e-06,
      "loss": 0.7034,
      "mean_token_accuracy": 0.7971277639269829,
      "num_tokens": 14968080.0,
      "step": 12460
    },
    {
      "entropy": 1.8536121487617492,
      "epoch": 0.038655908936970254,
      "grad_norm": 12.425955772399902,
      "learning_rate": 6.0394265232974914e-06,
      "loss": 0.571,
      "mean_token_accuracy": 0.8072732269763947,
      "num_tokens": 14980128.0,
      "step": 12470
    },
    {
      "entropy": 1.853446225821972,
      "epoch": 0.03868690806201995,
      "grad_norm": 5.7373366355896,
      "learning_rate": 6.044270076528141e-06,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8060764700174332,
      "num_tokens": 14991962.0,
      "step": 12480
    },
    {
      "entropy": 1.792759819328785,
      "epoch": 0.03871790718706965,
      "grad_norm": 11.446489334106445,
      "learning_rate": 6.049113629758791e-06,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8205834746360778,
      "num_tokens": 15004797.0,
      "step": 12490
    },
    {
      "entropy": 1.8071394935250282,
      "epoch": 0.038748906312119344,
      "grad_norm": 10.317464828491211,
      "learning_rate": 6.053957182989442e-06,
      "loss": 0.5875,
      "mean_token_accuracy": 0.817318132519722,
      "num_tokens": 15017391.0,
      "step": 12500
    },
    {
      "entropy": 1.8499668538570404,
      "epoch": 0.038779905437169034,
      "grad_norm": 11.95541763305664,
      "learning_rate": 6.058800736220092e-06,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8172942087054252,
      "num_tokens": 15028762.0,
      "step": 12510
    },
    {
      "entropy": 1.8761356472969055,
      "epoch": 0.03881090456221873,
      "grad_norm": 11.275162696838379,
      "learning_rate": 6.063644289450742e-06,
      "loss": 0.615,
      "mean_token_accuracy": 0.8089762479066849,
      "num_tokens": 15040328.0,
      "step": 12520
    },
    {
      "entropy": 1.8349191531538964,
      "epoch": 0.03884190368726843,
      "grad_norm": 12.247875213623047,
      "learning_rate": 6.0684878426813924e-06,
      "loss": 0.628,
      "mean_token_accuracy": 0.8030536040663719,
      "num_tokens": 15052829.0,
      "step": 12530
    },
    {
      "entropy": 1.8827327758073806,
      "epoch": 0.03887290281231812,
      "grad_norm": 13.34996509552002,
      "learning_rate": 6.073331395912041e-06,
      "loss": 0.5946,
      "mean_token_accuracy": 0.8162320896983146,
      "num_tokens": 15064110.0,
      "step": 12540
    },
    {
      "entropy": 1.825656296312809,
      "epoch": 0.03890390193736782,
      "grad_norm": 9.6633939743042,
      "learning_rate": 6.078174949142691e-06,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8281683087348938,
      "num_tokens": 15076335.0,
      "step": 12550
    },
    {
      "entropy": 1.8427977129817008,
      "epoch": 0.038934901062417517,
      "grad_norm": 7.553597927093506,
      "learning_rate": 6.083018502373341e-06,
      "loss": 0.6407,
      "mean_token_accuracy": 0.8068282768130303,
      "num_tokens": 15088344.0,
      "step": 12560
    },
    {
      "entropy": 1.7809947788715363,
      "epoch": 0.038965900187467206,
      "grad_norm": 5.229321002960205,
      "learning_rate": 6.087862055603991e-06,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8179076954722404,
      "num_tokens": 15101575.0,
      "step": 12570
    },
    {
      "entropy": 1.9495843350887299,
      "epoch": 0.0389968993125169,
      "grad_norm": 10.395573616027832,
      "learning_rate": 6.092705608834642e-06,
      "loss": 0.6253,
      "mean_token_accuracy": 0.8120292127132416,
      "num_tokens": 15112551.0,
      "step": 12580
    },
    {
      "entropy": 1.8639352589845657,
      "epoch": 0.0390278984375666,
      "grad_norm": 14.738471984863281,
      "learning_rate": 6.097549162065292e-06,
      "loss": 0.6255,
      "mean_token_accuracy": 0.8092112436890602,
      "num_tokens": 15124473.0,
      "step": 12590
    },
    {
      "entropy": 1.8332767322659493,
      "epoch": 0.039058897562616296,
      "grad_norm": 6.293068885803223,
      "learning_rate": 6.102392715295942e-06,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8051409214735031,
      "num_tokens": 15136268.0,
      "step": 12600
    },
    {
      "entropy": 1.8370594762265682,
      "epoch": 0.03908989668766599,
      "grad_norm": 11.538333892822266,
      "learning_rate": 6.1072362685265915e-06,
      "loss": 0.5955,
      "mean_token_accuracy": 0.8151108309626579,
      "num_tokens": 15147270.0,
      "step": 12610
    },
    {
      "entropy": 1.8287363216280936,
      "epoch": 0.03912089581271569,
      "grad_norm": 12.497879981994629,
      "learning_rate": 6.112079821757242e-06,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8114735826849937,
      "num_tokens": 15159305.0,
      "step": 12620
    },
    {
      "entropy": 1.841526921093464,
      "epoch": 0.03915189493776538,
      "grad_norm": 15.22536563873291,
      "learning_rate": 6.116923374987892e-06,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8125358670949936,
      "num_tokens": 15170956.0,
      "step": 12630
    },
    {
      "entropy": 1.8685049675405025,
      "epoch": 0.039182894062815075,
      "grad_norm": 10.67180347442627,
      "learning_rate": 6.121766928218542e-06,
      "loss": 0.6239,
      "mean_token_accuracy": 0.8018307209014892,
      "num_tokens": 15183129.0,
      "step": 12640
    },
    {
      "entropy": 1.8348088413476944,
      "epoch": 0.03921389318786477,
      "grad_norm": 6.374260902404785,
      "learning_rate": 6.126610481449191e-06,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8142863139510155,
      "num_tokens": 15195411.0,
      "step": 12650
    },
    {
      "entropy": 1.9104970484972,
      "epoch": 0.03924489231291447,
      "grad_norm": 11.84320068359375,
      "learning_rate": 6.131454034679841e-06,
      "loss": 0.7097,
      "mean_token_accuracy": 0.7972751423716545,
      "num_tokens": 15206633.0,
      "step": 12660
    },
    {
      "entropy": 1.8191149190068245,
      "epoch": 0.039275891437964165,
      "grad_norm": 14.073987007141113,
      "learning_rate": 6.136297587910492e-06,
      "loss": 0.6266,
      "mean_token_accuracy": 0.8092659756541252,
      "num_tokens": 15218732.0,
      "step": 12670
    },
    {
      "entropy": 1.839244581758976,
      "epoch": 0.03930689056301386,
      "grad_norm": 9.864472389221191,
      "learning_rate": 6.1411411411411415e-06,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8092280417680741,
      "num_tokens": 15231394.0,
      "step": 12680
    },
    {
      "entropy": 1.8042413115501403,
      "epoch": 0.03933788968806355,
      "grad_norm": 12.334712982177734,
      "learning_rate": 6.145984694371791e-06,
      "loss": 0.6022,
      "mean_token_accuracy": 0.8171849220991134,
      "num_tokens": 15242791.0,
      "step": 12690
    },
    {
      "entropy": 1.8627389699220658,
      "epoch": 0.03936888881311325,
      "grad_norm": 13.441896438598633,
      "learning_rate": 6.150828247602441e-06,
      "loss": 0.5941,
      "mean_token_accuracy": 0.8153049975633622,
      "num_tokens": 15254161.0,
      "step": 12700
    },
    {
      "entropy": 1.8331183463335037,
      "epoch": 0.039399887938162945,
      "grad_norm": 12.164008140563965,
      "learning_rate": 6.155671800833092e-06,
      "loss": 0.5919,
      "mean_token_accuracy": 0.8250399813055992,
      "num_tokens": 15265756.0,
      "step": 12710
    },
    {
      "entropy": 1.8537740007042884,
      "epoch": 0.03943088706321264,
      "grad_norm": 11.0955810546875,
      "learning_rate": 6.160515354063742e-06,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8286507219076157,
      "num_tokens": 15277513.0,
      "step": 12720
    },
    {
      "entropy": 1.8841774478554725,
      "epoch": 0.03946188618826234,
      "grad_norm": 6.4962663650512695,
      "learning_rate": 6.165358907294392e-06,
      "loss": 0.6102,
      "mean_token_accuracy": 0.8118945837020874,
      "num_tokens": 15288556.0,
      "step": 12730
    },
    {
      "entropy": 1.8026639834046363,
      "epoch": 0.039492885313312034,
      "grad_norm": 12.76574420928955,
      "learning_rate": 6.1702024605250425e-06,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8168751522898674,
      "num_tokens": 15301444.0,
      "step": 12740
    },
    {
      "entropy": 1.8253360256552695,
      "epoch": 0.039523884438361724,
      "grad_norm": 5.973870754241943,
      "learning_rate": 6.175046013755691e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8278811991214752,
      "num_tokens": 15313764.0,
      "step": 12750
    },
    {
      "entropy": 1.7969791740179062,
      "epoch": 0.03955488356341142,
      "grad_norm": 14.835697174072266,
      "learning_rate": 6.1798895669863415e-06,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8214199602603912,
      "num_tokens": 15326494.0,
      "step": 12760
    },
    {
      "entropy": 1.8358357205986977,
      "epoch": 0.03958588268846112,
      "grad_norm": 6.574784278869629,
      "learning_rate": 6.184733120216991e-06,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8155452191829682,
      "num_tokens": 15338454.0,
      "step": 12770
    },
    {
      "entropy": 1.7818957820534707,
      "epoch": 0.039616881813510814,
      "grad_norm": 5.757773399353027,
      "learning_rate": 6.189576673447641e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8222944095730782,
      "num_tokens": 15351199.0,
      "step": 12780
    },
    {
      "entropy": 1.7962487503886222,
      "epoch": 0.03964788093856051,
      "grad_norm": 11.691226959228516,
      "learning_rate": 6.194420226678292e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8140451073646545,
      "num_tokens": 15363661.0,
      "step": 12790
    },
    {
      "entropy": 1.869248776137829,
      "epoch": 0.03967888006361021,
      "grad_norm": 13.061457633972168,
      "learning_rate": 6.199263779908942e-06,
      "loss": 0.6036,
      "mean_token_accuracy": 0.8111810341477395,
      "num_tokens": 15375191.0,
      "step": 12800
    },
    {
      "entropy": 1.9534028589725494,
      "epoch": 0.0397098791886599,
      "grad_norm": 12.596964836120605,
      "learning_rate": 6.204107333139592e-06,
      "loss": 0.7028,
      "mean_token_accuracy": 0.7962926715612412,
      "num_tokens": 15386223.0,
      "step": 12810
    },
    {
      "entropy": 1.8273673072457313,
      "epoch": 0.03974087831370959,
      "grad_norm": 6.692715644836426,
      "learning_rate": 6.208950886370242e-06,
      "loss": 0.6008,
      "mean_token_accuracy": 0.8225413948297501,
      "num_tokens": 15398818.0,
      "step": 12820
    },
    {
      "entropy": 1.8166001297533512,
      "epoch": 0.03977187743875929,
      "grad_norm": 13.053438186645508,
      "learning_rate": 6.213794439600892e-06,
      "loss": 0.5571,
      "mean_token_accuracy": 0.825442411005497,
      "num_tokens": 15411042.0,
      "step": 12830
    },
    {
      "entropy": 1.8038362354040145,
      "epoch": 0.03980287656380899,
      "grad_norm": 14.547704696655273,
      "learning_rate": 6.218637992831542e-06,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8112800106406212,
      "num_tokens": 15423517.0,
      "step": 12840
    },
    {
      "entropy": 1.8522241950035094,
      "epoch": 0.03983387568885868,
      "grad_norm": 13.58086109161377,
      "learning_rate": 6.223481546062192e-06,
      "loss": 0.6283,
      "mean_token_accuracy": 0.8121964901685714,
      "num_tokens": 15434996.0,
      "step": 12850
    },
    {
      "entropy": 1.8666557878255845,
      "epoch": 0.03986487481390838,
      "grad_norm": 12.256643295288086,
      "learning_rate": 6.228325099292841e-06,
      "loss": 0.6245,
      "mean_token_accuracy": 0.7993756279349327,
      "num_tokens": 15446746.0,
      "step": 12860
    },
    {
      "entropy": 1.8094900250434875,
      "epoch": 0.039895873938958076,
      "grad_norm": 12.334588050842285,
      "learning_rate": 6.233168652523491e-06,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8260209634900093,
      "num_tokens": 15459054.0,
      "step": 12870
    },
    {
      "entropy": 1.8770345002412796,
      "epoch": 0.039926873064007766,
      "grad_norm": 12.718692779541016,
      "learning_rate": 6.238012205754142e-06,
      "loss": 0.6855,
      "mean_token_accuracy": 0.7952335923910141,
      "num_tokens": 15469948.0,
      "step": 12880
    },
    {
      "entropy": 1.8727562785148621,
      "epoch": 0.03995787218905746,
      "grad_norm": 9.794669151306152,
      "learning_rate": 6.242855758984792e-06,
      "loss": 0.6304,
      "mean_token_accuracy": 0.816286937892437,
      "num_tokens": 15482541.0,
      "step": 12890
    },
    {
      "entropy": 1.8690064042806624,
      "epoch": 0.03998887131410716,
      "grad_norm": 11.124152183532715,
      "learning_rate": 6.2476993122154415e-06,
      "loss": 0.589,
      "mean_token_accuracy": 0.8171432822942734,
      "num_tokens": 15494339.0,
      "step": 12900
    },
    {
      "entropy": 1.815145094692707,
      "epoch": 0.040019870439156856,
      "grad_norm": 11.725676536560059,
      "learning_rate": 6.252542865446091e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.822598272562027,
      "num_tokens": 15506838.0,
      "step": 12910
    },
    {
      "entropy": 1.7810951352119446,
      "epoch": 0.04005086956420655,
      "grad_norm": 6.2188520431518555,
      "learning_rate": 6.257386418676742e-06,
      "loss": 0.5694,
      "mean_token_accuracy": 0.8137658298015594,
      "num_tokens": 15520324.0,
      "step": 12920
    },
    {
      "entropy": 1.8406692005693912,
      "epoch": 0.04008186868925625,
      "grad_norm": 10.676637649536133,
      "learning_rate": 6.262229971907392e-06,
      "loss": 0.5757,
      "mean_token_accuracy": 0.8103611707687378,
      "num_tokens": 15533133.0,
      "step": 12930
    },
    {
      "entropy": 1.9034065037965775,
      "epoch": 0.04011286781430594,
      "grad_norm": 12.021625518798828,
      "learning_rate": 6.267073525138042e-06,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8219456240534783,
      "num_tokens": 15543784.0,
      "step": 12940
    },
    {
      "entropy": 1.859899564087391,
      "epoch": 0.040143866939355635,
      "grad_norm": 5.559539794921875,
      "learning_rate": 6.271917078368693e-06,
      "loss": 0.6117,
      "mean_token_accuracy": 0.8060070484876632,
      "num_tokens": 15555512.0,
      "step": 12950
    },
    {
      "entropy": 1.8671554252505302,
      "epoch": 0.04017486606440533,
      "grad_norm": 4.944991111755371,
      "learning_rate": 6.276760631599341e-06,
      "loss": 0.6074,
      "mean_token_accuracy": 0.8140358075499534,
      "num_tokens": 15567319.0,
      "step": 12960
    },
    {
      "entropy": 1.8478776171803475,
      "epoch": 0.04020586518945503,
      "grad_norm": 6.253074645996094,
      "learning_rate": 6.2816041848299916e-06,
      "loss": 0.6098,
      "mean_token_accuracy": 0.8053146034479142,
      "num_tokens": 15579682.0,
      "step": 12970
    },
    {
      "entropy": 1.9027994275093079,
      "epoch": 0.040236864314504725,
      "grad_norm": 11.593799591064453,
      "learning_rate": 6.2864477380606414e-06,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8165087446570396,
      "num_tokens": 15590423.0,
      "step": 12980
    },
    {
      "entropy": 1.8880388498306275,
      "epoch": 0.04026786343955442,
      "grad_norm": 12.472247123718262,
      "learning_rate": 6.291291291291291e-06,
      "loss": 0.6243,
      "mean_token_accuracy": 0.8147116348147392,
      "num_tokens": 15601491.0,
      "step": 12990
    },
    {
      "entropy": 1.8754856497049333,
      "epoch": 0.04029886256460411,
      "grad_norm": 10.099672317504883,
      "learning_rate": 6.296134844521942e-06,
      "loss": 0.615,
      "mean_token_accuracy": 0.8218295946717262,
      "num_tokens": 15611664.0,
      "step": 13000
    },
    {
      "entropy": 1.8529302537441255,
      "epoch": 0.04032986168965381,
      "grad_norm": 11.330911636352539,
      "learning_rate": 6.300978397752592e-06,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8136036321520805,
      "num_tokens": 15623729.0,
      "step": 13010
    },
    {
      "entropy": 1.8294827699661256,
      "epoch": 0.040360860814703504,
      "grad_norm": 12.999713897705078,
      "learning_rate": 6.305821950983242e-06,
      "loss": 0.6562,
      "mean_token_accuracy": 0.8110322475433349,
      "num_tokens": 15634983.0,
      "step": 13020
    },
    {
      "entropy": 1.8562739863991737,
      "epoch": 0.0403918599397532,
      "grad_norm": 9.199260711669922,
      "learning_rate": 6.310665504213892e-06,
      "loss": 0.5929,
      "mean_token_accuracy": 0.8223294109106064,
      "num_tokens": 15646086.0,
      "step": 13030
    },
    {
      "entropy": 1.8410490170121192,
      "epoch": 0.0404228590648029,
      "grad_norm": 12.867775917053223,
      "learning_rate": 6.3155090574445424e-06,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8097691163420677,
      "num_tokens": 15658322.0,
      "step": 13040
    },
    {
      "entropy": 1.770823860168457,
      "epoch": 0.040453858189852594,
      "grad_norm": 5.460647106170654,
      "learning_rate": 6.320352610675192e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8246093928813935,
      "num_tokens": 15671766.0,
      "step": 13050
    },
    {
      "entropy": 1.856894339621067,
      "epoch": 0.040484857314902284,
      "grad_norm": 12.062054634094238,
      "learning_rate": 6.325196163905842e-06,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8271098077297211,
      "num_tokens": 15683812.0,
      "step": 13060
    },
    {
      "entropy": 1.8751189470291139,
      "epoch": 0.04051585643995198,
      "grad_norm": 6.459671974182129,
      "learning_rate": 6.330039717136491e-06,
      "loss": 0.6067,
      "mean_token_accuracy": 0.8083498194813729,
      "num_tokens": 15695486.0,
      "step": 13070
    },
    {
      "entropy": 1.8417337238788605,
      "epoch": 0.04054685556500168,
      "grad_norm": 6.831528663635254,
      "learning_rate": 6.334883270367141e-06,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8124615296721458,
      "num_tokens": 15707915.0,
      "step": 13080
    },
    {
      "entropy": 1.854557254910469,
      "epoch": 0.040577854690051374,
      "grad_norm": 10.786755561828613,
      "learning_rate": 6.339726823597792e-06,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8192732855677605,
      "num_tokens": 15720535.0,
      "step": 13090
    },
    {
      "entropy": 1.9224106088280677,
      "epoch": 0.04060885381510107,
      "grad_norm": 12.00683879852295,
      "learning_rate": 6.344570376828442e-06,
      "loss": 0.698,
      "mean_token_accuracy": 0.7983192473649978,
      "num_tokens": 15732872.0,
      "step": 13100
    },
    {
      "entropy": 1.8551457852125168,
      "epoch": 0.04063985294015077,
      "grad_norm": 7.283522605895996,
      "learning_rate": 6.349413930059092e-06,
      "loss": 0.6393,
      "mean_token_accuracy": 0.8065184399485588,
      "num_tokens": 15745391.0,
      "step": 13110
    },
    {
      "entropy": 1.8762536928057671,
      "epoch": 0.04067085206520046,
      "grad_norm": 5.411208629608154,
      "learning_rate": 6.3542574832897415e-06,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8252108618617058,
      "num_tokens": 15756896.0,
      "step": 13120
    },
    {
      "entropy": 1.9300520285964011,
      "epoch": 0.04070185119025015,
      "grad_norm": 11.322684288024902,
      "learning_rate": 6.359101036520392e-06,
      "loss": 0.6418,
      "mean_token_accuracy": 0.8126945987343788,
      "num_tokens": 15768386.0,
      "step": 13130
    },
    {
      "entropy": 1.7394558861851692,
      "epoch": 0.04073285031529985,
      "grad_norm": 3.4418604373931885,
      "learning_rate": 6.363944589751042e-06,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8297542706131935,
      "num_tokens": 15782338.0,
      "step": 13140
    },
    {
      "entropy": 1.886605440080166,
      "epoch": 0.040763849440349546,
      "grad_norm": 10.883418083190918,
      "learning_rate": 6.368788142981692e-06,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8118449002504349,
      "num_tokens": 15794126.0,
      "step": 13150
    },
    {
      "entropy": 1.833412842452526,
      "epoch": 0.04079484856539924,
      "grad_norm": 5.032846450805664,
      "learning_rate": 6.373631696212343e-06,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8236513167619706,
      "num_tokens": 15806365.0,
      "step": 13160
    },
    {
      "entropy": 1.8451514735817909,
      "epoch": 0.04082584769044894,
      "grad_norm": 12.876490592956543,
      "learning_rate": 6.378475249442991e-06,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8131675496697426,
      "num_tokens": 15817789.0,
      "step": 13170
    },
    {
      "entropy": 1.9239716470241546,
      "epoch": 0.04085684681549863,
      "grad_norm": 10.991585731506348,
      "learning_rate": 6.383318802673642e-06,
      "loss": 0.6288,
      "mean_token_accuracy": 0.8010996967554093,
      "num_tokens": 15828472.0,
      "step": 13180
    },
    {
      "entropy": 1.861264744400978,
      "epoch": 0.040887845940548326,
      "grad_norm": 12.282683372497559,
      "learning_rate": 6.3881623559042915e-06,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8243414610624313,
      "num_tokens": 15840805.0,
      "step": 13190
    },
    {
      "entropy": 1.7732505962252616,
      "epoch": 0.04091884506559802,
      "grad_norm": 10.902586936950684,
      "learning_rate": 6.3930059091349414e-06,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8143455445766449,
      "num_tokens": 15854669.0,
      "step": 13200
    },
    {
      "entropy": 1.8966855704784393,
      "epoch": 0.04094984419064772,
      "grad_norm": 15.284767150878906,
      "learning_rate": 6.397849462365592e-06,
      "loss": 0.6365,
      "mean_token_accuracy": 0.8180679067969322,
      "num_tokens": 15866135.0,
      "step": 13210
    },
    {
      "entropy": 1.8465805247426033,
      "epoch": 0.040980843315697416,
      "grad_norm": 10.624847412109375,
      "learning_rate": 6.402693015596242e-06,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8213504150509834,
      "num_tokens": 15878224.0,
      "step": 13220
    },
    {
      "entropy": 1.9168522462248803,
      "epoch": 0.04101184244074711,
      "grad_norm": 10.14426326751709,
      "learning_rate": 6.407536568826892e-06,
      "loss": 0.6252,
      "mean_token_accuracy": 0.8033525243401527,
      "num_tokens": 15891104.0,
      "step": 13230
    },
    {
      "entropy": 1.902791763842106,
      "epoch": 0.04104284156579681,
      "grad_norm": 12.496599197387695,
      "learning_rate": 6.412380122057542e-06,
      "loss": 0.5903,
      "mean_token_accuracy": 0.809710368514061,
      "num_tokens": 15903392.0,
      "step": 13240
    },
    {
      "entropy": 1.834274485707283,
      "epoch": 0.0410738406908465,
      "grad_norm": 10.14233684539795,
      "learning_rate": 6.4172236752881925e-06,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8176365941762924,
      "num_tokens": 15916354.0,
      "step": 13250
    },
    {
      "entropy": 1.7371825829148293,
      "epoch": 0.041104839815896195,
      "grad_norm": 4.078171730041504,
      "learning_rate": 6.4220672285188424e-06,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8276883006095886,
      "num_tokens": 15929845.0,
      "step": 13260
    },
    {
      "entropy": 1.906320759654045,
      "epoch": 0.04113583894094589,
      "grad_norm": 13.177295684814453,
      "learning_rate": 6.426910781749492e-06,
      "loss": 0.6577,
      "mean_token_accuracy": 0.803479115664959,
      "num_tokens": 15941475.0,
      "step": 13270
    },
    {
      "entropy": 1.86391938328743,
      "epoch": 0.04116683806599559,
      "grad_norm": 12.480244636535645,
      "learning_rate": 6.431754334980141e-06,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8072491884231567,
      "num_tokens": 15953798.0,
      "step": 13280
    },
    {
      "entropy": 1.8636515244841576,
      "epoch": 0.041197837191045285,
      "grad_norm": 11.828653335571289,
      "learning_rate": 6.436597888210791e-06,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8113080978393554,
      "num_tokens": 15965803.0,
      "step": 13290
    },
    {
      "entropy": 1.8319952994585038,
      "epoch": 0.04122883631609498,
      "grad_norm": 10.738685607910156,
      "learning_rate": 6.441441441441442e-06,
      "loss": 0.6163,
      "mean_token_accuracy": 0.8101550653576851,
      "num_tokens": 15978557.0,
      "step": 13300
    },
    {
      "entropy": 1.8741496190428735,
      "epoch": 0.04125983544114467,
      "grad_norm": 11.56949234008789,
      "learning_rate": 6.446284994672092e-06,
      "loss": 0.5879,
      "mean_token_accuracy": 0.811277537047863,
      "num_tokens": 15990542.0,
      "step": 13310
    },
    {
      "entropy": 1.9263570591807366,
      "epoch": 0.04129083456619437,
      "grad_norm": 5.565122127532959,
      "learning_rate": 6.451128547902742e-06,
      "loss": 0.6913,
      "mean_token_accuracy": 0.7948109433054924,
      "num_tokens": 16001976.0,
      "step": 13320
    },
    {
      "entropy": 1.9291939318180085,
      "epoch": 0.041321833691244064,
      "grad_norm": 11.844070434570312,
      "learning_rate": 6.455972101133392e-06,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8096133157610893,
      "num_tokens": 16012910.0,
      "step": 13330
    },
    {
      "entropy": 1.8042861357331277,
      "epoch": 0.04135283281629376,
      "grad_norm": 10.28139877319336,
      "learning_rate": 6.460815654364042e-06,
      "loss": 0.5329,
      "mean_token_accuracy": 0.825696873664856,
      "num_tokens": 16025765.0,
      "step": 13340
    },
    {
      "entropy": 1.9436326138675213,
      "epoch": 0.04138383194134346,
      "grad_norm": 12.31657886505127,
      "learning_rate": 6.465659207594692e-06,
      "loss": 0.6457,
      "mean_token_accuracy": 0.7972278758883476,
      "num_tokens": 16037425.0,
      "step": 13350
    },
    {
      "entropy": 1.8341301783919335,
      "epoch": 0.041414831066393154,
      "grad_norm": 11.382768630981445,
      "learning_rate": 6.470502760825342e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8359068840742111,
      "num_tokens": 16050108.0,
      "step": 13360
    },
    {
      "entropy": 1.9616862878203392,
      "epoch": 0.041445830191442844,
      "grad_norm": 11.222855567932129,
      "learning_rate": 6.475346314055993e-06,
      "loss": 0.6498,
      "mean_token_accuracy": 0.8021759241819382,
      "num_tokens": 16061548.0,
      "step": 13370
    },
    {
      "entropy": 1.9346465274691582,
      "epoch": 0.04147682931649254,
      "grad_norm": 10.253416061401367,
      "learning_rate": 6.480189867286641e-06,
      "loss": 0.6582,
      "mean_token_accuracy": 0.8080699786543846,
      "num_tokens": 16072805.0,
      "step": 13380
    },
    {
      "entropy": 1.8889733031392097,
      "epoch": 0.04150782844154224,
      "grad_norm": 10.119233131408691,
      "learning_rate": 6.485033420517292e-06,
      "loss": 0.6144,
      "mean_token_accuracy": 0.8134203433990479,
      "num_tokens": 16084811.0,
      "step": 13390
    },
    {
      "entropy": 1.935821218788624,
      "epoch": 0.041538827566591933,
      "grad_norm": 11.066147804260254,
      "learning_rate": 6.489876973747942e-06,
      "loss": 0.6034,
      "mean_token_accuracy": 0.8229247480630875,
      "num_tokens": 16095975.0,
      "step": 13400
    },
    {
      "entropy": 1.8281036272644997,
      "epoch": 0.04156982669164163,
      "grad_norm": 12.909806251525879,
      "learning_rate": 6.4947205269785915e-06,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8127323508262634,
      "num_tokens": 16108978.0,
      "step": 13410
    },
    {
      "entropy": 1.915538875758648,
      "epoch": 0.04160082581669133,
      "grad_norm": 11.429049491882324,
      "learning_rate": 6.499564080209242e-06,
      "loss": 0.6331,
      "mean_token_accuracy": 0.8143791824579238,
      "num_tokens": 16121160.0,
      "step": 13420
    },
    {
      "entropy": 1.8657421082258225,
      "epoch": 0.041631824941741016,
      "grad_norm": 5.312261581420898,
      "learning_rate": 6.504407633439892e-06,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8279088050127029,
      "num_tokens": 16133755.0,
      "step": 13430
    },
    {
      "entropy": 1.9448801666498183,
      "epoch": 0.04166282406679071,
      "grad_norm": 11.335988998413086,
      "learning_rate": 6.509251186670542e-06,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8046508118510246,
      "num_tokens": 16145359.0,
      "step": 13440
    },
    {
      "entropy": 1.837200105190277,
      "epoch": 0.04169382319184041,
      "grad_norm": 8.97873592376709,
      "learning_rate": 6.514094739901192e-06,
      "loss": 0.577,
      "mean_token_accuracy": 0.823949719965458,
      "num_tokens": 16157112.0,
      "step": 13450
    },
    {
      "entropy": 1.8169792860746383,
      "epoch": 0.041724822316890106,
      "grad_norm": 13.009424209594727,
      "learning_rate": 6.518938293131843e-06,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8228383541107178,
      "num_tokens": 16169936.0,
      "step": 13460
    },
    {
      "entropy": 1.7715341180562973,
      "epoch": 0.0417558214419398,
      "grad_norm": 4.483689308166504,
      "learning_rate": 6.5237818463624925e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8237165853381156,
      "num_tokens": 16184134.0,
      "step": 13470
    },
    {
      "entropy": 1.9055643543601035,
      "epoch": 0.0417868205669895,
      "grad_norm": 11.176976203918457,
      "learning_rate": 6.528625399593142e-06,
      "loss": 0.616,
      "mean_token_accuracy": 0.807232391834259,
      "num_tokens": 16195929.0,
      "step": 13480
    },
    {
      "entropy": 1.7436404943466186,
      "epoch": 0.04181781969203919,
      "grad_norm": 11.869129180908203,
      "learning_rate": 6.5334689528237915e-06,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8363062143325806,
      "num_tokens": 16210306.0,
      "step": 13490
    },
    {
      "entropy": 1.8251071318984031,
      "epoch": 0.041848818817088886,
      "grad_norm": 6.912018775939941,
      "learning_rate": 6.538312506054441e-06,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8134068369865417,
      "num_tokens": 16222376.0,
      "step": 13500
    },
    {
      "entropy": 1.8436161801218987,
      "epoch": 0.04187981794213858,
      "grad_norm": 10.873340606689453,
      "learning_rate": 6.543156059285092e-06,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8138954237103462,
      "num_tokens": 16235581.0,
      "step": 13510
    },
    {
      "entropy": 1.9201088815927505,
      "epoch": 0.04191081706718828,
      "grad_norm": 12.203166007995605,
      "learning_rate": 6.547999612515742e-06,
      "loss": 0.6399,
      "mean_token_accuracy": 0.809160690009594,
      "num_tokens": 16246598.0,
      "step": 13520
    },
    {
      "entropy": 1.9009541541337966,
      "epoch": 0.041941816192237975,
      "grad_norm": 12.666475296020508,
      "learning_rate": 6.552843165746392e-06,
      "loss": 0.6352,
      "mean_token_accuracy": 0.7980454340577126,
      "num_tokens": 16258072.0,
      "step": 13530
    },
    {
      "entropy": 1.8785468250513078,
      "epoch": 0.04197281531728767,
      "grad_norm": 10.72907829284668,
      "learning_rate": 6.557686718977042e-06,
      "loss": 0.6576,
      "mean_token_accuracy": 0.8070585578680038,
      "num_tokens": 16269003.0,
      "step": 13540
    },
    {
      "entropy": 1.7919475421309472,
      "epoch": 0.04200381444233736,
      "grad_norm": 12.568443298339844,
      "learning_rate": 6.5625302722076924e-06,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8193079605698586,
      "num_tokens": 16281900.0,
      "step": 13550
    },
    {
      "entropy": 1.865149575471878,
      "epoch": 0.04203481356738706,
      "grad_norm": 11.324207305908203,
      "learning_rate": 6.567373825438342e-06,
      "loss": 0.6203,
      "mean_token_accuracy": 0.810130500793457,
      "num_tokens": 16293834.0,
      "step": 13560
    },
    {
      "entropy": 1.9343768432736397,
      "epoch": 0.042065812692436755,
      "grad_norm": 13.249279975891113,
      "learning_rate": 6.572217378668992e-06,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8070745259523392,
      "num_tokens": 16305965.0,
      "step": 13570
    },
    {
      "entropy": 1.916192325949669,
      "epoch": 0.04209681181748645,
      "grad_norm": 12.691089630126953,
      "learning_rate": 6.577060931899643e-06,
      "loss": 0.5989,
      "mean_token_accuracy": 0.8105158194899559,
      "num_tokens": 16317096.0,
      "step": 13580
    },
    {
      "entropy": 1.888793683052063,
      "epoch": 0.04212781094253615,
      "grad_norm": 9.996535301208496,
      "learning_rate": 6.581904485130292e-06,
      "loss": 0.6723,
      "mean_token_accuracy": 0.8087509065866471,
      "num_tokens": 16329359.0,
      "step": 13590
    },
    {
      "entropy": 1.8468269050121306,
      "epoch": 0.042158810067585845,
      "grad_norm": 10.06303596496582,
      "learning_rate": 6.586748038360942e-06,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8199773550033569,
      "num_tokens": 16341773.0,
      "step": 13600
    },
    {
      "entropy": 1.8991917297244072,
      "epoch": 0.04218980919263554,
      "grad_norm": 3.874143362045288,
      "learning_rate": 6.591591591591592e-06,
      "loss": 0.6202,
      "mean_token_accuracy": 0.8165913313627243,
      "num_tokens": 16353377.0,
      "step": 13610
    },
    {
      "entropy": 1.9188966274261474,
      "epoch": 0.04222080831768523,
      "grad_norm": 11.466548919677734,
      "learning_rate": 6.596435144822242e-06,
      "loss": 0.6185,
      "mean_token_accuracy": 0.8085856437683105,
      "num_tokens": 16365134.0,
      "step": 13620
    },
    {
      "entropy": 1.9046609073877334,
      "epoch": 0.04225180744273493,
      "grad_norm": 10.0979585647583,
      "learning_rate": 6.601278698052892e-06,
      "loss": 0.5963,
      "mean_token_accuracy": 0.8179086208343506,
      "num_tokens": 16376299.0,
      "step": 13630
    },
    {
      "entropy": 1.828147941827774,
      "epoch": 0.042282806567784624,
      "grad_norm": 10.194396018981934,
      "learning_rate": 6.606122251283542e-06,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8191608220338822,
      "num_tokens": 16388768.0,
      "step": 13640
    },
    {
      "entropy": 1.8713184520602226,
      "epoch": 0.04231380569283432,
      "grad_norm": 12.538405418395996,
      "learning_rate": 6.610965804514192e-06,
      "loss": 0.6416,
      "mean_token_accuracy": 0.8056328803300857,
      "num_tokens": 16400797.0,
      "step": 13650
    },
    {
      "entropy": 1.8713348254561424,
      "epoch": 0.04234480481788402,
      "grad_norm": 12.18179988861084,
      "learning_rate": 6.615809357744842e-06,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8290207415819169,
      "num_tokens": 16412316.0,
      "step": 13660
    },
    {
      "entropy": 1.7908138126134872,
      "epoch": 0.042375803942933714,
      "grad_norm": 11.162075996398926,
      "learning_rate": 6.620652910975493e-06,
      "loss": 0.548,
      "mean_token_accuracy": 0.8253503978252411,
      "num_tokens": 16426142.0,
      "step": 13670
    },
    {
      "entropy": 1.8941476494073868,
      "epoch": 0.042406803067983403,
      "grad_norm": 15.437644958496094,
      "learning_rate": 6.625496464206143e-06,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8184331357479095,
      "num_tokens": 16438321.0,
      "step": 13680
    },
    {
      "entropy": 1.9231675088405609,
      "epoch": 0.0424378021930331,
      "grad_norm": 12.137004852294922,
      "learning_rate": 6.6303400174367925e-06,
      "loss": 0.6361,
      "mean_token_accuracy": 0.814939396083355,
      "num_tokens": 16449946.0,
      "step": 13690
    },
    {
      "entropy": 1.8700689136981965,
      "epoch": 0.0424688013180828,
      "grad_norm": 10.420080184936523,
      "learning_rate": 6.6351835706674416e-06,
      "loss": 0.5667,
      "mean_token_accuracy": 0.8213317602872848,
      "num_tokens": 16461823.0,
      "step": 13700
    },
    {
      "entropy": 1.9616285428404807,
      "epoch": 0.04249980044313249,
      "grad_norm": 11.999855995178223,
      "learning_rate": 6.6400271238980914e-06,
      "loss": 0.6274,
      "mean_token_accuracy": 0.8095705136656761,
      "num_tokens": 16473326.0,
      "step": 13710
    },
    {
      "entropy": 1.8622044518589973,
      "epoch": 0.04253079956818219,
      "grad_norm": 6.192050457000732,
      "learning_rate": 6.644870677128742e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8142581716179847,
      "num_tokens": 16486260.0,
      "step": 13720
    },
    {
      "entropy": 1.9137049853801726,
      "epoch": 0.042561798693231886,
      "grad_norm": 10.177970886230469,
      "learning_rate": 6.649714230359392e-06,
      "loss": 0.6307,
      "mean_token_accuracy": 0.8191399827599526,
      "num_tokens": 16497734.0,
      "step": 13730
    },
    {
      "entropy": 1.8745167449116706,
      "epoch": 0.042592797818281576,
      "grad_norm": 10.241938591003418,
      "learning_rate": 6.654557783590042e-06,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8165997639298439,
      "num_tokens": 16509288.0,
      "step": 13740
    },
    {
      "entropy": 1.93837161809206,
      "epoch": 0.04262379694333127,
      "grad_norm": 12.580158233642578,
      "learning_rate": 6.659401336820692e-06,
      "loss": 0.6656,
      "mean_token_accuracy": 0.7997888430953026,
      "num_tokens": 16520584.0,
      "step": 13750
    },
    {
      "entropy": 1.9007338181138038,
      "epoch": 0.04265479606838097,
      "grad_norm": 10.395405769348145,
      "learning_rate": 6.6642448900513426e-06,
      "loss": 0.6244,
      "mean_token_accuracy": 0.8126766815781593,
      "num_tokens": 16532328.0,
      "step": 13760
    },
    {
      "entropy": 1.9072567522525787,
      "epoch": 0.042685795193430666,
      "grad_norm": 12.813956260681152,
      "learning_rate": 6.6690884432819924e-06,
      "loss": 0.6346,
      "mean_token_accuracy": 0.81207554936409,
      "num_tokens": 16543530.0,
      "step": 13770
    },
    {
      "entropy": 1.9019085496664048,
      "epoch": 0.04271679431848036,
      "grad_norm": 12.529980659484863,
      "learning_rate": 6.673931996512642e-06,
      "loss": 0.66,
      "mean_token_accuracy": 0.8073429599404335,
      "num_tokens": 16554610.0,
      "step": 13780
    },
    {
      "entropy": 1.915639691054821,
      "epoch": 0.04274779344353006,
      "grad_norm": 6.579958438873291,
      "learning_rate": 6.678775549743293e-06,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8055771961808205,
      "num_tokens": 16566534.0,
      "step": 13790
    },
    {
      "entropy": 1.897673524916172,
      "epoch": 0.04277879256857975,
      "grad_norm": 13.530618667602539,
      "learning_rate": 6.683619102973943e-06,
      "loss": 0.5869,
      "mean_token_accuracy": 0.821336168050766,
      "num_tokens": 16577725.0,
      "step": 13800
    },
    {
      "entropy": 1.8759017661213875,
      "epoch": 0.042809791693629445,
      "grad_norm": 10.226716995239258,
      "learning_rate": 6.688462656204592e-06,
      "loss": 0.5829,
      "mean_token_accuracy": 0.823052391409874,
      "num_tokens": 16589628.0,
      "step": 13810
    },
    {
      "entropy": 1.8439259082078934,
      "epoch": 0.04284079081867914,
      "grad_norm": 10.295611381530762,
      "learning_rate": 6.693306209435242e-06,
      "loss": 0.581,
      "mean_token_accuracy": 0.8230698376893997,
      "num_tokens": 16601483.0,
      "step": 13820
    },
    {
      "entropy": 1.9123721539974212,
      "epoch": 0.04287178994372884,
      "grad_norm": 4.241229057312012,
      "learning_rate": 6.698149762665892e-06,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8144331857562065,
      "num_tokens": 16613580.0,
      "step": 13830
    },
    {
      "entropy": 1.8960613742470742,
      "epoch": 0.042902789068778535,
      "grad_norm": 6.314639091491699,
      "learning_rate": 6.7029933158965425e-06,
      "loss": 0.591,
      "mean_token_accuracy": 0.8104770466685295,
      "num_tokens": 16626132.0,
      "step": 13840
    },
    {
      "entropy": 1.9086899921298026,
      "epoch": 0.04293378819382823,
      "grad_norm": 5.750073432922363,
      "learning_rate": 6.707836869127192e-06,
      "loss": 0.6769,
      "mean_token_accuracy": 0.8074093982577324,
      "num_tokens": 16637715.0,
      "step": 13850
    },
    {
      "entropy": 1.9291119009256363,
      "epoch": 0.04296478731887792,
      "grad_norm": 10.997711181640625,
      "learning_rate": 6.712680422357842e-06,
      "loss": 0.64,
      "mean_token_accuracy": 0.8164422243833542,
      "num_tokens": 16648790.0,
      "step": 13860
    },
    {
      "entropy": 1.7718395471572876,
      "epoch": 0.04299578644392762,
      "grad_norm": 5.029774188995361,
      "learning_rate": 6.717523975588492e-06,
      "loss": 0.5216,
      "mean_token_accuracy": 0.824312150478363,
      "num_tokens": 16661991.0,
      "step": 13870
    },
    {
      "entropy": 1.897067406773567,
      "epoch": 0.043026785568977315,
      "grad_norm": 13.85363483428955,
      "learning_rate": 6.722367528819143e-06,
      "loss": 0.6711,
      "mean_token_accuracy": 0.8090870007872581,
      "num_tokens": 16672565.0,
      "step": 13880
    },
    {
      "entropy": 1.8265616819262505,
      "epoch": 0.04305778469402701,
      "grad_norm": 13.524333000183105,
      "learning_rate": 6.727211082049793e-06,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8150227144360542,
      "num_tokens": 16685456.0,
      "step": 13890
    },
    {
      "entropy": 1.8813450008630752,
      "epoch": 0.04308878381907671,
      "grad_norm": 12.762773513793945,
      "learning_rate": 6.732054635280443e-06,
      "loss": 0.5886,
      "mean_token_accuracy": 0.8163325443863869,
      "num_tokens": 16697034.0,
      "step": 13900
    },
    {
      "entropy": 1.8345113858580588,
      "epoch": 0.043119782944126404,
      "grad_norm": 12.19513988494873,
      "learning_rate": 6.736898188511092e-06,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8157405987381935,
      "num_tokens": 16708871.0,
      "step": 13910
    },
    {
      "entropy": 1.8650273010134697,
      "epoch": 0.043150782069176094,
      "grad_norm": 12.387640953063965,
      "learning_rate": 6.7417417417417415e-06,
      "loss": 0.6953,
      "mean_token_accuracy": 0.8010308906435967,
      "num_tokens": 16720631.0,
      "step": 13920
    },
    {
      "entropy": 1.8238304048776626,
      "epoch": 0.04318178119422579,
      "grad_norm": 10.031027793884277,
      "learning_rate": 6.746585294972392e-06,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8242588087916374,
      "num_tokens": 16733234.0,
      "step": 13930
    },
    {
      "entropy": 1.8230559036135674,
      "epoch": 0.04321278031927549,
      "grad_norm": 12.307608604431152,
      "learning_rate": 6.751428848203042e-06,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8173372864723205,
      "num_tokens": 16745599.0,
      "step": 13940
    },
    {
      "entropy": 1.9341980874538423,
      "epoch": 0.043243779444325184,
      "grad_norm": 12.861936569213867,
      "learning_rate": 6.756272401433692e-06,
      "loss": 0.6583,
      "mean_token_accuracy": 0.8072870954871177,
      "num_tokens": 16756209.0,
      "step": 13950
    },
    {
      "entropy": 1.8573587149381638,
      "epoch": 0.04327477856937488,
      "grad_norm": 11.555964469909668,
      "learning_rate": 6.761115954664342e-06,
      "loss": 0.5883,
      "mean_token_accuracy": 0.8267058104276657,
      "num_tokens": 16767433.0,
      "step": 13960
    },
    {
      "entropy": 1.8955969214439392,
      "epoch": 0.04330577769442458,
      "grad_norm": 11.410296440124512,
      "learning_rate": 6.765959507894993e-06,
      "loss": 0.5956,
      "mean_token_accuracy": 0.8207617849111557,
      "num_tokens": 16778097.0,
      "step": 13970
    },
    {
      "entropy": 1.876904509961605,
      "epoch": 0.043336776819474274,
      "grad_norm": 10.337623596191406,
      "learning_rate": 6.7708030611256425e-06,
      "loss": 0.636,
      "mean_token_accuracy": 0.8040330380201339,
      "num_tokens": 16789800.0,
      "step": 13980
    },
    {
      "entropy": 1.9401722326874733,
      "epoch": 0.04336777594452396,
      "grad_norm": 11.589249610900879,
      "learning_rate": 6.7756466143562924e-06,
      "loss": 0.6594,
      "mean_token_accuracy": 0.8035330668091774,
      "num_tokens": 16801605.0,
      "step": 13990
    },
    {
      "entropy": 1.890137755870819,
      "epoch": 0.04339877506957366,
      "grad_norm": 10.558548927307129,
      "learning_rate": 6.780490167586943e-06,
      "loss": 0.6248,
      "mean_token_accuracy": 0.8102419808506965,
      "num_tokens": 16813740.0,
      "step": 14000
    },
    {
      "entropy": 1.888036273419857,
      "epoch": 0.043429774194623356,
      "grad_norm": 6.69431734085083,
      "learning_rate": 6.785333720817593e-06,
      "loss": 0.6036,
      "mean_token_accuracy": 0.81003537774086,
      "num_tokens": 16826251.0,
      "step": 14010
    },
    {
      "entropy": 1.949825246632099,
      "epoch": 0.04346077331967305,
      "grad_norm": 11.190848350524902,
      "learning_rate": 6.790177274048242e-06,
      "loss": 0.6626,
      "mean_token_accuracy": 0.8077096566557884,
      "num_tokens": 16837543.0,
      "step": 14020
    },
    {
      "entropy": 1.8271329566836356,
      "epoch": 0.04349177244472275,
      "grad_norm": 5.520564556121826,
      "learning_rate": 6.795020827278892e-06,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8192524507641792,
      "num_tokens": 16850237.0,
      "step": 14030
    },
    {
      "entropy": 1.8742077186703683,
      "epoch": 0.043522771569772446,
      "grad_norm": 12.861546516418457,
      "learning_rate": 6.799864380509542e-06,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8112540423870087,
      "num_tokens": 16861883.0,
      "step": 14040
    },
    {
      "entropy": 1.813956792652607,
      "epoch": 0.043553770694822136,
      "grad_norm": 13.800911903381348,
      "learning_rate": 6.8047079337401926e-06,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8147607937455177,
      "num_tokens": 16874756.0,
      "step": 14050
    },
    {
      "entropy": 1.9326902404427528,
      "epoch": 0.04358476981987183,
      "grad_norm": 11.105195045471191,
      "learning_rate": 6.8095514869708425e-06,
      "loss": 0.6847,
      "mean_token_accuracy": 0.793427674472332,
      "num_tokens": 16885758.0,
      "step": 14060
    },
    {
      "entropy": 1.9657369270920753,
      "epoch": 0.04361576894492153,
      "grad_norm": 11.325508117675781,
      "learning_rate": 6.814395040201492e-06,
      "loss": 0.6517,
      "mean_token_accuracy": 0.7974928095936775,
      "num_tokens": 16896989.0,
      "step": 14070
    },
    {
      "entropy": 1.7869946122169496,
      "epoch": 0.043646768069971226,
      "grad_norm": 10.723480224609375,
      "learning_rate": 6.819238593432142e-06,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8241583168506622,
      "num_tokens": 16910317.0,
      "step": 14080
    },
    {
      "entropy": 1.886980764567852,
      "epoch": 0.04367776719502092,
      "grad_norm": 10.611824989318848,
      "learning_rate": 6.824082146662793e-06,
      "loss": 0.5972,
      "mean_token_accuracy": 0.821302755177021,
      "num_tokens": 16921546.0,
      "step": 14090
    },
    {
      "entropy": 1.795514563471079,
      "epoch": 0.04370876632007062,
      "grad_norm": 3.9574859142303467,
      "learning_rate": 6.828925699893443e-06,
      "loss": 0.5667,
      "mean_token_accuracy": 0.8096185430884362,
      "num_tokens": 16934810.0,
      "step": 14100
    },
    {
      "entropy": 1.8915703102946282,
      "epoch": 0.04373976544512031,
      "grad_norm": 12.432307243347168,
      "learning_rate": 6.833769253124093e-06,
      "loss": 0.6694,
      "mean_token_accuracy": 0.7994834899902343,
      "num_tokens": 16946379.0,
      "step": 14110
    },
    {
      "entropy": 1.872760045528412,
      "epoch": 0.043770764570170005,
      "grad_norm": 12.616002082824707,
      "learning_rate": 6.838612806354742e-06,
      "loss": 0.6253,
      "mean_token_accuracy": 0.8020053207874298,
      "num_tokens": 16958537.0,
      "step": 14120
    },
    {
      "entropy": 1.8190848156809807,
      "epoch": 0.0438017636952197,
      "grad_norm": 6.749497890472412,
      "learning_rate": 6.843456359585392e-06,
      "loss": 0.566,
      "mean_token_accuracy": 0.8224000081419944,
      "num_tokens": 16971281.0,
      "step": 14130
    },
    {
      "entropy": 1.7950861573219299,
      "epoch": 0.0438327628202694,
      "grad_norm": 5.5427021980285645,
      "learning_rate": 6.848299912816042e-06,
      "loss": 0.523,
      "mean_token_accuracy": 0.821921581029892,
      "num_tokens": 16983378.0,
      "step": 14140
    },
    {
      "entropy": 1.7081642150878906,
      "epoch": 0.043863761945319095,
      "grad_norm": 14.56338882446289,
      "learning_rate": 6.853143466046692e-06,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8312652423977852,
      "num_tokens": 16996461.0,
      "step": 14150
    },
    {
      "entropy": 1.7555221557617187,
      "epoch": 0.04389476107036879,
      "grad_norm": 5.769707679748535,
      "learning_rate": 6.857987019277342e-06,
      "loss": 0.5165,
      "mean_token_accuracy": 0.81901466101408,
      "num_tokens": 17009223.0,
      "step": 14160
    },
    {
      "entropy": 1.8228839874267577,
      "epoch": 0.04392576019541848,
      "grad_norm": 14.009598731994629,
      "learning_rate": 6.862830572507992e-06,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8186228111386299,
      "num_tokens": 17021369.0,
      "step": 14170
    },
    {
      "entropy": 1.8004897370934487,
      "epoch": 0.04395675932046818,
      "grad_norm": 13.954216957092285,
      "learning_rate": 6.867674125738643e-06,
      "loss": 0.6168,
      "mean_token_accuracy": 0.8179115906357766,
      "num_tokens": 17033773.0,
      "step": 14180
    },
    {
      "entropy": 1.8727715358138084,
      "epoch": 0.043987758445517874,
      "grad_norm": 5.836693286895752,
      "learning_rate": 6.872517678969293e-06,
      "loss": 0.63,
      "mean_token_accuracy": 0.8152304857969284,
      "num_tokens": 17045247.0,
      "step": 14190
    },
    {
      "entropy": 1.9000848352909088,
      "epoch": 0.04401875757056757,
      "grad_norm": 13.797229766845703,
      "learning_rate": 6.8773612321999425e-06,
      "loss": 0.6389,
      "mean_token_accuracy": 0.8070854544639587,
      "num_tokens": 17056322.0,
      "step": 14200
    },
    {
      "entropy": 1.8419507443904877,
      "epoch": 0.04404975669561727,
      "grad_norm": 10.07559585571289,
      "learning_rate": 6.882204785430593e-06,
      "loss": 0.607,
      "mean_token_accuracy": 0.8120106473565102,
      "num_tokens": 17068663.0,
      "step": 14210
    },
    {
      "entropy": 1.8738445043563843,
      "epoch": 0.044080755820666964,
      "grad_norm": 12.083559036254883,
      "learning_rate": 6.887048338661243e-06,
      "loss": 0.597,
      "mean_token_accuracy": 0.8096369743347168,
      "num_tokens": 17080468.0,
      "step": 14220
    },
    {
      "entropy": 1.8636516377329826,
      "epoch": 0.044111754945716654,
      "grad_norm": 10.079278945922852,
      "learning_rate": 6.891891891891892e-06,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8174061790108681,
      "num_tokens": 17091833.0,
      "step": 14230
    },
    {
      "entropy": 1.8882641837000846,
      "epoch": 0.04414275407076635,
      "grad_norm": 13.434144973754883,
      "learning_rate": 6.896735445122542e-06,
      "loss": 0.6377,
      "mean_token_accuracy": 0.8117697656154632,
      "num_tokens": 17102951.0,
      "step": 14240
    },
    {
      "entropy": 1.9334753528237343,
      "epoch": 0.04417375319581605,
      "grad_norm": 9.094237327575684,
      "learning_rate": 6.901578998353192e-06,
      "loss": 0.6594,
      "mean_token_accuracy": 0.8049651876091957,
      "num_tokens": 17114582.0,
      "step": 14250
    },
    {
      "entropy": 1.8502332031726838,
      "epoch": 0.044204752320865744,
      "grad_norm": 9.20546817779541,
      "learning_rate": 6.906422551583843e-06,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8162536874413491,
      "num_tokens": 17127285.0,
      "step": 14260
    },
    {
      "entropy": 1.8943143799901008,
      "epoch": 0.04423575144591544,
      "grad_norm": 10.137446403503418,
      "learning_rate": 6.9112661048144926e-06,
      "loss": 0.6203,
      "mean_token_accuracy": 0.8087219893932343,
      "num_tokens": 17139089.0,
      "step": 14270
    },
    {
      "entropy": 1.887183803319931,
      "epoch": 0.04426675057096514,
      "grad_norm": 10.977912902832031,
      "learning_rate": 6.9161096580451424e-06,
      "loss": 0.6091,
      "mean_token_accuracy": 0.8069622635841369,
      "num_tokens": 17151186.0,
      "step": 14280
    },
    {
      "entropy": 1.8306689888238907,
      "epoch": 0.04429774969601483,
      "grad_norm": 11.67708969116211,
      "learning_rate": 6.920953211275792e-06,
      "loss": 0.567,
      "mean_token_accuracy": 0.8185241803526878,
      "num_tokens": 17162472.0,
      "step": 14290
    },
    {
      "entropy": 1.9653134107589723,
      "epoch": 0.04432874882106452,
      "grad_norm": 12.62784481048584,
      "learning_rate": 6.925796764506443e-06,
      "loss": 0.7129,
      "mean_token_accuracy": 0.7954209297895432,
      "num_tokens": 17173491.0,
      "step": 14300
    },
    {
      "entropy": 1.8972666263580322,
      "epoch": 0.04435974794611422,
      "grad_norm": 10.280572891235352,
      "learning_rate": 6.930640317737093e-06,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8118947297334671,
      "num_tokens": 17184219.0,
      "step": 14310
    },
    {
      "entropy": 1.8529438108205796,
      "epoch": 0.044390747071163916,
      "grad_norm": 13.168403625488281,
      "learning_rate": 6.935483870967743e-06,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8159810289740562,
      "num_tokens": 17195740.0,
      "step": 14320
    },
    {
      "entropy": 1.8751512482762336,
      "epoch": 0.04442174619621361,
      "grad_norm": 5.939009666442871,
      "learning_rate": 6.940327424198392e-06,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8138178676366806,
      "num_tokens": 17207488.0,
      "step": 14330
    },
    {
      "entropy": 1.9205424144864083,
      "epoch": 0.04445274532126331,
      "grad_norm": 9.95887565612793,
      "learning_rate": 6.945170977429042e-06,
      "loss": 0.6255,
      "mean_token_accuracy": 0.8147617742419243,
      "num_tokens": 17218552.0,
      "step": 14340
    },
    {
      "entropy": 1.9514837980270385,
      "epoch": 0.044483744446313006,
      "grad_norm": 11.541547775268555,
      "learning_rate": 6.9500145306596925e-06,
      "loss": 0.7025,
      "mean_token_accuracy": 0.7957576259970665,
      "num_tokens": 17229630.0,
      "step": 14350
    },
    {
      "entropy": 1.9413245290517807,
      "epoch": 0.044514743571362696,
      "grad_norm": 11.1705322265625,
      "learning_rate": 6.954858083890342e-06,
      "loss": 0.6673,
      "mean_token_accuracy": 0.8048760443925858,
      "num_tokens": 17240633.0,
      "step": 14360
    },
    {
      "entropy": 1.8450009673833847,
      "epoch": 0.04454574269641239,
      "grad_norm": 6.041028022766113,
      "learning_rate": 6.959701637120992e-06,
      "loss": 0.6093,
      "mean_token_accuracy": 0.8054246112704277,
      "num_tokens": 17253698.0,
      "step": 14370
    },
    {
      "entropy": 1.833503720164299,
      "epoch": 0.04457674182146209,
      "grad_norm": 4.289499759674072,
      "learning_rate": 6.964545190351643e-06,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8216146498918533,
      "num_tokens": 17266350.0,
      "step": 14380
    },
    {
      "entropy": 1.7655683636665345,
      "epoch": 0.044607740946511786,
      "grad_norm": 12.362275123596191,
      "learning_rate": 6.969388743582293e-06,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8274104654788971,
      "num_tokens": 17279397.0,
      "step": 14390
    },
    {
      "entropy": 1.8648739516735078,
      "epoch": 0.04463874007156148,
      "grad_norm": 11.831520080566406,
      "learning_rate": 6.974232296812943e-06,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8219448134303093,
      "num_tokens": 17291500.0,
      "step": 14400
    },
    {
      "entropy": 1.8236894220113755,
      "epoch": 0.04466973919661118,
      "grad_norm": 10.356793403625488,
      "learning_rate": 6.979075850043593e-06,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8268671408295631,
      "num_tokens": 17304068.0,
      "step": 14410
    },
    {
      "entropy": 1.8797018930315972,
      "epoch": 0.04470073832166087,
      "grad_norm": 9.828136444091797,
      "learning_rate": 6.983919403274243e-06,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8069930672645569,
      "num_tokens": 17316443.0,
      "step": 14420
    },
    {
      "entropy": 1.9420302003622054,
      "epoch": 0.044731737446710565,
      "grad_norm": 12.722853660583496,
      "learning_rate": 6.988762956504893e-06,
      "loss": 0.6601,
      "mean_token_accuracy": 0.8111462906002999,
      "num_tokens": 17326906.0,
      "step": 14430
    },
    {
      "entropy": 1.8793850436806678,
      "epoch": 0.04476273657176026,
      "grad_norm": 13.556890487670898,
      "learning_rate": 6.993606509735542e-06,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8150814548134804,
      "num_tokens": 17338259.0,
      "step": 14440
    },
    {
      "entropy": 1.8735877990722656,
      "epoch": 0.04479373569680996,
      "grad_norm": 10.880188941955566,
      "learning_rate": 6.998450062966192e-06,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8156931459903717,
      "num_tokens": 17351172.0,
      "step": 14450
    },
    {
      "entropy": 1.8445069909095764,
      "epoch": 0.044824734821859655,
      "grad_norm": 11.748570442199707,
      "learning_rate": 7.003293616196842e-06,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8234645172953605,
      "num_tokens": 17363083.0,
      "step": 14460
    },
    {
      "entropy": 1.8664287984371186,
      "epoch": 0.04485573394690935,
      "grad_norm": 11.027554512023926,
      "learning_rate": 7.008137169427493e-06,
      "loss": 0.6284,
      "mean_token_accuracy": 0.8028878584504128,
      "num_tokens": 17374802.0,
      "step": 14470
    },
    {
      "entropy": 1.8942121878266334,
      "epoch": 0.04488673307195904,
      "grad_norm": 12.925647735595703,
      "learning_rate": 7.012980722658143e-06,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8129858180880547,
      "num_tokens": 17386778.0,
      "step": 14480
    },
    {
      "entropy": 1.8534621268510818,
      "epoch": 0.04491773219700874,
      "grad_norm": 11.046483039855957,
      "learning_rate": 7.0178242758887926e-06,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8115181505680085,
      "num_tokens": 17399462.0,
      "step": 14490
    },
    {
      "entropy": 1.8930703341960906,
      "epoch": 0.044948731322058434,
      "grad_norm": 14.628288269042969,
      "learning_rate": 7.0226678291194424e-06,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8067412465810776,
      "num_tokens": 17411740.0,
      "step": 14500
    },
    {
      "entropy": 1.9142782092094421,
      "epoch": 0.04497973044710813,
      "grad_norm": 13.652180671691895,
      "learning_rate": 7.027511382350093e-06,
      "loss": 0.6096,
      "mean_token_accuracy": 0.8202868893742561,
      "num_tokens": 17423288.0,
      "step": 14510
    },
    {
      "entropy": 1.9144923388957977,
      "epoch": 0.04501072957215783,
      "grad_norm": 9.822308540344238,
      "learning_rate": 7.032354935580743e-06,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8314526736736297,
      "num_tokens": 17434455.0,
      "step": 14520
    },
    {
      "entropy": 1.926996847987175,
      "epoch": 0.045041728697207524,
      "grad_norm": 12.479947090148926,
      "learning_rate": 7.037198488811393e-06,
      "loss": 0.6008,
      "mean_token_accuracy": 0.8161266520619392,
      "num_tokens": 17445979.0,
      "step": 14530
    },
    {
      "entropy": 1.9374195352196693,
      "epoch": 0.045072727822257214,
      "grad_norm": 15.353309631347656,
      "learning_rate": 7.042042042042042e-06,
      "loss": 0.6577,
      "mean_token_accuracy": 0.8026217222213745,
      "num_tokens": 17457698.0,
      "step": 14540
    },
    {
      "entropy": 1.960123062133789,
      "epoch": 0.04510372694730691,
      "grad_norm": 13.066014289855957,
      "learning_rate": 7.046885595272692e-06,
      "loss": 0.664,
      "mean_token_accuracy": 0.8060840681195259,
      "num_tokens": 17468922.0,
      "step": 14550
    },
    {
      "entropy": 1.9222340703010559,
      "epoch": 0.04513472607235661,
      "grad_norm": 11.0694580078125,
      "learning_rate": 7.051729148503343e-06,
      "loss": 0.6043,
      "mean_token_accuracy": 0.8135023102164268,
      "num_tokens": 17480520.0,
      "step": 14560
    },
    {
      "entropy": 1.8666946336627006,
      "epoch": 0.0451657251974063,
      "grad_norm": 7.747961044311523,
      "learning_rate": 7.0565727017339925e-06,
      "loss": 0.5063,
      "mean_token_accuracy": 0.837678550183773,
      "num_tokens": 17493323.0,
      "step": 14570
    },
    {
      "entropy": 1.8420935034751893,
      "epoch": 0.045196724322456,
      "grad_norm": 10.135159492492676,
      "learning_rate": 7.061416254964642e-06,
      "loss": 0.502,
      "mean_token_accuracy": 0.8299189880490303,
      "num_tokens": 17506110.0,
      "step": 14580
    },
    {
      "entropy": 1.9041017875075341,
      "epoch": 0.0452277234475057,
      "grad_norm": 9.9268159866333,
      "learning_rate": 7.066259808195293e-06,
      "loss": 0.6489,
      "mean_token_accuracy": 0.8125099197030068,
      "num_tokens": 17517261.0,
      "step": 14590
    },
    {
      "entropy": 1.8802707374095917,
      "epoch": 0.045258722572555386,
      "grad_norm": 9.548727989196777,
      "learning_rate": 7.071103361425943e-06,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8187322363257408,
      "num_tokens": 17529058.0,
      "step": 14600
    },
    {
      "entropy": 1.9222388476133347,
      "epoch": 0.04528972169760508,
      "grad_norm": 13.59824275970459,
      "learning_rate": 7.075946914656593e-06,
      "loss": 0.6298,
      "mean_token_accuracy": 0.8107534229755402,
      "num_tokens": 17540313.0,
      "step": 14610
    },
    {
      "entropy": 1.8212895065546035,
      "epoch": 0.04532072082265478,
      "grad_norm": 5.862724304199219,
      "learning_rate": 7.080790467887243e-06,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8119758501648903,
      "num_tokens": 17553154.0,
      "step": 14620
    },
    {
      "entropy": 1.919022636115551,
      "epoch": 0.045351719947704476,
      "grad_norm": 11.380285263061523,
      "learning_rate": 7.0856340211178935e-06,
      "loss": 0.6269,
      "mean_token_accuracy": 0.804240868985653,
      "num_tokens": 17564581.0,
      "step": 14630
    },
    {
      "entropy": 1.8831742450594902,
      "epoch": 0.04538271907275417,
      "grad_norm": 12.382061004638672,
      "learning_rate": 7.090477574348543e-06,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8190920054912567,
      "num_tokens": 17576403.0,
      "step": 14640
    },
    {
      "entropy": 1.8691495105624198,
      "epoch": 0.04541371819780387,
      "grad_norm": 5.398478984832764,
      "learning_rate": 7.095321127579192e-06,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8149119704961777,
      "num_tokens": 17587726.0,
      "step": 14650
    },
    {
      "entropy": 1.7776469945907594,
      "epoch": 0.045444717322853566,
      "grad_norm": 11.435405731201172,
      "learning_rate": 7.100164680809842e-06,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8274336785078049,
      "num_tokens": 17600178.0,
      "step": 14660
    },
    {
      "entropy": 1.8829714879393578,
      "epoch": 0.045475716447903256,
      "grad_norm": 6.995675563812256,
      "learning_rate": 7.105008234040492e-06,
      "loss": 0.6475,
      "mean_token_accuracy": 0.8128258779644966,
      "num_tokens": 17611382.0,
      "step": 14670
    },
    {
      "entropy": 1.906617347896099,
      "epoch": 0.04550671557295295,
      "grad_norm": 6.328146457672119,
      "learning_rate": 7.109851787271143e-06,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8149865731596947,
      "num_tokens": 17623005.0,
      "step": 14680
    },
    {
      "entropy": 1.8847622662782668,
      "epoch": 0.04553771469800265,
      "grad_norm": 12.374866485595703,
      "learning_rate": 7.114695340501793e-06,
      "loss": 0.5861,
      "mean_token_accuracy": 0.8213684529066085,
      "num_tokens": 17634543.0,
      "step": 14690
    },
    {
      "entropy": 1.8927973687648774,
      "epoch": 0.045568713823052345,
      "grad_norm": 11.521233558654785,
      "learning_rate": 7.119538893732443e-06,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8219080328941345,
      "num_tokens": 17645810.0,
      "step": 14700
    },
    {
      "entropy": 1.9088530361652374,
      "epoch": 0.04559971294810204,
      "grad_norm": 11.580865859985352,
      "learning_rate": 7.1243824469630925e-06,
      "loss": 0.6404,
      "mean_token_accuracy": 0.8121732175350189,
      "num_tokens": 17657230.0,
      "step": 14710
    },
    {
      "entropy": 1.906784760951996,
      "epoch": 0.04563071207315174,
      "grad_norm": 10.853860855102539,
      "learning_rate": 7.129226000193743e-06,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8171251997351646,
      "num_tokens": 17669157.0,
      "step": 14720
    },
    {
      "entropy": 1.8272288024425507,
      "epoch": 0.04566171119820143,
      "grad_norm": 11.097689628601074,
      "learning_rate": 7.134069553424393e-06,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8214239537715912,
      "num_tokens": 17682522.0,
      "step": 14730
    },
    {
      "entropy": 1.9602118730545044,
      "epoch": 0.045692710323251125,
      "grad_norm": 11.482585906982422,
      "learning_rate": 7.138913106655043e-06,
      "loss": 0.6694,
      "mean_token_accuracy": 0.8125876560807228,
      "num_tokens": 17693830.0,
      "step": 14740
    },
    {
      "entropy": 1.8845744907855988,
      "epoch": 0.04572370944830082,
      "grad_norm": 11.898690223693848,
      "learning_rate": 7.143756659885692e-06,
      "loss": 0.621,
      "mean_token_accuracy": 0.8045976728200912,
      "num_tokens": 17706045.0,
      "step": 14750
    },
    {
      "entropy": 1.8579593807458878,
      "epoch": 0.04575470857335052,
      "grad_norm": 5.022470474243164,
      "learning_rate": 7.148600213116342e-06,
      "loss": 0.6451,
      "mean_token_accuracy": 0.804808932542801,
      "num_tokens": 17718119.0,
      "step": 14760
    },
    {
      "entropy": 1.9066347777843475,
      "epoch": 0.045785707698400215,
      "grad_norm": 10.700624465942383,
      "learning_rate": 7.153443766346993e-06,
      "loss": 0.61,
      "mean_token_accuracy": 0.8180598929524422,
      "num_tokens": 17729029.0,
      "step": 14770
    },
    {
      "entropy": 1.8993641972541808,
      "epoch": 0.04581670682344991,
      "grad_norm": 10.812292098999023,
      "learning_rate": 7.1582873195776426e-06,
      "loss": 0.6628,
      "mean_token_accuracy": 0.8115692853927612,
      "num_tokens": 17740072.0,
      "step": 14780
    },
    {
      "entropy": 1.8603774085640907,
      "epoch": 0.0458477059484996,
      "grad_norm": 13.853246688842773,
      "learning_rate": 7.1631308728082925e-06,
      "loss": 0.609,
      "mean_token_accuracy": 0.8194773256778717,
      "num_tokens": 17751523.0,
      "step": 14790
    },
    {
      "entropy": 1.7647764384746552,
      "epoch": 0.0458787050735493,
      "grad_norm": 13.947328567504883,
      "learning_rate": 7.167974426038943e-06,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8233131125569344,
      "num_tokens": 17763817.0,
      "step": 14800
    },
    {
      "entropy": 1.750529208779335,
      "epoch": 0.045909704198598994,
      "grad_norm": 5.576157569885254,
      "learning_rate": 7.172817979269593e-06,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8132794231176377,
      "num_tokens": 17777124.0,
      "step": 14810
    },
    {
      "entropy": 1.851732324063778,
      "epoch": 0.04594070332364869,
      "grad_norm": 11.042110443115234,
      "learning_rate": 7.177661532500243e-06,
      "loss": 0.6467,
      "mean_token_accuracy": 0.8058951094746589,
      "num_tokens": 17788612.0,
      "step": 14820
    },
    {
      "entropy": 1.7839807882905006,
      "epoch": 0.04597170244869839,
      "grad_norm": 10.22685718536377,
      "learning_rate": 7.182505085730893e-06,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8168496385216713,
      "num_tokens": 17801156.0,
      "step": 14830
    },
    {
      "entropy": 1.8429742008447647,
      "epoch": 0.046002701573748084,
      "grad_norm": 5.943838596343994,
      "learning_rate": 7.1873486389615436e-06,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8247615218162536,
      "num_tokens": 17813245.0,
      "step": 14840
    },
    {
      "entropy": 1.8560289978981017,
      "epoch": 0.04603370069879777,
      "grad_norm": 11.962748527526855,
      "learning_rate": 7.1921921921921935e-06,
      "loss": 0.651,
      "mean_token_accuracy": 0.8067975297570229,
      "num_tokens": 17824749.0,
      "step": 14850
    },
    {
      "entropy": 1.7992481097579003,
      "epoch": 0.04606469982384747,
      "grad_norm": 13.64804744720459,
      "learning_rate": 7.1970357454228425e-06,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8238918125629425,
      "num_tokens": 17838276.0,
      "step": 14860
    },
    {
      "entropy": 1.925189484655857,
      "epoch": 0.04609569894889717,
      "grad_norm": 10.546503067016602,
      "learning_rate": 7.201879298653492e-06,
      "loss": 0.6402,
      "mean_token_accuracy": 0.8055410549044609,
      "num_tokens": 17849457.0,
      "step": 14870
    },
    {
      "entropy": 1.939789319038391,
      "epoch": 0.04612669807394686,
      "grad_norm": 13.753273963928223,
      "learning_rate": 7.206722851884142e-06,
      "loss": 0.6455,
      "mean_token_accuracy": 0.807138554751873,
      "num_tokens": 17860102.0,
      "step": 14880
    },
    {
      "entropy": 1.8846666172146798,
      "epoch": 0.04615769719899656,
      "grad_norm": 11.280816078186035,
      "learning_rate": 7.211566405114793e-06,
      "loss": 0.6067,
      "mean_token_accuracy": 0.8096478402614593,
      "num_tokens": 17871606.0,
      "step": 14890
    },
    {
      "entropy": 1.918617771565914,
      "epoch": 0.046188696324046256,
      "grad_norm": 10.672183990478516,
      "learning_rate": 7.216409958345443e-06,
      "loss": 0.6779,
      "mean_token_accuracy": 0.8008549973368645,
      "num_tokens": 17882968.0,
      "step": 14900
    },
    {
      "entropy": 1.9250801861286164,
      "epoch": 0.046219695449095946,
      "grad_norm": 10.703913688659668,
      "learning_rate": 7.221253511576093e-06,
      "loss": 0.6541,
      "mean_token_accuracy": 0.8061710268259048,
      "num_tokens": 17894617.0,
      "step": 14910
    },
    {
      "entropy": 1.8236558228731155,
      "epoch": 0.04625069457414564,
      "grad_norm": 11.690652847290039,
      "learning_rate": 7.226097064806743e-06,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8214540228247642,
      "num_tokens": 17907854.0,
      "step": 14920
    },
    {
      "entropy": 1.8554262310266494,
      "epoch": 0.04628169369919534,
      "grad_norm": 10.404338836669922,
      "learning_rate": 7.230940618037393e-06,
      "loss": 0.5821,
      "mean_token_accuracy": 0.8083539769053459,
      "num_tokens": 17920382.0,
      "step": 14930
    },
    {
      "entropy": 1.855211953818798,
      "epoch": 0.046312692824245036,
      "grad_norm": 10.875576972961426,
      "learning_rate": 7.235784171268043e-06,
      "loss": 0.5979,
      "mean_token_accuracy": 0.8178864941000938,
      "num_tokens": 17932205.0,
      "step": 14940
    },
    {
      "entropy": 1.8454056218266488,
      "epoch": 0.04634369194929473,
      "grad_norm": 11.040106773376465,
      "learning_rate": 7.240627724498693e-06,
      "loss": 0.58,
      "mean_token_accuracy": 0.8090640112757683,
      "num_tokens": 17943421.0,
      "step": 14950
    },
    {
      "entropy": 1.901095400750637,
      "epoch": 0.04637469107434443,
      "grad_norm": 10.230569839477539,
      "learning_rate": 7.245471277729342e-06,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8283701628446579,
      "num_tokens": 17954715.0,
      "step": 14960
    },
    {
      "entropy": 1.8909978330135346,
      "epoch": 0.04640569019939412,
      "grad_norm": 12.003896713256836,
      "learning_rate": 7.250314830959992e-06,
      "loss": 0.6415,
      "mean_token_accuracy": 0.809289188683033,
      "num_tokens": 17966078.0,
      "step": 14970
    },
    {
      "entropy": 1.893742746114731,
      "epoch": 0.046436689324443815,
      "grad_norm": 12.008511543273926,
      "learning_rate": 7.255158384190643e-06,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8165367186069489,
      "num_tokens": 17977838.0,
      "step": 14980
    },
    {
      "entropy": 1.9413186386227608,
      "epoch": 0.04646768844949351,
      "grad_norm": 11.98163890838623,
      "learning_rate": 7.260001937421293e-06,
      "loss": 0.632,
      "mean_token_accuracy": 0.8125703752040863,
      "num_tokens": 17989242.0,
      "step": 14990
    },
    {
      "entropy": 1.8333601012825966,
      "epoch": 0.04649868757454321,
      "grad_norm": 5.240977764129639,
      "learning_rate": 7.2648454906519426e-06,
      "loss": 0.537,
      "mean_token_accuracy": 0.8350656241178512,
      "num_tokens": 18001491.0,
      "step": 15000
    },
    {
      "entropy": 1.8348119348287582,
      "epoch": 0.046529686699592905,
      "grad_norm": 5.844879150390625,
      "learning_rate": 7.269689043882593e-06,
      "loss": 0.602,
      "mean_token_accuracy": 0.8153130680322647,
      "num_tokens": 18014233.0,
      "step": 15010
    },
    {
      "entropy": 1.8076252147555352,
      "epoch": 0.0465606858246426,
      "grad_norm": 5.277684688568115,
      "learning_rate": 7.274532597113243e-06,
      "loss": 0.5404,
      "mean_token_accuracy": 0.824401643872261,
      "num_tokens": 18027418.0,
      "step": 15020
    },
    {
      "entropy": 1.9218341365456582,
      "epoch": 0.0465916849496923,
      "grad_norm": 10.615427017211914,
      "learning_rate": 7.279376150343893e-06,
      "loss": 0.608,
      "mean_token_accuracy": 0.8177351862192154,
      "num_tokens": 18038854.0,
      "step": 15030
    },
    {
      "entropy": 1.9281207531690598,
      "epoch": 0.04662268407474199,
      "grad_norm": 9.128558158874512,
      "learning_rate": 7.284219703574543e-06,
      "loss": 0.598,
      "mean_token_accuracy": 0.8243363618850708,
      "num_tokens": 18049861.0,
      "step": 15040
    },
    {
      "entropy": 1.8921822875738143,
      "epoch": 0.046653683199791685,
      "grad_norm": 13.774589538574219,
      "learning_rate": 7.289063256805194e-06,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8057242497801781,
      "num_tokens": 18061596.0,
      "step": 15050
    },
    {
      "entropy": 1.832139255106449,
      "epoch": 0.04668468232484138,
      "grad_norm": 10.49471378326416,
      "learning_rate": 7.2939068100358436e-06,
      "loss": 0.5717,
      "mean_token_accuracy": 0.816807533800602,
      "num_tokens": 18073371.0,
      "step": 15060
    },
    {
      "entropy": 1.8594244837760925,
      "epoch": 0.04671568144989108,
      "grad_norm": 11.264952659606934,
      "learning_rate": 7.298750363266493e-06,
      "loss": 0.6349,
      "mean_token_accuracy": 0.8005358681082726,
      "num_tokens": 18086010.0,
      "step": 15070
    },
    {
      "entropy": 1.8758624821901322,
      "epoch": 0.046746680574940774,
      "grad_norm": 13.645959854125977,
      "learning_rate": 7.3035939164971425e-06,
      "loss": 0.5787,
      "mean_token_accuracy": 0.8165410995483399,
      "num_tokens": 18097503.0,
      "step": 15080
    },
    {
      "entropy": 1.8632541045546531,
      "epoch": 0.04677767969999047,
      "grad_norm": 12.606849670410156,
      "learning_rate": 7.308437469727792e-06,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8230202242732048,
      "num_tokens": 18109025.0,
      "step": 15090
    },
    {
      "entropy": 1.8468170419335366,
      "epoch": 0.04680867882504016,
      "grad_norm": 5.763012409210205,
      "learning_rate": 7.313281022958443e-06,
      "loss": 0.5543,
      "mean_token_accuracy": 0.82147196829319,
      "num_tokens": 18120920.0,
      "step": 15100
    },
    {
      "entropy": 1.8019221499562263,
      "epoch": 0.04683967795008986,
      "grad_norm": 12.194572448730469,
      "learning_rate": 7.318124576189093e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8267902106046676,
      "num_tokens": 18134004.0,
      "step": 15110
    },
    {
      "entropy": 1.9301525950431824,
      "epoch": 0.046870677075139554,
      "grad_norm": 11.044305801391602,
      "learning_rate": 7.322968129419743e-06,
      "loss": 0.7477,
      "mean_token_accuracy": 0.798159584403038,
      "num_tokens": 18145789.0,
      "step": 15120
    },
    {
      "entropy": 1.9049590498209,
      "epoch": 0.04690167620018925,
      "grad_norm": 12.671975135803223,
      "learning_rate": 7.327811682650393e-06,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8189090758562088,
      "num_tokens": 18157088.0,
      "step": 15130
    },
    {
      "entropy": 1.8682396605610847,
      "epoch": 0.04693267532523895,
      "grad_norm": 9.93550968170166,
      "learning_rate": 7.3326552358810435e-06,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8179777503013611,
      "num_tokens": 18169072.0,
      "step": 15140
    },
    {
      "entropy": 1.8468847021460533,
      "epoch": 0.046963674450288644,
      "grad_norm": 9.431316375732422,
      "learning_rate": 7.337498789111693e-06,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8228641420602798,
      "num_tokens": 18181019.0,
      "step": 15150
    },
    {
      "entropy": 1.775972270965576,
      "epoch": 0.04699467357533833,
      "grad_norm": 10.788688659667969,
      "learning_rate": 7.342342342342343e-06,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8265146508812904,
      "num_tokens": 18194155.0,
      "step": 15160
    },
    {
      "entropy": 1.8457147806882859,
      "epoch": 0.04702567270038803,
      "grad_norm": 9.854268074035645,
      "learning_rate": 7.347185895572992e-06,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8127473801374435,
      "num_tokens": 18206750.0,
      "step": 15170
    },
    {
      "entropy": 1.7577089250087738,
      "epoch": 0.047056671825437726,
      "grad_norm": 5.139494895935059,
      "learning_rate": 7.352029448803642e-06,
      "loss": 0.4973,
      "mean_token_accuracy": 0.8430294960737228,
      "num_tokens": 18219202.0,
      "step": 15180
    },
    {
      "entropy": 1.8935526996850967,
      "epoch": 0.04708767095048742,
      "grad_norm": 5.4716057777404785,
      "learning_rate": 7.356873002034293e-06,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8127918854355812,
      "num_tokens": 18230861.0,
      "step": 15190
    },
    {
      "entropy": 1.8892264723777772,
      "epoch": 0.04711867007553712,
      "grad_norm": 12.728769302368164,
      "learning_rate": 7.361716555264943e-06,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8149558499455452,
      "num_tokens": 18242915.0,
      "step": 15200
    },
    {
      "entropy": 1.8911462113261224,
      "epoch": 0.047149669200586816,
      "grad_norm": 9.735945701599121,
      "learning_rate": 7.366560108495593e-06,
      "loss": 0.6442,
      "mean_token_accuracy": 0.8063351511955261,
      "num_tokens": 18253977.0,
      "step": 15210
    },
    {
      "entropy": 1.9077469795942306,
      "epoch": 0.047180668325636506,
      "grad_norm": 13.403579711914062,
      "learning_rate": 7.371403661726243e-06,
      "loss": 0.6695,
      "mean_token_accuracy": 0.8085703745484352,
      "num_tokens": 18264673.0,
      "step": 15220
    },
    {
      "entropy": 1.86230625808239,
      "epoch": 0.0472116674506862,
      "grad_norm": 5.826930999755859,
      "learning_rate": 7.376247214956893e-06,
      "loss": 0.5957,
      "mean_token_accuracy": 0.8189520880579948,
      "num_tokens": 18276799.0,
      "step": 15230
    },
    {
      "entropy": 1.9372797414660454,
      "epoch": 0.0472426665757359,
      "grad_norm": 12.060833930969238,
      "learning_rate": 7.381090768187543e-06,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8150049105286599,
      "num_tokens": 18288796.0,
      "step": 15240
    },
    {
      "entropy": 1.788297127187252,
      "epoch": 0.047273665700785596,
      "grad_norm": 5.706685543060303,
      "learning_rate": 7.385934321418193e-06,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8224897965788841,
      "num_tokens": 18301726.0,
      "step": 15250
    },
    {
      "entropy": 1.9241210684180259,
      "epoch": 0.04730466482583529,
      "grad_norm": 11.742621421813965,
      "learning_rate": 7.390777874648844e-06,
      "loss": 0.6246,
      "mean_token_accuracy": 0.810030820965767,
      "num_tokens": 18312680.0,
      "step": 15260
    },
    {
      "entropy": 1.9467894092202187,
      "epoch": 0.04733566395088499,
      "grad_norm": 9.387283325195312,
      "learning_rate": 7.395621427879494e-06,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8097245275974274,
      "num_tokens": 18324139.0,
      "step": 15270
    },
    {
      "entropy": 1.9100206598639489,
      "epoch": 0.04736666307593468,
      "grad_norm": 10.441351890563965,
      "learning_rate": 7.400464981110143e-06,
      "loss": 0.6332,
      "mean_token_accuracy": 0.7996808186173439,
      "num_tokens": 18336228.0,
      "step": 15280
    },
    {
      "entropy": 1.96844242811203,
      "epoch": 0.047397662200984375,
      "grad_norm": 11.180095672607422,
      "learning_rate": 7.405308534340793e-06,
      "loss": 0.675,
      "mean_token_accuracy": 0.8107746347784996,
      "num_tokens": 18346950.0,
      "step": 15290
    },
    {
      "entropy": 1.9150108516216278,
      "epoch": 0.04742866132603407,
      "grad_norm": 9.653234481811523,
      "learning_rate": 7.4101520875714425e-06,
      "loss": 0.6054,
      "mean_token_accuracy": 0.8183826908469201,
      "num_tokens": 18358537.0,
      "step": 15300
    },
    {
      "entropy": 1.8632504418492317,
      "epoch": 0.04745966045108377,
      "grad_norm": 13.043754577636719,
      "learning_rate": 7.414995640802093e-06,
      "loss": 0.5891,
      "mean_token_accuracy": 0.8238464057445526,
      "num_tokens": 18371462.0,
      "step": 15310
    },
    {
      "entropy": 1.8496581763029099,
      "epoch": 0.047490659576133465,
      "grad_norm": 11.938183784484863,
      "learning_rate": 7.419839194032743e-06,
      "loss": 0.5633,
      "mean_token_accuracy": 0.827662679553032,
      "num_tokens": 18384026.0,
      "step": 15320
    },
    {
      "entropy": 1.8263291805982589,
      "epoch": 0.04752165870118316,
      "grad_norm": 10.78771686553955,
      "learning_rate": 7.424682747263393e-06,
      "loss": 0.5709,
      "mean_token_accuracy": 0.815859617292881,
      "num_tokens": 18396414.0,
      "step": 15330
    },
    {
      "entropy": 1.8266072757542133,
      "epoch": 0.04755265782623285,
      "grad_norm": 11.682705879211426,
      "learning_rate": 7.429526300494043e-06,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8147292017936707,
      "num_tokens": 18409403.0,
      "step": 15340
    },
    {
      "entropy": 1.8509322211146355,
      "epoch": 0.04758365695128255,
      "grad_norm": 4.418774127960205,
      "learning_rate": 7.434369853724694e-06,
      "loss": 0.5744,
      "mean_token_accuracy": 0.822237353026867,
      "num_tokens": 18421350.0,
      "step": 15350
    },
    {
      "entropy": 1.8988280490040779,
      "epoch": 0.047614656076332244,
      "grad_norm": 11.46986198425293,
      "learning_rate": 7.4392134069553435e-06,
      "loss": 0.5755,
      "mean_token_accuracy": 0.8086514964699745,
      "num_tokens": 18433565.0,
      "step": 15360
    },
    {
      "entropy": 1.883128097653389,
      "epoch": 0.04764565520138194,
      "grad_norm": 10.333133697509766,
      "learning_rate": 7.444056960185993e-06,
      "loss": 0.6242,
      "mean_token_accuracy": 0.8101354941725731,
      "num_tokens": 18445233.0,
      "step": 15370
    },
    {
      "entropy": 1.7941233783960342,
      "epoch": 0.04767665432643164,
      "grad_norm": 4.953238010406494,
      "learning_rate": 7.448900513416642e-06,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8319120317697525,
      "num_tokens": 18457975.0,
      "step": 15380
    },
    {
      "entropy": 1.9092483133077622,
      "epoch": 0.047707653451481334,
      "grad_norm": 11.126029014587402,
      "learning_rate": 7.453744066647292e-06,
      "loss": 0.6611,
      "mean_token_accuracy": 0.8040424823760987,
      "num_tokens": 18469319.0,
      "step": 15390
    },
    {
      "entropy": 1.8364340022206307,
      "epoch": 0.04773865257653103,
      "grad_norm": 5.549501419067383,
      "learning_rate": 7.458587619877943e-06,
      "loss": 0.6241,
      "mean_token_accuracy": 0.8029653668403626,
      "num_tokens": 18482789.0,
      "step": 15400
    },
    {
      "entropy": 1.8426788106560708,
      "epoch": 0.04776965170158072,
      "grad_norm": 11.887903213500977,
      "learning_rate": 7.463431173108593e-06,
      "loss": 0.5929,
      "mean_token_accuracy": 0.8010544419288635,
      "num_tokens": 18495234.0,
      "step": 15410
    },
    {
      "entropy": 1.898276437819004,
      "epoch": 0.04780065082663042,
      "grad_norm": 11.306375503540039,
      "learning_rate": 7.468274726339243e-06,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8193996638059616,
      "num_tokens": 18507024.0,
      "step": 15420
    },
    {
      "entropy": 1.9499488294124603,
      "epoch": 0.047831649951680114,
      "grad_norm": 13.077889442443848,
      "learning_rate": 7.4731182795698935e-06,
      "loss": 0.7473,
      "mean_token_accuracy": 0.7906885385513306,
      "num_tokens": 18518699.0,
      "step": 15430
    },
    {
      "entropy": 1.8603081166744233,
      "epoch": 0.04786264907672981,
      "grad_norm": 11.552556037902832,
      "learning_rate": 7.477961832800543e-06,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8289436608552933,
      "num_tokens": 18530154.0,
      "step": 15440
    },
    {
      "entropy": 1.8968264564871788,
      "epoch": 0.04789364820177951,
      "grad_norm": 12.209922790527344,
      "learning_rate": 7.482805386031193e-06,
      "loss": 0.6208,
      "mean_token_accuracy": 0.8118745431303978,
      "num_tokens": 18541201.0,
      "step": 15450
    },
    {
      "entropy": 1.7692169472575188,
      "epoch": 0.0479246473268292,
      "grad_norm": 11.662141799926758,
      "learning_rate": 7.487648939261843e-06,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8198059305548668,
      "num_tokens": 18554224.0,
      "step": 15460
    },
    {
      "entropy": 1.927010752260685,
      "epoch": 0.04795564645187889,
      "grad_norm": 10.837479591369629,
      "learning_rate": 7.492492492492494e-06,
      "loss": 0.5931,
      "mean_token_accuracy": 0.8046490982174873,
      "num_tokens": 18566184.0,
      "step": 15470
    },
    {
      "entropy": 1.854862241446972,
      "epoch": 0.04798664557692859,
      "grad_norm": 9.293188095092773,
      "learning_rate": 7.497336045723144e-06,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8218624517321587,
      "num_tokens": 18578244.0,
      "step": 15480
    },
    {
      "entropy": 1.8932930827140808,
      "epoch": 0.048017644701978286,
      "grad_norm": 10.518685340881348,
      "learning_rate": 7.502179598953793e-06,
      "loss": 0.6514,
      "mean_token_accuracy": 0.8079543009400367,
      "num_tokens": 18589205.0,
      "step": 15490
    },
    {
      "entropy": 1.8627614587545396,
      "epoch": 0.04804864382702798,
      "grad_norm": 14.004125595092773,
      "learning_rate": 7.507023152184443e-06,
      "loss": 0.6486,
      "mean_token_accuracy": 0.8162886321544647,
      "num_tokens": 18600091.0,
      "step": 15500
    },
    {
      "entropy": 1.8322096571326256,
      "epoch": 0.04807964295207768,
      "grad_norm": 10.797554016113281,
      "learning_rate": 7.5118667054150926e-06,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8135915264487267,
      "num_tokens": 18611780.0,
      "step": 15510
    },
    {
      "entropy": 1.7482316687703132,
      "epoch": 0.048110642077127376,
      "grad_norm": 11.08513069152832,
      "learning_rate": 7.516710258645743e-06,
      "loss": 0.581,
      "mean_token_accuracy": 0.8262606129050255,
      "num_tokens": 18624882.0,
      "step": 15520
    },
    {
      "entropy": 1.8912085920572281,
      "epoch": 0.048141641202177066,
      "grad_norm": 12.140768051147461,
      "learning_rate": 7.521553811876393e-06,
      "loss": 0.6816,
      "mean_token_accuracy": 0.8002603054046631,
      "num_tokens": 18636954.0,
      "step": 15530
    },
    {
      "entropy": 1.8268685653805732,
      "epoch": 0.04817264032722676,
      "grad_norm": 12.404664039611816,
      "learning_rate": 7.526397365107043e-06,
      "loss": 0.5911,
      "mean_token_accuracy": 0.8149344757199287,
      "num_tokens": 18649266.0,
      "step": 15540
    },
    {
      "entropy": 1.8767468333244324,
      "epoch": 0.04820363945227646,
      "grad_norm": 10.459026336669922,
      "learning_rate": 7.531240918337693e-06,
      "loss": 0.6271,
      "mean_token_accuracy": 0.8109247386455536,
      "num_tokens": 18660751.0,
      "step": 15550
    },
    {
      "entropy": 1.831296342611313,
      "epoch": 0.048234638577326155,
      "grad_norm": 10.499650955200195,
      "learning_rate": 7.536084471568344e-06,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8176342695951462,
      "num_tokens": 18673033.0,
      "step": 15560
    },
    {
      "entropy": 1.8214371725916862,
      "epoch": 0.04826563770237585,
      "grad_norm": 14.110614776611328,
      "learning_rate": 7.5409280247989936e-06,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8230008244514465,
      "num_tokens": 18685183.0,
      "step": 15570
    },
    {
      "entropy": 1.8123807251453399,
      "epoch": 0.04829663682742555,
      "grad_norm": 10.320479393005371,
      "learning_rate": 7.5457715780296435e-06,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8221219345927239,
      "num_tokens": 18696924.0,
      "step": 15580
    },
    {
      "entropy": 1.7216524064540863,
      "epoch": 0.04832763595247524,
      "grad_norm": 9.836614608764648,
      "learning_rate": 7.5506151312602925e-06,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8300070941448212,
      "num_tokens": 18710414.0,
      "step": 15590
    },
    {
      "entropy": 1.889795495569706,
      "epoch": 0.048358635077524935,
      "grad_norm": 11.627612113952637,
      "learning_rate": 7.555458684490942e-06,
      "loss": 0.6584,
      "mean_token_accuracy": 0.8059937000274658,
      "num_tokens": 18721801.0,
      "step": 15600
    },
    {
      "entropy": 1.8819514483213424,
      "epoch": 0.04838963420257463,
      "grad_norm": 6.095297813415527,
      "learning_rate": 7.560302237721593e-06,
      "loss": 0.6108,
      "mean_token_accuracy": 0.807888326048851,
      "num_tokens": 18733560.0,
      "step": 15610
    },
    {
      "entropy": 1.8655756518244744,
      "epoch": 0.04842063332762433,
      "grad_norm": 10.622197151184082,
      "learning_rate": 7.565145790952243e-06,
      "loss": 0.6113,
      "mean_token_accuracy": 0.8073986008763313,
      "num_tokens": 18746189.0,
      "step": 15620
    },
    {
      "entropy": 1.8603959396481513,
      "epoch": 0.048451632452674025,
      "grad_norm": 11.854666709899902,
      "learning_rate": 7.569989344182893e-06,
      "loss": 0.5933,
      "mean_token_accuracy": 0.817948243021965,
      "num_tokens": 18757924.0,
      "step": 15630
    },
    {
      "entropy": 1.9059595987200737,
      "epoch": 0.04848263157772372,
      "grad_norm": 9.098125457763672,
      "learning_rate": 7.574832897413544e-06,
      "loss": 0.6782,
      "mean_token_accuracy": 0.8045271784067154,
      "num_tokens": 18769446.0,
      "step": 15640
    },
    {
      "entropy": 1.7989111676812173,
      "epoch": 0.04851363070277341,
      "grad_norm": 11.551828384399414,
      "learning_rate": 7.5796764506441935e-06,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8248863905668259,
      "num_tokens": 18782700.0,
      "step": 15650
    },
    {
      "entropy": 1.8210396483540534,
      "epoch": 0.04854462982782311,
      "grad_norm": 4.9011759757995605,
      "learning_rate": 7.584520003874843e-06,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8225024446845055,
      "num_tokens": 18795390.0,
      "step": 15660
    },
    {
      "entropy": 1.811398984491825,
      "epoch": 0.048575628952872804,
      "grad_norm": 11.048978805541992,
      "learning_rate": 7.589363557105493e-06,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8189741969108582,
      "num_tokens": 18807396.0,
      "step": 15670
    },
    {
      "entropy": 1.892940777540207,
      "epoch": 0.0486066280779225,
      "grad_norm": 12.325209617614746,
      "learning_rate": 7.594207110336144e-06,
      "loss": 0.6986,
      "mean_token_accuracy": 0.7981882244348526,
      "num_tokens": 18819378.0,
      "step": 15680
    },
    {
      "entropy": 1.886229231953621,
      "epoch": 0.0486376272029722,
      "grad_norm": 10.576592445373535,
      "learning_rate": 7.599050663566794e-06,
      "loss": 0.6074,
      "mean_token_accuracy": 0.8123501881957054,
      "num_tokens": 18830551.0,
      "step": 15690
    },
    {
      "entropy": 1.8603569209575652,
      "epoch": 0.048668626328021894,
      "grad_norm": 11.795734405517578,
      "learning_rate": 7.603894216797443e-06,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8348536252975464,
      "num_tokens": 18842716.0,
      "step": 15700
    },
    {
      "entropy": 1.9385449796915055,
      "epoch": 0.04869962545307159,
      "grad_norm": 10.914395332336426,
      "learning_rate": 7.608737770028093e-06,
      "loss": 0.6356,
      "mean_token_accuracy": 0.8061470225453377,
      "num_tokens": 18853817.0,
      "step": 15710
    },
    {
      "entropy": 1.8827759057283402,
      "epoch": 0.04873062457812128,
      "grad_norm": 14.19170093536377,
      "learning_rate": 7.613581323258743e-06,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8078219577670097,
      "num_tokens": 18865592.0,
      "step": 15720
    },
    {
      "entropy": 1.8566590577363968,
      "epoch": 0.04876162370317098,
      "grad_norm": 13.042546272277832,
      "learning_rate": 7.618424876489393e-06,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8074685573577881,
      "num_tokens": 18878717.0,
      "step": 15730
    },
    {
      "entropy": 1.8778809905052185,
      "epoch": 0.04879262282822067,
      "grad_norm": 9.526302337646484,
      "learning_rate": 7.623268429720043e-06,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8302160531282425,
      "num_tokens": 18890522.0,
      "step": 15740
    },
    {
      "entropy": 1.843679629266262,
      "epoch": 0.04882362195327037,
      "grad_norm": 11.040970802307129,
      "learning_rate": 7.628111982950693e-06,
      "loss": 0.6254,
      "mean_token_accuracy": 0.812717217206955,
      "num_tokens": 18902667.0,
      "step": 15750
    },
    {
      "entropy": 1.9100946336984634,
      "epoch": 0.04885462107832007,
      "grad_norm": 7.509261608123779,
      "learning_rate": 7.632955536181344e-06,
      "loss": 0.6673,
      "mean_token_accuracy": 0.8177873358130455,
      "num_tokens": 18913868.0,
      "step": 15760
    },
    {
      "entropy": 1.8826175585389138,
      "epoch": 0.04888562020336976,
      "grad_norm": 11.155854225158691,
      "learning_rate": 7.637799089411994e-06,
      "loss": 0.6481,
      "mean_token_accuracy": 0.8087216004729271,
      "num_tokens": 18925988.0,
      "step": 15770
    },
    {
      "entropy": 1.7796615183353424,
      "epoch": 0.04891661932841945,
      "grad_norm": 5.291562080383301,
      "learning_rate": 7.642642642642644e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8381139814853669,
      "num_tokens": 18938466.0,
      "step": 15780
    },
    {
      "entropy": 1.9144254103302956,
      "epoch": 0.04894761845346915,
      "grad_norm": 10.03088092803955,
      "learning_rate": 7.647486195873294e-06,
      "loss": 0.6672,
      "mean_token_accuracy": 0.8007161229848861,
      "num_tokens": 18949499.0,
      "step": 15790
    },
    {
      "entropy": 1.8293558046221734,
      "epoch": 0.048978617578518846,
      "grad_norm": 10.299362182617188,
      "learning_rate": 7.652329749103943e-06,
      "loss": 0.6104,
      "mean_token_accuracy": 0.8068207755684853,
      "num_tokens": 18961354.0,
      "step": 15800
    },
    {
      "entropy": 1.843856942653656,
      "epoch": 0.04900961670356854,
      "grad_norm": 5.878364562988281,
      "learning_rate": 7.657173302334593e-06,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8132404252886772,
      "num_tokens": 18973459.0,
      "step": 15810
    },
    {
      "entropy": 1.8029664367437364,
      "epoch": 0.04904061582861824,
      "grad_norm": 13.001435279846191,
      "learning_rate": 7.662016855565243e-06,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8184084549546242,
      "num_tokens": 18986940.0,
      "step": 15820
    },
    {
      "entropy": 1.836187256872654,
      "epoch": 0.049071614953667936,
      "grad_norm": 14.026951789855957,
      "learning_rate": 7.666860408795893e-06,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8184718936681747,
      "num_tokens": 19000053.0,
      "step": 15830
    },
    {
      "entropy": 2.0127352684736253,
      "epoch": 0.049102614078717625,
      "grad_norm": 10.804215431213379,
      "learning_rate": 7.671703962026543e-06,
      "loss": 0.6748,
      "mean_token_accuracy": 0.8046320468187332,
      "num_tokens": 19010864.0,
      "step": 15840
    },
    {
      "entropy": 1.8914284870028495,
      "epoch": 0.04913361320376732,
      "grad_norm": 12.271016120910645,
      "learning_rate": 7.676547515257193e-06,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8181375965476037,
      "num_tokens": 19022131.0,
      "step": 15850
    },
    {
      "entropy": 1.8896478191018105,
      "epoch": 0.04916461232881702,
      "grad_norm": 10.453010559082031,
      "learning_rate": 7.681391068487843e-06,
      "loss": 0.6347,
      "mean_token_accuracy": 0.8188618034124374,
      "num_tokens": 19033759.0,
      "step": 15860
    },
    {
      "entropy": 1.9267440363764763,
      "epoch": 0.049195611453866715,
      "grad_norm": 12.614462852478027,
      "learning_rate": 7.686234621718494e-06,
      "loss": 0.6087,
      "mean_token_accuracy": 0.819622540473938,
      "num_tokens": 19044630.0,
      "step": 15870
    },
    {
      "entropy": 1.8585731774568557,
      "epoch": 0.04922661057891641,
      "grad_norm": 11.191518783569336,
      "learning_rate": 7.691078174949144e-06,
      "loss": 0.6359,
      "mean_token_accuracy": 0.8025752380490303,
      "num_tokens": 19056706.0,
      "step": 15880
    },
    {
      "entropy": 1.8346818998456,
      "epoch": 0.04925760970396611,
      "grad_norm": 10.337366104125977,
      "learning_rate": 7.695921728179794e-06,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8142318353056908,
      "num_tokens": 19069370.0,
      "step": 15890
    },
    {
      "entropy": 1.8798568680882455,
      "epoch": 0.0492886088290158,
      "grad_norm": 11.572778701782227,
      "learning_rate": 7.700765281410444e-06,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8030693680047989,
      "num_tokens": 19081586.0,
      "step": 15900
    },
    {
      "entropy": 1.8789875194430352,
      "epoch": 0.049319607954065495,
      "grad_norm": 6.301096439361572,
      "learning_rate": 7.705608834641092e-06,
      "loss": 0.618,
      "mean_token_accuracy": 0.8181715860962868,
      "num_tokens": 19093093.0,
      "step": 15910
    },
    {
      "entropy": 1.7927953436970712,
      "epoch": 0.04935060707911519,
      "grad_norm": 6.900021076202393,
      "learning_rate": 7.710452387871744e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8337122991681098,
      "num_tokens": 19105589.0,
      "step": 15920
    },
    {
      "entropy": 1.8942086696624756,
      "epoch": 0.04938160620416489,
      "grad_norm": 9.592936515808105,
      "learning_rate": 7.715295941102394e-06,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8135825991630554,
      "num_tokens": 19117515.0,
      "step": 15930
    },
    {
      "entropy": 1.8845902875065803,
      "epoch": 0.049412605329214584,
      "grad_norm": 5.827622890472412,
      "learning_rate": 7.720139494333044e-06,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8127764299511909,
      "num_tokens": 19129539.0,
      "step": 15940
    },
    {
      "entropy": 1.8030100539326668,
      "epoch": 0.04944360445426428,
      "grad_norm": 10.93111515045166,
      "learning_rate": 7.724983047563693e-06,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8306656986474991,
      "num_tokens": 19142450.0,
      "step": 15950
    },
    {
      "entropy": 1.8265103816986084,
      "epoch": 0.04947460357931397,
      "grad_norm": 4.990983486175537,
      "learning_rate": 7.729826600794343e-06,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8204995647072792,
      "num_tokens": 19154745.0,
      "step": 15960
    },
    {
      "entropy": 1.8918519958853721,
      "epoch": 0.04950560270436367,
      "grad_norm": 9.628335952758789,
      "learning_rate": 7.734670154024993e-06,
      "loss": 0.6261,
      "mean_token_accuracy": 0.8113234505057335,
      "num_tokens": 19166521.0,
      "step": 15970
    },
    {
      "entropy": 1.886893954873085,
      "epoch": 0.049536601829413364,
      "grad_norm": 11.60335922241211,
      "learning_rate": 7.739513707255643e-06,
      "loss": 0.612,
      "mean_token_accuracy": 0.8159346550703048,
      "num_tokens": 19177833.0,
      "step": 15980
    },
    {
      "entropy": 1.831916256248951,
      "epoch": 0.04956760095446306,
      "grad_norm": 9.314860343933105,
      "learning_rate": 7.744357260486293e-06,
      "loss": 0.5776,
      "mean_token_accuracy": 0.816097392141819,
      "num_tokens": 19189742.0,
      "step": 15990
    },
    {
      "entropy": 1.9015469133853913,
      "epoch": 0.04959860007951276,
      "grad_norm": 11.75980281829834,
      "learning_rate": 7.749200813716945e-06,
      "loss": 0.6715,
      "mean_token_accuracy": 0.8024095699191094,
      "num_tokens": 19200772.0,
      "step": 16000
    },
    {
      "entropy": 1.8420792520046234,
      "epoch": 0.049629599204562454,
      "grad_norm": 10.332155227661133,
      "learning_rate": 7.754044366947593e-06,
      "loss": 0.5571,
      "mean_token_accuracy": 0.8230875134468079,
      "num_tokens": 19212754.0,
      "step": 16010
    },
    {
      "entropy": 1.9134253069758416,
      "epoch": 0.04966059832961214,
      "grad_norm": 12.374646186828613,
      "learning_rate": 7.758887920178243e-06,
      "loss": 0.6448,
      "mean_token_accuracy": 0.8036070108413697,
      "num_tokens": 19224725.0,
      "step": 16020
    },
    {
      "entropy": 1.8665786176919936,
      "epoch": 0.04969159745466184,
      "grad_norm": 4.8785624504089355,
      "learning_rate": 7.763731473408892e-06,
      "loss": 0.6109,
      "mean_token_accuracy": 0.8155334115028381,
      "num_tokens": 19236500.0,
      "step": 16030
    },
    {
      "entropy": 1.827447460591793,
      "epoch": 0.04972259657971154,
      "grad_norm": 10.244511604309082,
      "learning_rate": 7.768575026639544e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8148609265685082,
      "num_tokens": 19249091.0,
      "step": 16040
    },
    {
      "entropy": 1.875395241379738,
      "epoch": 0.04975359570476123,
      "grad_norm": 11.924630165100098,
      "learning_rate": 7.773418579870194e-06,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8249801099300385,
      "num_tokens": 19260725.0,
      "step": 16050
    },
    {
      "entropy": 1.8245926171541214,
      "epoch": 0.04978459482981093,
      "grad_norm": 5.5314226150512695,
      "learning_rate": 7.778262133100844e-06,
      "loss": 0.5986,
      "mean_token_accuracy": 0.8085296213626861,
      "num_tokens": 19272815.0,
      "step": 16060
    },
    {
      "entropy": 1.879467526078224,
      "epoch": 0.049815593954860626,
      "grad_norm": 10.704133033752441,
      "learning_rate": 7.783105686331494e-06,
      "loss": 0.6219,
      "mean_token_accuracy": 0.8161226466298104,
      "num_tokens": 19284571.0,
      "step": 16070
    },
    {
      "entropy": 1.8190572828054428,
      "epoch": 0.04984659307991032,
      "grad_norm": 12.365442276000977,
      "learning_rate": 7.787949239562144e-06,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8116471886634826,
      "num_tokens": 19297247.0,
      "step": 16080
    },
    {
      "entropy": 1.8453327685594558,
      "epoch": 0.04987759220496001,
      "grad_norm": 11.743905067443848,
      "learning_rate": 7.792792792792793e-06,
      "loss": 0.5898,
      "mean_token_accuracy": 0.8130872502923012,
      "num_tokens": 19309118.0,
      "step": 16090
    },
    {
      "entropy": 1.8390842065215112,
      "epoch": 0.04990859133000971,
      "grad_norm": 11.728593826293945,
      "learning_rate": 7.797636346023443e-06,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8115570530295372,
      "num_tokens": 19321049.0,
      "step": 16100
    },
    {
      "entropy": 1.8872535049915313,
      "epoch": 0.049939590455059406,
      "grad_norm": 9.616110801696777,
      "learning_rate": 7.802479899254093e-06,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8136581242084503,
      "num_tokens": 19332620.0,
      "step": 16110
    },
    {
      "entropy": 1.9285971507430077,
      "epoch": 0.0499705895801091,
      "grad_norm": 10.771924018859863,
      "learning_rate": 7.807323452484743e-06,
      "loss": 0.6794,
      "mean_token_accuracy": 0.7982939928770065,
      "num_tokens": 19344422.0,
      "step": 16120
    },
    {
      "entropy": 1.8309599250555038,
      "epoch": 0.0500015887051588,
      "grad_norm": 11.928840637207031,
      "learning_rate": 7.812167005715393e-06,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8183940485119819,
      "num_tokens": 19356711.0,
      "step": 16130
    },
    {
      "entropy": 1.7304350897669791,
      "epoch": 0.050032587830208496,
      "grad_norm": 11.800362586975098,
      "learning_rate": 7.817010558946043e-06,
      "loss": 0.5798,
      "mean_token_accuracy": 0.8197820097208023,
      "num_tokens": 19370369.0,
      "step": 16140
    },
    {
      "entropy": 1.8792952716350555,
      "epoch": 0.050063586955258185,
      "grad_norm": 12.090376853942871,
      "learning_rate": 7.821854112176693e-06,
      "loss": 0.6114,
      "mean_token_accuracy": 0.8166804388165474,
      "num_tokens": 19381499.0,
      "step": 16150
    },
    {
      "entropy": 1.8935234278440476,
      "epoch": 0.05009458608030788,
      "grad_norm": 11.883420944213867,
      "learning_rate": 7.826697665407343e-06,
      "loss": 0.6121,
      "mean_token_accuracy": 0.8147160053253174,
      "num_tokens": 19392659.0,
      "step": 16160
    },
    {
      "entropy": 1.846675968170166,
      "epoch": 0.05012558520535758,
      "grad_norm": 11.219228744506836,
      "learning_rate": 7.831541218637994e-06,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8102226793766022,
      "num_tokens": 19404388.0,
      "step": 16170
    },
    {
      "entropy": 1.7764321342110634,
      "epoch": 0.050156584330407275,
      "grad_norm": 10.584576606750488,
      "learning_rate": 7.836384771868644e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.823691800236702,
      "num_tokens": 19417440.0,
      "step": 16180
    },
    {
      "entropy": 1.82427935898304,
      "epoch": 0.05018758345545697,
      "grad_norm": 5.237261772155762,
      "learning_rate": 7.841228325099294e-06,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8294490948319435,
      "num_tokens": 19429964.0,
      "step": 16190
    },
    {
      "entropy": 1.830063909292221,
      "epoch": 0.05021858258050667,
      "grad_norm": 10.078315734863281,
      "learning_rate": 7.846071878329944e-06,
      "loss": 0.5906,
      "mean_token_accuracy": 0.816213846206665,
      "num_tokens": 19441944.0,
      "step": 16200
    },
    {
      "entropy": 1.8124560460448265,
      "epoch": 0.05024958170555636,
      "grad_norm": 7.47139835357666,
      "learning_rate": 7.850915431560594e-06,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8125569224357605,
      "num_tokens": 19454488.0,
      "step": 16210
    },
    {
      "entropy": 1.8454759851098061,
      "epoch": 0.050280580830606054,
      "grad_norm": 11.923384666442871,
      "learning_rate": 7.855758984791244e-06,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8168366953730584,
      "num_tokens": 19466582.0,
      "step": 16220
    },
    {
      "entropy": 1.8621361300349235,
      "epoch": 0.05031157995565575,
      "grad_norm": 11.48657512664795,
      "learning_rate": 7.860602538021894e-06,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8186299324035644,
      "num_tokens": 19478249.0,
      "step": 16230
    },
    {
      "entropy": 1.8037197709083557,
      "epoch": 0.05034257908070545,
      "grad_norm": 5.228190898895264,
      "learning_rate": 7.865446091252543e-06,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8192895010113717,
      "num_tokens": 19489977.0,
      "step": 16240
    },
    {
      "entropy": 1.9085147067904473,
      "epoch": 0.050373578205755144,
      "grad_norm": 12.668362617492676,
      "learning_rate": 7.870289644483193e-06,
      "loss": 0.6271,
      "mean_token_accuracy": 0.8066832914948463,
      "num_tokens": 19501682.0,
      "step": 16250
    },
    {
      "entropy": 1.8951284110546112,
      "epoch": 0.05040457733080484,
      "grad_norm": 5.59218168258667,
      "learning_rate": 7.875133197713843e-06,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8141255229711533,
      "num_tokens": 19513281.0,
      "step": 16260
    },
    {
      "entropy": 1.8662694096565247,
      "epoch": 0.05043557645585453,
      "grad_norm": 11.334489822387695,
      "learning_rate": 7.879976750944493e-06,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8265201568603515,
      "num_tokens": 19525550.0,
      "step": 16270
    },
    {
      "entropy": 1.838777382671833,
      "epoch": 0.05046657558090423,
      "grad_norm": 12.854705810546875,
      "learning_rate": 7.884820304175143e-06,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8137482151389122,
      "num_tokens": 19537201.0,
      "step": 16280
    },
    {
      "entropy": 1.8109822481870652,
      "epoch": 0.050497574705953924,
      "grad_norm": 11.828731536865234,
      "learning_rate": 7.889663857405795e-06,
      "loss": 0.6238,
      "mean_token_accuracy": 0.8079608172178269,
      "num_tokens": 19550217.0,
      "step": 16290
    },
    {
      "entropy": 1.8977823466062547,
      "epoch": 0.05052857383100362,
      "grad_norm": 6.278931140899658,
      "learning_rate": 7.894507410636444e-06,
      "loss": 0.63,
      "mean_token_accuracy": 0.8086236611008644,
      "num_tokens": 19562455.0,
      "step": 16300
    },
    {
      "entropy": 1.885308313369751,
      "epoch": 0.05055957295605332,
      "grad_norm": 12.805869102478027,
      "learning_rate": 7.899350963867094e-06,
      "loss": 0.6222,
      "mean_token_accuracy": 0.8057803988456727,
      "num_tokens": 19573567.0,
      "step": 16310
    },
    {
      "entropy": 1.8054832592606544,
      "epoch": 0.050590572081103014,
      "grad_norm": 10.450759887695312,
      "learning_rate": 7.904194517097744e-06,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8225147247314453,
      "num_tokens": 19585912.0,
      "step": 16320
    },
    {
      "entropy": 1.836893168091774,
      "epoch": 0.0506215712061527,
      "grad_norm": 8.395682334899902,
      "learning_rate": 7.909038070328392e-06,
      "loss": 0.604,
      "mean_token_accuracy": 0.8111759915947914,
      "num_tokens": 19597962.0,
      "step": 16330
    },
    {
      "entropy": 1.8264792799949645,
      "epoch": 0.0506525703312024,
      "grad_norm": 9.940167427062988,
      "learning_rate": 7.913881623559044e-06,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8128548488020897,
      "num_tokens": 19610522.0,
      "step": 16340
    },
    {
      "entropy": 1.7342656940221786,
      "epoch": 0.050683569456252096,
      "grad_norm": 5.958490371704102,
      "learning_rate": 7.918725176789694e-06,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8350143611431122,
      "num_tokens": 19623800.0,
      "step": 16350
    },
    {
      "entropy": 1.8424322217702866,
      "epoch": 0.05071456858130179,
      "grad_norm": 10.20142936706543,
      "learning_rate": 7.923568730020344e-06,
      "loss": 0.6181,
      "mean_token_accuracy": 0.8129007190465927,
      "num_tokens": 19635543.0,
      "step": 16360
    },
    {
      "entropy": 1.8671317547559738,
      "epoch": 0.05074556770635149,
      "grad_norm": 11.634441375732422,
      "learning_rate": 7.928412283250994e-06,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8180768474936485,
      "num_tokens": 19646796.0,
      "step": 16370
    },
    {
      "entropy": 1.8461050242185593,
      "epoch": 0.050776566831401186,
      "grad_norm": 5.321835041046143,
      "learning_rate": 7.933255836481643e-06,
      "loss": 0.6055,
      "mean_token_accuracy": 0.8090061485767365,
      "num_tokens": 19658855.0,
      "step": 16380
    },
    {
      "entropy": 1.8895393520593644,
      "epoch": 0.050807565956450876,
      "grad_norm": 12.092207908630371,
      "learning_rate": 7.938099389712293e-06,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8007647573947907,
      "num_tokens": 19670438.0,
      "step": 16390
    },
    {
      "entropy": 1.93761787712574,
      "epoch": 0.05083856508150057,
      "grad_norm": 13.244606971740723,
      "learning_rate": 7.942942942942943e-06,
      "loss": 0.6953,
      "mean_token_accuracy": 0.7985577821731568,
      "num_tokens": 19681443.0,
      "step": 16400
    },
    {
      "entropy": 1.8733505114912987,
      "epoch": 0.05086956420655027,
      "grad_norm": 10.528936386108398,
      "learning_rate": 7.947786496173593e-06,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8295532435178756,
      "num_tokens": 19693237.0,
      "step": 16410
    },
    {
      "entropy": 1.8836349695920944,
      "epoch": 0.050900563331599966,
      "grad_norm": 9.953451156616211,
      "learning_rate": 7.952630049404245e-06,
      "loss": 0.5955,
      "mean_token_accuracy": 0.8276330158114433,
      "num_tokens": 19704482.0,
      "step": 16420
    },
    {
      "entropy": 1.8280857503414154,
      "epoch": 0.05093156245664966,
      "grad_norm": 11.806560516357422,
      "learning_rate": 7.957473602634893e-06,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8197454378008843,
      "num_tokens": 19716072.0,
      "step": 16430
    },
    {
      "entropy": 1.868168619275093,
      "epoch": 0.05096256158169936,
      "grad_norm": 14.613922119140625,
      "learning_rate": 7.962317155865543e-06,
      "loss": 0.6294,
      "mean_token_accuracy": 0.8104723706841469,
      "num_tokens": 19727908.0,
      "step": 16440
    },
    {
      "entropy": 1.8337607711553574,
      "epoch": 0.050993560706749055,
      "grad_norm": 11.018813133239746,
      "learning_rate": 7.967160709096193e-06,
      "loss": 0.5722,
      "mean_token_accuracy": 0.815605454146862,
      "num_tokens": 19740124.0,
      "step": 16450
    },
    {
      "entropy": 1.8944316014647484,
      "epoch": 0.051024559831798745,
      "grad_norm": 12.068158149719238,
      "learning_rate": 7.972004262326844e-06,
      "loss": 0.611,
      "mean_token_accuracy": 0.8088446974754333,
      "num_tokens": 19752042.0,
      "step": 16460
    },
    {
      "entropy": 1.8141653031110763,
      "epoch": 0.05105555895684844,
      "grad_norm": 11.938932418823242,
      "learning_rate": 7.976847815557494e-06,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8189818143844605,
      "num_tokens": 19764559.0,
      "step": 16470
    },
    {
      "entropy": 1.8091833949089051,
      "epoch": 0.05108655808189814,
      "grad_norm": 6.31152868270874,
      "learning_rate": 7.981691368788144e-06,
      "loss": 0.6014,
      "mean_token_accuracy": 0.8165456935763359,
      "num_tokens": 19776260.0,
      "step": 16480
    },
    {
      "entropy": 1.812686663866043,
      "epoch": 0.051117557206947835,
      "grad_norm": 5.221765995025635,
      "learning_rate": 7.986534922018794e-06,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8260109111666679,
      "num_tokens": 19788934.0,
      "step": 16490
    },
    {
      "entropy": 1.8625101804733277,
      "epoch": 0.05114855633199753,
      "grad_norm": 11.6726713180542,
      "learning_rate": 7.991378475249444e-06,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8176575809717178,
      "num_tokens": 19800717.0,
      "step": 16500
    },
    {
      "entropy": 1.8866008058190347,
      "epoch": 0.05117955545704723,
      "grad_norm": 13.77607250213623,
      "learning_rate": 7.996222028480094e-06,
      "loss": 0.6522,
      "mean_token_accuracy": 0.7992734283208847,
      "num_tokens": 19812273.0,
      "step": 16510
    },
    {
      "entropy": 1.9345821738243103,
      "epoch": 0.05121055458209692,
      "grad_norm": 10.338969230651855,
      "learning_rate": 8.001065581710744e-06,
      "loss": 0.684,
      "mean_token_accuracy": 0.7988832354545593,
      "num_tokens": 19822905.0,
      "step": 16520
    },
    {
      "entropy": 1.9000392645597457,
      "epoch": 0.051241553707146614,
      "grad_norm": 10.453362464904785,
      "learning_rate": 8.005909134941393e-06,
      "loss": 0.6257,
      "mean_token_accuracy": 0.8212269991636276,
      "num_tokens": 19833662.0,
      "step": 16530
    },
    {
      "entropy": 1.8661847695708276,
      "epoch": 0.05127255283219631,
      "grad_norm": 10.485684394836426,
      "learning_rate": 8.010752688172043e-06,
      "loss": 0.643,
      "mean_token_accuracy": 0.8103910699486733,
      "num_tokens": 19845416.0,
      "step": 16540
    },
    {
      "entropy": 1.8219714432954788,
      "epoch": 0.05130355195724601,
      "grad_norm": 12.01816463470459,
      "learning_rate": 8.015596241402693e-06,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8123174890875816,
      "num_tokens": 19857098.0,
      "step": 16550
    },
    {
      "entropy": 1.8485499978065492,
      "epoch": 0.051334551082295704,
      "grad_norm": 9.451250076293945,
      "learning_rate": 8.020439794633343e-06,
      "loss": 0.6128,
      "mean_token_accuracy": 0.8077960431575775,
      "num_tokens": 19869230.0,
      "step": 16560
    },
    {
      "entropy": 1.8622397229075431,
      "epoch": 0.0513655502073454,
      "grad_norm": 10.47313404083252,
      "learning_rate": 8.025283347863993e-06,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8121888309717178,
      "num_tokens": 19881415.0,
      "step": 16570
    },
    {
      "entropy": 1.763058114051819,
      "epoch": 0.05139654933239509,
      "grad_norm": 5.442761421203613,
      "learning_rate": 8.030126901094643e-06,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8221294581890106,
      "num_tokens": 19894415.0,
      "step": 16580
    },
    {
      "entropy": 1.7793406203389168,
      "epoch": 0.05142754845744479,
      "grad_norm": 12.98482894897461,
      "learning_rate": 8.034970454325294e-06,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8247652933001518,
      "num_tokens": 19907687.0,
      "step": 16590
    },
    {
      "entropy": 1.874410592019558,
      "epoch": 0.051458547582494484,
      "grad_norm": 4.950082778930664,
      "learning_rate": 8.039814007555944e-06,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8031128868460655,
      "num_tokens": 19919926.0,
      "step": 16600
    },
    {
      "entropy": 1.8641373217105865,
      "epoch": 0.05148954670754418,
      "grad_norm": 11.791984558105469,
      "learning_rate": 8.044657560786594e-06,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8170140698552132,
      "num_tokens": 19932297.0,
      "step": 16610
    },
    {
      "entropy": 1.8777590736746788,
      "epoch": 0.05152054583259388,
      "grad_norm": 14.65937614440918,
      "learning_rate": 8.049501114017244e-06,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8093700706958771,
      "num_tokens": 19944296.0,
      "step": 16620
    },
    {
      "entropy": 1.83215441852808,
      "epoch": 0.05155154495764357,
      "grad_norm": 18.492868423461914,
      "learning_rate": 8.054344667247894e-06,
      "loss": 0.5367,
      "mean_token_accuracy": 0.83001299649477,
      "num_tokens": 19956294.0,
      "step": 16630
    },
    {
      "entropy": 1.8398912638425826,
      "epoch": 0.05158254408269326,
      "grad_norm": 12.522979736328125,
      "learning_rate": 8.059188220478544e-06,
      "loss": 0.615,
      "mean_token_accuracy": 0.8146472930908203,
      "num_tokens": 19968456.0,
      "step": 16640
    },
    {
      "entropy": 1.8897781684994697,
      "epoch": 0.05161354320774296,
      "grad_norm": 11.673059463500977,
      "learning_rate": 8.064031773709194e-06,
      "loss": 0.6458,
      "mean_token_accuracy": 0.8108509004116058,
      "num_tokens": 19980146.0,
      "step": 16650
    },
    {
      "entropy": 1.9111702471971512,
      "epoch": 0.051644542332792656,
      "grad_norm": 11.109082221984863,
      "learning_rate": 8.068875326939844e-06,
      "loss": 0.633,
      "mean_token_accuracy": 0.8093647316098214,
      "num_tokens": 19992383.0,
      "step": 16660
    },
    {
      "entropy": 1.9211183845996858,
      "epoch": 0.05167554145784235,
      "grad_norm": 11.176039695739746,
      "learning_rate": 8.073718880170494e-06,
      "loss": 0.6448,
      "mean_token_accuracy": 0.8067373007535934,
      "num_tokens": 20002938.0,
      "step": 16670
    },
    {
      "entropy": 1.8170073732733727,
      "epoch": 0.05170654058289205,
      "grad_norm": 11.94314193725586,
      "learning_rate": 8.078562433401143e-06,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8301214516162873,
      "num_tokens": 20014929.0,
      "step": 16680
    },
    {
      "entropy": 1.8377844780683517,
      "epoch": 0.051737539707941746,
      "grad_norm": 9.638590812683105,
      "learning_rate": 8.083405986631793e-06,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8208416193723679,
      "num_tokens": 20026901.0,
      "step": 16690
    },
    {
      "entropy": 1.866816473007202,
      "epoch": 0.051768538832991436,
      "grad_norm": 10.777813911437988,
      "learning_rate": 8.088249539862443e-06,
      "loss": 0.639,
      "mean_token_accuracy": 0.8086013346910477,
      "num_tokens": 20038260.0,
      "step": 16700
    },
    {
      "entropy": 1.8705125212669373,
      "epoch": 0.05179953795804113,
      "grad_norm": 12.59802532196045,
      "learning_rate": 8.093093093093095e-06,
      "loss": 0.5945,
      "mean_token_accuracy": 0.8204304948449135,
      "num_tokens": 20049362.0,
      "step": 16710
    },
    {
      "entropy": 1.81406359821558,
      "epoch": 0.05183053708309083,
      "grad_norm": 14.44949722290039,
      "learning_rate": 8.097936646323745e-06,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8081631407141685,
      "num_tokens": 20062608.0,
      "step": 16720
    },
    {
      "entropy": 1.9008635878562927,
      "epoch": 0.051861536208140525,
      "grad_norm": 9.96707534790039,
      "learning_rate": 8.102780199554395e-06,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8111796498298645,
      "num_tokens": 20074235.0,
      "step": 16730
    },
    {
      "entropy": 1.8784829795360565,
      "epoch": 0.05189253533319022,
      "grad_norm": 11.282867431640625,
      "learning_rate": 8.107623752785044e-06,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8110943511128426,
      "num_tokens": 20086392.0,
      "step": 16740
    },
    {
      "entropy": 1.915185084939003,
      "epoch": 0.05192353445823992,
      "grad_norm": 10.151687622070312,
      "learning_rate": 8.112467306015693e-06,
      "loss": 0.6498,
      "mean_token_accuracy": 0.8103260189294815,
      "num_tokens": 20097631.0,
      "step": 16750
    },
    {
      "entropy": 1.791477920114994,
      "epoch": 0.05195453358328961,
      "grad_norm": 5.157210350036621,
      "learning_rate": 8.117310859246344e-06,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8181603282690049,
      "num_tokens": 20111249.0,
      "step": 16760
    },
    {
      "entropy": 1.8779030591249466,
      "epoch": 0.051985532708339305,
      "grad_norm": 10.711991310119629,
      "learning_rate": 8.122154412476994e-06,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8104298338294029,
      "num_tokens": 20122999.0,
      "step": 16770
    },
    {
      "entropy": 1.8585640251636506,
      "epoch": 0.052016531833389,
      "grad_norm": 5.962745666503906,
      "learning_rate": 8.126997965707644e-06,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8122934922575951,
      "num_tokens": 20135004.0,
      "step": 16780
    },
    {
      "entropy": 1.8683840811252594,
      "epoch": 0.0520475309584387,
      "grad_norm": 10.104706764221191,
      "learning_rate": 8.131841518938294e-06,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8203515037894249,
      "num_tokens": 20146845.0,
      "step": 16790
    },
    {
      "entropy": 1.8777433142066002,
      "epoch": 0.052078530083488395,
      "grad_norm": 11.71458625793457,
      "learning_rate": 8.136685072168944e-06,
      "loss": 0.6173,
      "mean_token_accuracy": 0.816298334300518,
      "num_tokens": 20157635.0,
      "step": 16800
    },
    {
      "entropy": 1.8254910349845885,
      "epoch": 0.05210952920853809,
      "grad_norm": 12.500772476196289,
      "learning_rate": 8.141528625399594e-06,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8122817128896713,
      "num_tokens": 20169780.0,
      "step": 16810
    },
    {
      "entropy": 1.7841567173600197,
      "epoch": 0.05214052833358779,
      "grad_norm": 5.604972839355469,
      "learning_rate": 8.146372178630243e-06,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8187017843127251,
      "num_tokens": 20182608.0,
      "step": 16820
    },
    {
      "entropy": 1.8085433050990105,
      "epoch": 0.05217152745863748,
      "grad_norm": 12.112675666809082,
      "learning_rate": 8.151215731860895e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.819066570699215,
      "num_tokens": 20194982.0,
      "step": 16830
    },
    {
      "entropy": 1.8413061544299125,
      "epoch": 0.052202526583687174,
      "grad_norm": 12.422449111938477,
      "learning_rate": 8.156059285091545e-06,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8272005066275596,
      "num_tokens": 20206354.0,
      "step": 16840
    },
    {
      "entropy": 1.8800123199820518,
      "epoch": 0.05223352570873687,
      "grad_norm": 5.7091779708862305,
      "learning_rate": 8.160902838322193e-06,
      "loss": 0.6096,
      "mean_token_accuracy": 0.8087597385048866,
      "num_tokens": 20218011.0,
      "step": 16850
    },
    {
      "entropy": 1.7969383016228675,
      "epoch": 0.05226452483378657,
      "grad_norm": 10.848820686340332,
      "learning_rate": 8.165746391552843e-06,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8236014172434807,
      "num_tokens": 20230550.0,
      "step": 16860
    },
    {
      "entropy": 1.8589994862675667,
      "epoch": 0.052295523958836264,
      "grad_norm": 10.558844566345215,
      "learning_rate": 8.170589944783493e-06,
      "loss": 0.56,
      "mean_token_accuracy": 0.8191485345363617,
      "num_tokens": 20242485.0,
      "step": 16870
    },
    {
      "entropy": 1.8285365521907806,
      "epoch": 0.05232652308388596,
      "grad_norm": 10.265873908996582,
      "learning_rate": 8.175433498014144e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.827988238632679,
      "num_tokens": 20254416.0,
      "step": 16880
    },
    {
      "entropy": 1.7561014324426651,
      "epoch": 0.05235752220893565,
      "grad_norm": 10.504303932189941,
      "learning_rate": 8.180277051244794e-06,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8308162286877632,
      "num_tokens": 20267582.0,
      "step": 16890
    },
    {
      "entropy": 1.8213954389095306,
      "epoch": 0.05238852133398535,
      "grad_norm": 11.907408714294434,
      "learning_rate": 8.185120604475444e-06,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8134548336267471,
      "num_tokens": 20279358.0,
      "step": 16900
    },
    {
      "entropy": 1.855673785507679,
      "epoch": 0.05241952045903504,
      "grad_norm": 12.381454467773438,
      "learning_rate": 8.189964157706094e-06,
      "loss": 0.6341,
      "mean_token_accuracy": 0.8102763295173645,
      "num_tokens": 20291624.0,
      "step": 16910
    },
    {
      "entropy": 1.8303613483905792,
      "epoch": 0.05245051958408474,
      "grad_norm": 11.09730052947998,
      "learning_rate": 8.194807710936744e-06,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8215647727251053,
      "num_tokens": 20304029.0,
      "step": 16920
    },
    {
      "entropy": 1.7606044977903366,
      "epoch": 0.052481518709134437,
      "grad_norm": 14.970794677734375,
      "learning_rate": 8.199651264167394e-06,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8292558521032334,
      "num_tokens": 20317399.0,
      "step": 16930
    },
    {
      "entropy": 1.9114146530628204,
      "epoch": 0.05251251783418413,
      "grad_norm": 11.057762145996094,
      "learning_rate": 8.204494817398044e-06,
      "loss": 0.6605,
      "mean_token_accuracy": 0.8093030959367752,
      "num_tokens": 20329008.0,
      "step": 16940
    },
    {
      "entropy": 1.9093146950006485,
      "epoch": 0.05254351695923382,
      "grad_norm": 10.629437446594238,
      "learning_rate": 8.209338370628694e-06,
      "loss": 0.6731,
      "mean_token_accuracy": 0.8056405037641525,
      "num_tokens": 20340231.0,
      "step": 16950
    },
    {
      "entropy": 1.8436574935913086,
      "epoch": 0.05257451608428352,
      "grad_norm": 11.923999786376953,
      "learning_rate": 8.214181923859344e-06,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8199576199054718,
      "num_tokens": 20352072.0,
      "step": 16960
    },
    {
      "entropy": 1.8853593990206718,
      "epoch": 0.052605515209333216,
      "grad_norm": 10.41568660736084,
      "learning_rate": 8.219025477089993e-06,
      "loss": 0.6458,
      "mean_token_accuracy": 0.8017956003546715,
      "num_tokens": 20363499.0,
      "step": 16970
    },
    {
      "entropy": 1.918022060394287,
      "epoch": 0.05263651433438291,
      "grad_norm": 8.378069877624512,
      "learning_rate": 8.223869030320643e-06,
      "loss": 0.6415,
      "mean_token_accuracy": 0.8088062942028046,
      "num_tokens": 20375464.0,
      "step": 16980
    },
    {
      "entropy": 1.8862889647483825,
      "epoch": 0.05266751345943261,
      "grad_norm": 10.294898986816406,
      "learning_rate": 8.228712583551293e-06,
      "loss": 0.636,
      "mean_token_accuracy": 0.8037711903452873,
      "num_tokens": 20387540.0,
      "step": 16990
    },
    {
      "entropy": 1.9208069905638694,
      "epoch": 0.052698512584482306,
      "grad_norm": 10.81685733795166,
      "learning_rate": 8.233556136781943e-06,
      "loss": 0.7022,
      "mean_token_accuracy": 0.8014640405774116,
      "num_tokens": 20399798.0,
      "step": 17000
    },
    {
      "entropy": 1.8465980917215348,
      "epoch": 0.052729511709531995,
      "grad_norm": 12.846884727478027,
      "learning_rate": 8.238399690012595e-06,
      "loss": 0.592,
      "mean_token_accuracy": 0.820027408003807,
      "num_tokens": 20411783.0,
      "step": 17010
    },
    {
      "entropy": 1.8712544098496438,
      "epoch": 0.05276051083458169,
      "grad_norm": 12.44199275970459,
      "learning_rate": 8.243243243243245e-06,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8144369840621948,
      "num_tokens": 20423720.0,
      "step": 17020
    },
    {
      "entropy": 1.861243399977684,
      "epoch": 0.05279150995963139,
      "grad_norm": 10.822843551635742,
      "learning_rate": 8.248086796473894e-06,
      "loss": 0.6249,
      "mean_token_accuracy": 0.8173502340912819,
      "num_tokens": 20435655.0,
      "step": 17030
    },
    {
      "entropy": 1.831656913459301,
      "epoch": 0.052822509084681085,
      "grad_norm": 10.160820960998535,
      "learning_rate": 8.252930349704544e-06,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8063213124871254,
      "num_tokens": 20449025.0,
      "step": 17040
    },
    {
      "entropy": 1.78798815459013,
      "epoch": 0.05285350820973078,
      "grad_norm": 5.9856414794921875,
      "learning_rate": 8.257773902935194e-06,
      "loss": 0.5513,
      "mean_token_accuracy": 0.8148110911250115,
      "num_tokens": 20461447.0,
      "step": 17050
    },
    {
      "entropy": 1.86494023501873,
      "epoch": 0.05288450733478048,
      "grad_norm": 9.526678085327148,
      "learning_rate": 8.262617456165844e-06,
      "loss": 0.6113,
      "mean_token_accuracy": 0.815309202671051,
      "num_tokens": 20472983.0,
      "step": 17060
    },
    {
      "entropy": 1.8194462105631828,
      "epoch": 0.05291550645983017,
      "grad_norm": 5.433764934539795,
      "learning_rate": 8.267461009396494e-06,
      "loss": 0.5941,
      "mean_token_accuracy": 0.8140083074569702,
      "num_tokens": 20485617.0,
      "step": 17070
    },
    {
      "entropy": 1.876285555958748,
      "epoch": 0.052946505584879865,
      "grad_norm": 5.180232524871826,
      "learning_rate": 8.272304562627144e-06,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8107918843626976,
      "num_tokens": 20496764.0,
      "step": 17080
    },
    {
      "entropy": 1.7992719858884811,
      "epoch": 0.05297750470992956,
      "grad_norm": 10.29720687866211,
      "learning_rate": 8.277148115857794e-06,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8232490435242653,
      "num_tokens": 20509654.0,
      "step": 17090
    },
    {
      "entropy": 1.8305759117007256,
      "epoch": 0.05300850383497926,
      "grad_norm": 11.629636764526367,
      "learning_rate": 8.281991669088444e-06,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8326339006423951,
      "num_tokens": 20521330.0,
      "step": 17100
    },
    {
      "entropy": 1.8921547949314117,
      "epoch": 0.053039502960028954,
      "grad_norm": 10.69015884399414,
      "learning_rate": 8.286835222319093e-06,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8138958439230919,
      "num_tokens": 20532830.0,
      "step": 17110
    },
    {
      "entropy": 1.8374129354953765,
      "epoch": 0.05307050208507865,
      "grad_norm": 14.065573692321777,
      "learning_rate": 8.291678775549743e-06,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8190833449363708,
      "num_tokens": 20545189.0,
      "step": 17120
    },
    {
      "entropy": 1.8749100968241692,
      "epoch": 0.05310150121012834,
      "grad_norm": 11.14418888092041,
      "learning_rate": 8.296522328780395e-06,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8282869830727577,
      "num_tokens": 20557028.0,
      "step": 17130
    },
    {
      "entropy": 1.9455086424946786,
      "epoch": 0.05313250033517804,
      "grad_norm": 6.915708065032959,
      "learning_rate": 8.301365882011045e-06,
      "loss": 0.6958,
      "mean_token_accuracy": 0.7915912076830864,
      "num_tokens": 20567980.0,
      "step": 17140
    },
    {
      "entropy": 1.962218326330185,
      "epoch": 0.053163499460227734,
      "grad_norm": 12.134001731872559,
      "learning_rate": 8.306209435241695e-06,
      "loss": 0.6929,
      "mean_token_accuracy": 0.8000232353806496,
      "num_tokens": 20578839.0,
      "step": 17150
    },
    {
      "entropy": 1.845068359375,
      "epoch": 0.05319449858527743,
      "grad_norm": 11.644062995910645,
      "learning_rate": 8.311052988472345e-06,
      "loss": 0.6593,
      "mean_token_accuracy": 0.8096413478255272,
      "num_tokens": 20590558.0,
      "step": 17160
    },
    {
      "entropy": 1.8619411289691925,
      "epoch": 0.05322549771032713,
      "grad_norm": 10.61169719696045,
      "learning_rate": 8.315896541702993e-06,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8225790098309517,
      "num_tokens": 20601969.0,
      "step": 17170
    },
    {
      "entropy": 1.8861952975392342,
      "epoch": 0.053256496835376824,
      "grad_norm": 5.271222114562988,
      "learning_rate": 8.320740094933644e-06,
      "loss": 0.6097,
      "mean_token_accuracy": 0.8090833589434624,
      "num_tokens": 20614378.0,
      "step": 17180
    },
    {
      "entropy": 1.7966760843992233,
      "epoch": 0.05328749596042652,
      "grad_norm": 10.908940315246582,
      "learning_rate": 8.325583648164294e-06,
      "loss": 0.5047,
      "mean_token_accuracy": 0.8201152965426445,
      "num_tokens": 20628222.0,
      "step": 17190
    },
    {
      "entropy": 1.8242161065340041,
      "epoch": 0.05331849508547621,
      "grad_norm": 5.578505992889404,
      "learning_rate": 8.330427201394944e-06,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8250388026237487,
      "num_tokens": 20641356.0,
      "step": 17200
    },
    {
      "entropy": 1.9080461800098418,
      "epoch": 0.053349494210525907,
      "grad_norm": 10.300594329833984,
      "learning_rate": 8.335270754625594e-06,
      "loss": 0.6218,
      "mean_token_accuracy": 0.8083810389041901,
      "num_tokens": 20653007.0,
      "step": 17210
    },
    {
      "entropy": 1.9093635857105256,
      "epoch": 0.0533804933355756,
      "grad_norm": 11.184460639953613,
      "learning_rate": 8.340114307856244e-06,
      "loss": 0.6377,
      "mean_token_accuracy": 0.8104760378599167,
      "num_tokens": 20663792.0,
      "step": 17220
    },
    {
      "entropy": 1.9165063306689263,
      "epoch": 0.0534114924606253,
      "grad_norm": 10.442662239074707,
      "learning_rate": 8.344957861086894e-06,
      "loss": 0.5942,
      "mean_token_accuracy": 0.8143234044313431,
      "num_tokens": 20675785.0,
      "step": 17230
    },
    {
      "entropy": 1.9359146520495414,
      "epoch": 0.053442491585674996,
      "grad_norm": 8.756479263305664,
      "learning_rate": 8.349801414317544e-06,
      "loss": 0.6133,
      "mean_token_accuracy": 0.8071959316730499,
      "num_tokens": 20687331.0,
      "step": 17240
    },
    {
      "entropy": 1.945779299736023,
      "epoch": 0.05347349071072469,
      "grad_norm": 11.406476020812988,
      "learning_rate": 8.354644967548195e-06,
      "loss": 0.6008,
      "mean_token_accuracy": 0.8256149128079414,
      "num_tokens": 20697727.0,
      "step": 17250
    },
    {
      "entropy": 1.8896187216043472,
      "epoch": 0.05350448983577438,
      "grad_norm": 12.807340621948242,
      "learning_rate": 8.359488520778845e-06,
      "loss": 0.6007,
      "mean_token_accuracy": 0.8113436698913574,
      "num_tokens": 20709839.0,
      "step": 17260
    },
    {
      "entropy": 1.8658792898058891,
      "epoch": 0.05353548896082408,
      "grad_norm": 11.434633255004883,
      "learning_rate": 8.364332074009493e-06,
      "loss": 0.6164,
      "mean_token_accuracy": 0.806533083319664,
      "num_tokens": 20721529.0,
      "step": 17270
    },
    {
      "entropy": 1.8530214801430702,
      "epoch": 0.053566488085873776,
      "grad_norm": 10.531983375549316,
      "learning_rate": 8.369175627240143e-06,
      "loss": 0.5375,
      "mean_token_accuracy": 0.819164864718914,
      "num_tokens": 20734410.0,
      "step": 17280
    },
    {
      "entropy": 1.864107219874859,
      "epoch": 0.05359748721092347,
      "grad_norm": 11.01248550415039,
      "learning_rate": 8.374019180470793e-06,
      "loss": 0.6512,
      "mean_token_accuracy": 0.8096802055835723,
      "num_tokens": 20746182.0,
      "step": 17290
    },
    {
      "entropy": 1.8512002035975457,
      "epoch": 0.05362848633597317,
      "grad_norm": 10.909806251525879,
      "learning_rate": 8.378862733701445e-06,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8117505341768265,
      "num_tokens": 20758266.0,
      "step": 17300
    },
    {
      "entropy": 1.9608330637216569,
      "epoch": 0.053659485461022866,
      "grad_norm": 9.702219009399414,
      "learning_rate": 8.383706286932095e-06,
      "loss": 0.6645,
      "mean_token_accuracy": 0.8040051057934761,
      "num_tokens": 20768751.0,
      "step": 17310
    },
    {
      "entropy": 1.9042405292391777,
      "epoch": 0.053690484586072555,
      "grad_norm": 10.729706764221191,
      "learning_rate": 8.388549840162744e-06,
      "loss": 0.6057,
      "mean_token_accuracy": 0.817810270190239,
      "num_tokens": 20780886.0,
      "step": 17320
    },
    {
      "entropy": 1.8193862006068229,
      "epoch": 0.05372148371112225,
      "grad_norm": 5.850473880767822,
      "learning_rate": 8.393393393393394e-06,
      "loss": 0.5504,
      "mean_token_accuracy": 0.8224448531866073,
      "num_tokens": 20793098.0,
      "step": 17330
    },
    {
      "entropy": 1.7956085950136185,
      "epoch": 0.05375248283617195,
      "grad_norm": 4.813007354736328,
      "learning_rate": 8.398236946624044e-06,
      "loss": 0.5571,
      "mean_token_accuracy": 0.818726347386837,
      "num_tokens": 20806408.0,
      "step": 17340
    },
    {
      "entropy": 1.8429106876254082,
      "epoch": 0.053783481961221645,
      "grad_norm": 4.898284435272217,
      "learning_rate": 8.403080499854694e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.828820888698101,
      "num_tokens": 20819353.0,
      "step": 17350
    },
    {
      "entropy": 1.8711024463176726,
      "epoch": 0.05381448108627134,
      "grad_norm": 6.471118927001953,
      "learning_rate": 8.407924053085344e-06,
      "loss": 0.588,
      "mean_token_accuracy": 0.8236263871192933,
      "num_tokens": 20830636.0,
      "step": 17360
    },
    {
      "entropy": 1.8240604743361473,
      "epoch": 0.05384548021132104,
      "grad_norm": 10.193392753601074,
      "learning_rate": 8.412767606315994e-06,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8229044646024704,
      "num_tokens": 20843265.0,
      "step": 17370
    },
    {
      "entropy": 1.9022013053297997,
      "epoch": 0.05387647933637073,
      "grad_norm": 11.367668151855469,
      "learning_rate": 8.417611159546644e-06,
      "loss": 0.6635,
      "mean_token_accuracy": 0.8005659952759743,
      "num_tokens": 20855224.0,
      "step": 17380
    },
    {
      "entropy": 1.874705444276333,
      "epoch": 0.053907478461420424,
      "grad_norm": 13.563992500305176,
      "learning_rate": 8.422454712777294e-06,
      "loss": 0.6151,
      "mean_token_accuracy": 0.8215208485722542,
      "num_tokens": 20866903.0,
      "step": 17390
    },
    {
      "entropy": 1.9510533064603806,
      "epoch": 0.05393847758647012,
      "grad_norm": 12.109170913696289,
      "learning_rate": 8.427298266007944e-06,
      "loss": 0.6787,
      "mean_token_accuracy": 0.793503324687481,
      "num_tokens": 20878367.0,
      "step": 17400
    },
    {
      "entropy": 1.877579514682293,
      "epoch": 0.05396947671151982,
      "grad_norm": 10.051459312438965,
      "learning_rate": 8.432141819238593e-06,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8256360620260239,
      "num_tokens": 20890360.0,
      "step": 17410
    },
    {
      "entropy": 1.9146401450037955,
      "epoch": 0.054000475836569514,
      "grad_norm": 10.120962142944336,
      "learning_rate": 8.436985372469243e-06,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8045231074094772,
      "num_tokens": 20902287.0,
      "step": 17420
    },
    {
      "entropy": 1.8735201194882394,
      "epoch": 0.05403147496161921,
      "grad_norm": 5.3620524406433105,
      "learning_rate": 8.441828925699895e-06,
      "loss": 0.6369,
      "mean_token_accuracy": 0.811458395421505,
      "num_tokens": 20913921.0,
      "step": 17430
    },
    {
      "entropy": 1.8790420606732368,
      "epoch": 0.0540624740866689,
      "grad_norm": 11.361139297485352,
      "learning_rate": 8.446672478930545e-06,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8240441083908081,
      "num_tokens": 20925521.0,
      "step": 17440
    },
    {
      "entropy": 1.8804828137159348,
      "epoch": 0.0540934732117186,
      "grad_norm": 12.176139831542969,
      "learning_rate": 8.451516032161195e-06,
      "loss": 0.5856,
      "mean_token_accuracy": 0.8197601407766342,
      "num_tokens": 20937530.0,
      "step": 17450
    },
    {
      "entropy": 1.8536333724856378,
      "epoch": 0.054124472336768294,
      "grad_norm": 8.680792808532715,
      "learning_rate": 8.456359585391845e-06,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8246132522821427,
      "num_tokens": 20949885.0,
      "step": 17460
    },
    {
      "entropy": 1.9130322173237801,
      "epoch": 0.05415547146181799,
      "grad_norm": 9.968182563781738,
      "learning_rate": 8.461203138622494e-06,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8138394117355346,
      "num_tokens": 20961126.0,
      "step": 17470
    },
    {
      "entropy": 1.8640489429235458,
      "epoch": 0.05418647058686769,
      "grad_norm": 14.096504211425781,
      "learning_rate": 8.466046691853144e-06,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8181562379002572,
      "num_tokens": 20973024.0,
      "step": 17480
    },
    {
      "entropy": 1.7738539189100266,
      "epoch": 0.05421746971191738,
      "grad_norm": 10.170340538024902,
      "learning_rate": 8.470890245083794e-06,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8179555460810661,
      "num_tokens": 20985918.0,
      "step": 17490
    },
    {
      "entropy": 1.8802258223295212,
      "epoch": 0.05424846883696708,
      "grad_norm": 10.039809226989746,
      "learning_rate": 8.475733798314444e-06,
      "loss": 0.6347,
      "mean_token_accuracy": 0.8079760164022446,
      "num_tokens": 20998102.0,
      "step": 17500
    },
    {
      "entropy": 1.9156350553035737,
      "epoch": 0.05427946796201677,
      "grad_norm": 9.540949821472168,
      "learning_rate": 8.480577351545094e-06,
      "loss": 0.6298,
      "mean_token_accuracy": 0.8177491262555122,
      "num_tokens": 21009239.0,
      "step": 17510
    },
    {
      "entropy": 1.804891037940979,
      "epoch": 0.054310467087066466,
      "grad_norm": 9.833329200744629,
      "learning_rate": 8.485420904775744e-06,
      "loss": 0.6428,
      "mean_token_accuracy": 0.808480116724968,
      "num_tokens": 21021894.0,
      "step": 17520
    },
    {
      "entropy": 1.9138577803969383,
      "epoch": 0.05434146621211616,
      "grad_norm": 6.600327968597412,
      "learning_rate": 8.490264458006394e-06,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8139492854475975,
      "num_tokens": 21033383.0,
      "step": 17530
    },
    {
      "entropy": 1.7570361971855164,
      "epoch": 0.05437246533716586,
      "grad_norm": 6.329018592834473,
      "learning_rate": 8.495108011237044e-06,
      "loss": 0.5523,
      "mean_token_accuracy": 0.818885837495327,
      "num_tokens": 21047301.0,
      "step": 17540
    },
    {
      "entropy": 1.9657986283302307,
      "epoch": 0.054403464462215556,
      "grad_norm": 10.000340461730957,
      "learning_rate": 8.499951564467695e-06,
      "loss": 0.7018,
      "mean_token_accuracy": 0.7982934132218361,
      "num_tokens": 21058156.0,
      "step": 17550
    },
    {
      "entropy": 1.881192384660244,
      "epoch": 0.05443446358726525,
      "grad_norm": 9.785136222839355,
      "learning_rate": 8.504795117698345e-06,
      "loss": 0.6109,
      "mean_token_accuracy": 0.8185616865754127,
      "num_tokens": 21070030.0,
      "step": 17560
    },
    {
      "entropy": 1.8533698439598083,
      "epoch": 0.05446546271231494,
      "grad_norm": 4.606391906738281,
      "learning_rate": 8.509638670928995e-06,
      "loss": 0.6352,
      "mean_token_accuracy": 0.8094547167420387,
      "num_tokens": 21082808.0,
      "step": 17570
    },
    {
      "entropy": 1.8658347025513649,
      "epoch": 0.05449646183736464,
      "grad_norm": 10.683303833007812,
      "learning_rate": 8.514482224159645e-06,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8231870085000992,
      "num_tokens": 21095739.0,
      "step": 17580
    },
    {
      "entropy": 1.9149486675858498,
      "epoch": 0.054527460962414336,
      "grad_norm": 7.256167411804199,
      "learning_rate": 8.519325777390293e-06,
      "loss": 0.6271,
      "mean_token_accuracy": 0.8080606922507286,
      "num_tokens": 21107649.0,
      "step": 17590
    },
    {
      "entropy": 1.857093758881092,
      "epoch": 0.05455846008746403,
      "grad_norm": 10.606932640075684,
      "learning_rate": 8.524169330620945e-06,
      "loss": 0.575,
      "mean_token_accuracy": 0.8211929991841316,
      "num_tokens": 21120550.0,
      "step": 17600
    },
    {
      "entropy": 1.8073217749595643,
      "epoch": 0.05458945921251373,
      "grad_norm": 10.49376392364502,
      "learning_rate": 8.529012883851594e-06,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8263013437390327,
      "num_tokens": 21134529.0,
      "step": 17610
    },
    {
      "entropy": 1.8306412398815155,
      "epoch": 0.054620458337563425,
      "grad_norm": 9.479334831237793,
      "learning_rate": 8.533856437082244e-06,
      "loss": 0.6191,
      "mean_token_accuracy": 0.8138529911637307,
      "num_tokens": 21148789.0,
      "step": 17620
    },
    {
      "entropy": 1.9288539364933968,
      "epoch": 0.054651457462613115,
      "grad_norm": 9.992156028747559,
      "learning_rate": 8.538699990312894e-06,
      "loss": 0.6597,
      "mean_token_accuracy": 0.8028325289487839,
      "num_tokens": 21160363.0,
      "step": 17630
    },
    {
      "entropy": 1.865708639472723,
      "epoch": 0.05468245658766281,
      "grad_norm": 5.635770797729492,
      "learning_rate": 8.543543543543544e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8194977372884751,
      "num_tokens": 21173133.0,
      "step": 17640
    },
    {
      "entropy": 1.8686399355530738,
      "epoch": 0.05471345571271251,
      "grad_norm": 14.837812423706055,
      "learning_rate": 8.548387096774194e-06,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8228330850601197,
      "num_tokens": 21185944.0,
      "step": 17650
    },
    {
      "entropy": 1.936301600933075,
      "epoch": 0.054744454837762205,
      "grad_norm": 9.631260871887207,
      "learning_rate": 8.553230650004844e-06,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8145893216133118,
      "num_tokens": 21197684.0,
      "step": 17660
    },
    {
      "entropy": 1.9080315500497818,
      "epoch": 0.0547754539628119,
      "grad_norm": 9.75123119354248,
      "learning_rate": 8.558074203235495e-06,
      "loss": 0.5961,
      "mean_token_accuracy": 0.8067243576049805,
      "num_tokens": 21209448.0,
      "step": 17670
    },
    {
      "entropy": 1.941755273938179,
      "epoch": 0.0548064530878616,
      "grad_norm": 7.447671413421631,
      "learning_rate": 8.562917756466145e-06,
      "loss": 0.6306,
      "mean_token_accuracy": 0.8096268206834794,
      "num_tokens": 21222011.0,
      "step": 17680
    },
    {
      "entropy": 1.9120370209217072,
      "epoch": 0.05483745221291129,
      "grad_norm": 5.153918266296387,
      "learning_rate": 8.567761309696794e-06,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8094914838671684,
      "num_tokens": 21234368.0,
      "step": 17690
    },
    {
      "entropy": 1.8817522883415223,
      "epoch": 0.054868451337960984,
      "grad_norm": 10.505568504333496,
      "learning_rate": 8.572604862927443e-06,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8192173585295677,
      "num_tokens": 21246765.0,
      "step": 17700
    },
    {
      "entropy": 1.8698602363467216,
      "epoch": 0.05489945046301068,
      "grad_norm": 10.905664443969727,
      "learning_rate": 8.577448416158093e-06,
      "loss": 0.5834,
      "mean_token_accuracy": 0.81576436907053,
      "num_tokens": 21258879.0,
      "step": 17710
    },
    {
      "entropy": 1.8876920908689498,
      "epoch": 0.05493044958806038,
      "grad_norm": 10.402029991149902,
      "learning_rate": 8.582291969388745e-06,
      "loss": 0.5738,
      "mean_token_accuracy": 0.8234436243772507,
      "num_tokens": 21270505.0,
      "step": 17720
    },
    {
      "entropy": 1.8734422281384469,
      "epoch": 0.054961448713110074,
      "grad_norm": 10.90580940246582,
      "learning_rate": 8.587135522619395e-06,
      "loss": 0.5968,
      "mean_token_accuracy": 0.8190149515867233,
      "num_tokens": 21282053.0,
      "step": 17730
    },
    {
      "entropy": 1.8915531307458877,
      "epoch": 0.05499244783815977,
      "grad_norm": 8.619651794433594,
      "learning_rate": 8.591979075850045e-06,
      "loss": 0.6141,
      "mean_token_accuracy": 0.8292980402708053,
      "num_tokens": 21293252.0,
      "step": 17740
    },
    {
      "entropy": 1.819742462038994,
      "epoch": 0.05502344696320946,
      "grad_norm": 9.859935760498047,
      "learning_rate": 8.596822629080695e-06,
      "loss": 0.5696,
      "mean_token_accuracy": 0.825517275929451,
      "num_tokens": 21305648.0,
      "step": 17750
    },
    {
      "entropy": 1.7411428809165954,
      "epoch": 0.05505444608825916,
      "grad_norm": 11.464165687561035,
      "learning_rate": 8.601666182311344e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8340383976697922,
      "num_tokens": 21320159.0,
      "step": 17760
    },
    {
      "entropy": 1.919995127618313,
      "epoch": 0.05508544521330885,
      "grad_norm": 4.660171985626221,
      "learning_rate": 8.606509735541994e-06,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8194687977433205,
      "num_tokens": 21331854.0,
      "step": 17770
    },
    {
      "entropy": 1.826239649951458,
      "epoch": 0.05511644433835855,
      "grad_norm": 6.208508014678955,
      "learning_rate": 8.611353288772644e-06,
      "loss": 0.519,
      "mean_token_accuracy": 0.8199943840503693,
      "num_tokens": 21344553.0,
      "step": 17780
    },
    {
      "entropy": 1.964565047621727,
      "epoch": 0.05514744346340825,
      "grad_norm": 11.809846878051758,
      "learning_rate": 8.616196842003294e-06,
      "loss": 0.6973,
      "mean_token_accuracy": 0.8042270794510842,
      "num_tokens": 21355485.0,
      "step": 17790
    },
    {
      "entropy": 1.9378721952438354,
      "epoch": 0.05517844258845794,
      "grad_norm": 9.962228775024414,
      "learning_rate": 8.621040395233944e-06,
      "loss": 0.6338,
      "mean_token_accuracy": 0.8136247247457504,
      "num_tokens": 21367049.0,
      "step": 17800
    },
    {
      "entropy": 1.894200675189495,
      "epoch": 0.05520944171350763,
      "grad_norm": 4.6598663330078125,
      "learning_rate": 8.625883948464594e-06,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8180066585540772,
      "num_tokens": 21378413.0,
      "step": 17810
    },
    {
      "entropy": 1.828821636736393,
      "epoch": 0.05524044083855733,
      "grad_norm": 5.244216442108154,
      "learning_rate": 8.630727501695244e-06,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8234935000538826,
      "num_tokens": 21391978.0,
      "step": 17820
    },
    {
      "entropy": 1.8749590143561363,
      "epoch": 0.055271439963607026,
      "grad_norm": 11.188740730285645,
      "learning_rate": 8.635571054925894e-06,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8215572342276574,
      "num_tokens": 21403442.0,
      "step": 17830
    },
    {
      "entropy": 1.792897927761078,
      "epoch": 0.05530243908865672,
      "grad_norm": 5.151137828826904,
      "learning_rate": 8.640414608156543e-06,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8246389016509056,
      "num_tokens": 21416621.0,
      "step": 17840
    },
    {
      "entropy": 1.9171807587146759,
      "epoch": 0.05533343821370642,
      "grad_norm": 10.319511413574219,
      "learning_rate": 8.645258161387195e-06,
      "loss": 0.7074,
      "mean_token_accuracy": 0.8059808582067489,
      "num_tokens": 21428271.0,
      "step": 17850
    },
    {
      "entropy": 1.8502112567424773,
      "epoch": 0.055364437338756116,
      "grad_norm": 9.720250129699707,
      "learning_rate": 8.650101714617845e-06,
      "loss": 0.636,
      "mean_token_accuracy": 0.8085378974676132,
      "num_tokens": 21439709.0,
      "step": 17860
    },
    {
      "entropy": 1.8581691205501556,
      "epoch": 0.05539543646380581,
      "grad_norm": 10.996453285217285,
      "learning_rate": 8.654945267848495e-06,
      "loss": 0.6044,
      "mean_token_accuracy": 0.8172203198075294,
      "num_tokens": 21451408.0,
      "step": 17870
    },
    {
      "entropy": 1.8972811639308929,
      "epoch": 0.0554264355888555,
      "grad_norm": 10.46535587310791,
      "learning_rate": 8.659788821079145e-06,
      "loss": 0.6667,
      "mean_token_accuracy": 0.8105733722448349,
      "num_tokens": 21463410.0,
      "step": 17880
    },
    {
      "entropy": 1.801578015089035,
      "epoch": 0.0554574347139052,
      "grad_norm": 11.458210945129395,
      "learning_rate": 8.664632374309795e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8232032969594002,
      "num_tokens": 21476435.0,
      "step": 17890
    },
    {
      "entropy": 1.8537754505872726,
      "epoch": 0.055488433838954895,
      "grad_norm": 9.625106811523438,
      "learning_rate": 8.669475927540444e-06,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8271110355854034,
      "num_tokens": 21488620.0,
      "step": 17900
    },
    {
      "entropy": 1.920037829875946,
      "epoch": 0.05551943296400459,
      "grad_norm": 11.167927742004395,
      "learning_rate": 8.674319480771094e-06,
      "loss": 0.633,
      "mean_token_accuracy": 0.8028773024678231,
      "num_tokens": 21500165.0,
      "step": 17910
    },
    {
      "entropy": 1.8313481405377388,
      "epoch": 0.05555043208905429,
      "grad_norm": 12.061219215393066,
      "learning_rate": 8.679163034001744e-06,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8166040554642677,
      "num_tokens": 21513677.0,
      "step": 17920
    },
    {
      "entropy": 2.0305935621261595,
      "epoch": 0.055581431214103985,
      "grad_norm": 10.752704620361328,
      "learning_rate": 8.684006587232394e-06,
      "loss": 0.6779,
      "mean_token_accuracy": 0.8040038585662842,
      "num_tokens": 21524307.0,
      "step": 17930
    },
    {
      "entropy": 1.8893893539905549,
      "epoch": 0.055612430339153675,
      "grad_norm": 11.609795570373535,
      "learning_rate": 8.688850140463044e-06,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8086724013090134,
      "num_tokens": 21536735.0,
      "step": 17940
    },
    {
      "entropy": 1.8628504529595376,
      "epoch": 0.05564342946420337,
      "grad_norm": 7.905481815338135,
      "learning_rate": 8.693693693693694e-06,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8129374206066131,
      "num_tokens": 21549205.0,
      "step": 17950
    },
    {
      "entropy": 1.884210042655468,
      "epoch": 0.05567442858925307,
      "grad_norm": 12.496581077575684,
      "learning_rate": 8.698537246924344e-06,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8244456693530082,
      "num_tokens": 21560770.0,
      "step": 17960
    },
    {
      "entropy": 1.8549162834882735,
      "epoch": 0.055705427714302765,
      "grad_norm": 10.498296737670898,
      "learning_rate": 8.703380800154995e-06,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8179799810051918,
      "num_tokens": 21572851.0,
      "step": 17970
    },
    {
      "entropy": 1.9428973540663719,
      "epoch": 0.05573642683935246,
      "grad_norm": 9.824682235717773,
      "learning_rate": 8.708224353385645e-06,
      "loss": 0.6371,
      "mean_token_accuracy": 0.8067533582448959,
      "num_tokens": 21584287.0,
      "step": 17980
    },
    {
      "entropy": 1.89179328083992,
      "epoch": 0.05576742596440216,
      "grad_norm": 5.299261093139648,
      "learning_rate": 8.713067906616295e-06,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8168211251497268,
      "num_tokens": 21596202.0,
      "step": 17990
    },
    {
      "entropy": 1.8720185294747353,
      "epoch": 0.05579842508945185,
      "grad_norm": 9.72613525390625,
      "learning_rate": 8.717911459846945e-06,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8193200185894967,
      "num_tokens": 21608507.0,
      "step": 18000
    },
    {
      "entropy": 1.90470210313797,
      "epoch": 0.055829424214501544,
      "grad_norm": 11.743047714233398,
      "learning_rate": 8.722755013077593e-06,
      "loss": 0.6317,
      "mean_token_accuracy": 0.8138886854052544,
      "num_tokens": 21619927.0,
      "step": 18010
    },
    {
      "entropy": 1.9255423635244369,
      "epoch": 0.05586042333955124,
      "grad_norm": 10.3324613571167,
      "learning_rate": 8.727598566308245e-06,
      "loss": 0.6288,
      "mean_token_accuracy": 0.8161659583449363,
      "num_tokens": 21631633.0,
      "step": 18020
    },
    {
      "entropy": 1.7495290905237197,
      "epoch": 0.05589142246460094,
      "grad_norm": 11.315978050231934,
      "learning_rate": 8.732442119538895e-06,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8315317690372467,
      "num_tokens": 21645071.0,
      "step": 18030
    },
    {
      "entropy": 1.8392060235142709,
      "epoch": 0.055922421589650634,
      "grad_norm": 10.65888786315918,
      "learning_rate": 8.737285672769545e-06,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8336151942610741,
      "num_tokens": 21657281.0,
      "step": 18040
    },
    {
      "entropy": 1.9446683064103127,
      "epoch": 0.05595342071470033,
      "grad_norm": 13.20337963104248,
      "learning_rate": 8.742129226000194e-06,
      "loss": 0.682,
      "mean_token_accuracy": 0.7979879945516586,
      "num_tokens": 21668367.0,
      "step": 18050
    },
    {
      "entropy": 1.8128132298588753,
      "epoch": 0.05598441983975002,
      "grad_norm": 11.751765251159668,
      "learning_rate": 8.746972779230844e-06,
      "loss": 0.585,
      "mean_token_accuracy": 0.8146183237433433,
      "num_tokens": 21681270.0,
      "step": 18060
    },
    {
      "entropy": 1.8214144945144652,
      "epoch": 0.05601541896479972,
      "grad_norm": 10.238035202026367,
      "learning_rate": 8.751816332461494e-06,
      "loss": 0.585,
      "mean_token_accuracy": 0.8190832883119583,
      "num_tokens": 21694477.0,
      "step": 18070
    },
    {
      "entropy": 1.8992233827710152,
      "epoch": 0.05604641808984941,
      "grad_norm": 3.5823521614074707,
      "learning_rate": 8.756659885692144e-06,
      "loss": 0.6326,
      "mean_token_accuracy": 0.8172223970293999,
      "num_tokens": 21706135.0,
      "step": 18080
    },
    {
      "entropy": 1.8768006905913353,
      "epoch": 0.05607741721489911,
      "grad_norm": 12.048982620239258,
      "learning_rate": 8.761503438922796e-06,
      "loss": 0.6044,
      "mean_token_accuracy": 0.8237545311450958,
      "num_tokens": 21717528.0,
      "step": 18090
    },
    {
      "entropy": 1.8304587841033935,
      "epoch": 0.056108416339948806,
      "grad_norm": 10.88693904876709,
      "learning_rate": 8.766346992153446e-06,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8200173109769822,
      "num_tokens": 21730335.0,
      "step": 18100
    },
    {
      "entropy": 1.7798964753746986,
      "epoch": 0.0561394154649985,
      "grad_norm": 13.059136390686035,
      "learning_rate": 8.771190545384094e-06,
      "loss": 0.508,
      "mean_token_accuracy": 0.8268529579043389,
      "num_tokens": 21743306.0,
      "step": 18110
    },
    {
      "entropy": 1.9081214651465417,
      "epoch": 0.05617041459004819,
      "grad_norm": 11.699897766113281,
      "learning_rate": 8.776034098614744e-06,
      "loss": 0.6163,
      "mean_token_accuracy": 0.8113586112856865,
      "num_tokens": 21754558.0,
      "step": 18120
    },
    {
      "entropy": 1.8716801404953003,
      "epoch": 0.05620141371509789,
      "grad_norm": 11.486919403076172,
      "learning_rate": 8.780877651845394e-06,
      "loss": 0.6531,
      "mean_token_accuracy": 0.7999576240777969,
      "num_tokens": 21766487.0,
      "step": 18130
    },
    {
      "entropy": 1.8077737405896186,
      "epoch": 0.056232412840147586,
      "grad_norm": 10.865653991699219,
      "learning_rate": 8.785721205076045e-06,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8171229988336564,
      "num_tokens": 21779437.0,
      "step": 18140
    },
    {
      "entropy": 1.93416608273983,
      "epoch": 0.05626341196519728,
      "grad_norm": 9.039263725280762,
      "learning_rate": 8.790564758306695e-06,
      "loss": 0.6632,
      "mean_token_accuracy": 0.8090454161167144,
      "num_tokens": 21790560.0,
      "step": 18150
    },
    {
      "entropy": 1.9330880522727967,
      "epoch": 0.05629441109024698,
      "grad_norm": 9.971395492553711,
      "learning_rate": 8.795408311537345e-06,
      "loss": 0.5888,
      "mean_token_accuracy": 0.8192548274993896,
      "num_tokens": 21802186.0,
      "step": 18160
    },
    {
      "entropy": 1.8966943442821502,
      "epoch": 0.056325410215296676,
      "grad_norm": 5.171525955200195,
      "learning_rate": 8.800251864767995e-06,
      "loss": 0.609,
      "mean_token_accuracy": 0.8174291670322418,
      "num_tokens": 21813141.0,
      "step": 18170
    },
    {
      "entropy": 1.9147083714604378,
      "epoch": 0.056356409340346365,
      "grad_norm": 10.910831451416016,
      "learning_rate": 8.805095417998645e-06,
      "loss": 0.6791,
      "mean_token_accuracy": 0.802092096209526,
      "num_tokens": 21824549.0,
      "step": 18180
    },
    {
      "entropy": 1.845213919878006,
      "epoch": 0.05638740846539606,
      "grad_norm": 11.23419189453125,
      "learning_rate": 8.809938971229295e-06,
      "loss": 0.541,
      "mean_token_accuracy": 0.836587692797184,
      "num_tokens": 21837064.0,
      "step": 18190
    },
    {
      "entropy": 1.9281988382339477,
      "epoch": 0.05641840759044576,
      "grad_norm": 12.443825721740723,
      "learning_rate": 8.814782524459944e-06,
      "loss": 0.6652,
      "mean_token_accuracy": 0.8060685142874717,
      "num_tokens": 21849025.0,
      "step": 18200
    },
    {
      "entropy": 1.985886165499687,
      "epoch": 0.056449406715495455,
      "grad_norm": 11.284612655639648,
      "learning_rate": 8.819626077690594e-06,
      "loss": 0.6496,
      "mean_token_accuracy": 0.8076219454407692,
      "num_tokens": 21859566.0,
      "step": 18210
    },
    {
      "entropy": 1.9329922825098038,
      "epoch": 0.05648040584054515,
      "grad_norm": 8.869474411010742,
      "learning_rate": 8.824469630921244e-06,
      "loss": 0.6058,
      "mean_token_accuracy": 0.810475055873394,
      "num_tokens": 21871056.0,
      "step": 18220
    },
    {
      "entropy": 1.920028594136238,
      "epoch": 0.05651140496559485,
      "grad_norm": 11.169597625732422,
      "learning_rate": 8.829313184151894e-06,
      "loss": 0.636,
      "mean_token_accuracy": 0.8107326775789261,
      "num_tokens": 21881630.0,
      "step": 18230
    },
    {
      "entropy": 1.8644825682044028,
      "epoch": 0.056542404090644545,
      "grad_norm": 10.988966941833496,
      "learning_rate": 8.834156737382544e-06,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8243206977844239,
      "num_tokens": 21893921.0,
      "step": 18240
    },
    {
      "entropy": 1.8154290586709976,
      "epoch": 0.056573403215694235,
      "grad_norm": 6.226930141448975,
      "learning_rate": 8.839000290613194e-06,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8170310199260712,
      "num_tokens": 21906253.0,
      "step": 18250
    },
    {
      "entropy": 1.859774151444435,
      "epoch": 0.05660440234074393,
      "grad_norm": 11.1119384765625,
      "learning_rate": 8.843843843843844e-06,
      "loss": 0.6233,
      "mean_token_accuracy": 0.8082350715994835,
      "num_tokens": 21918932.0,
      "step": 18260
    },
    {
      "entropy": 1.9023708343505858,
      "epoch": 0.05663540146579363,
      "grad_norm": 10.345152854919434,
      "learning_rate": 8.848687397074495e-06,
      "loss": 0.5928,
      "mean_token_accuracy": 0.8202927365899086,
      "num_tokens": 21930745.0,
      "step": 18270
    },
    {
      "entropy": 1.8009515717625617,
      "epoch": 0.056666400590843324,
      "grad_norm": 4.616584777832031,
      "learning_rate": 8.853530950305145e-06,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8291145578026772,
      "num_tokens": 21943477.0,
      "step": 18280
    },
    {
      "entropy": 1.910208411514759,
      "epoch": 0.05669739971589302,
      "grad_norm": 10.986778259277344,
      "learning_rate": 8.858374503535795e-06,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8097553431987763,
      "num_tokens": 21955368.0,
      "step": 18290
    },
    {
      "entropy": 1.8674479261040688,
      "epoch": 0.05672839884094272,
      "grad_norm": 12.478263854980469,
      "learning_rate": 8.863218056766445e-06,
      "loss": 0.592,
      "mean_token_accuracy": 0.8068441540002823,
      "num_tokens": 21968150.0,
      "step": 18300
    },
    {
      "entropy": 1.9316591426730156,
      "epoch": 0.05675939796599241,
      "grad_norm": 13.325005531311035,
      "learning_rate": 8.868061609997095e-06,
      "loss": 0.5848,
      "mean_token_accuracy": 0.8228493258357048,
      "num_tokens": 21979588.0,
      "step": 18310
    },
    {
      "entropy": 1.8889947533607483,
      "epoch": 0.056790397091042104,
      "grad_norm": 10.683235168457031,
      "learning_rate": 8.872905163227745e-06,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8133319228887558,
      "num_tokens": 21991387.0,
      "step": 18320
    },
    {
      "entropy": 1.83803388774395,
      "epoch": 0.0568213962160918,
      "grad_norm": 9.95289421081543,
      "learning_rate": 8.877748716458395e-06,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8165292799472809,
      "num_tokens": 22004539.0,
      "step": 18330
    },
    {
      "entropy": 1.8539226263761521,
      "epoch": 0.0568523953411415,
      "grad_norm": 11.232304573059082,
      "learning_rate": 8.882592269689044e-06,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8155701532959938,
      "num_tokens": 22017151.0,
      "step": 18340
    },
    {
      "entropy": 1.8634928598999978,
      "epoch": 0.056883394466191194,
      "grad_norm": 10.557002067565918,
      "learning_rate": 8.887435822919694e-06,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8263462707400322,
      "num_tokens": 22029875.0,
      "step": 18350
    },
    {
      "entropy": 1.8347053781151772,
      "epoch": 0.05691439359124089,
      "grad_norm": 5.2376389503479,
      "learning_rate": 8.892279376150344e-06,
      "loss": 0.553,
      "mean_token_accuracy": 0.82252157330513,
      "num_tokens": 22042520.0,
      "step": 18360
    },
    {
      "entropy": 1.9364128440618515,
      "epoch": 0.05694539271629058,
      "grad_norm": 5.532094955444336,
      "learning_rate": 8.897122929380994e-06,
      "loss": 0.673,
      "mean_token_accuracy": 0.800466650724411,
      "num_tokens": 22054370.0,
      "step": 18370
    },
    {
      "entropy": 1.9340432003140449,
      "epoch": 0.056976391841340276,
      "grad_norm": 11.368268966674805,
      "learning_rate": 8.901966482611644e-06,
      "loss": 0.6135,
      "mean_token_accuracy": 0.8022024169564247,
      "num_tokens": 22066153.0,
      "step": 18380
    },
    {
      "entropy": 1.8881090357899666,
      "epoch": 0.05700739096638997,
      "grad_norm": 5.1713972091674805,
      "learning_rate": 8.906810035842296e-06,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8244768619537354,
      "num_tokens": 22078654.0,
      "step": 18390
    },
    {
      "entropy": 1.919769637286663,
      "epoch": 0.05703839009143967,
      "grad_norm": 8.57338809967041,
      "learning_rate": 8.911653589072945e-06,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8175665631890296,
      "num_tokens": 22091226.0,
      "step": 18400
    },
    {
      "entropy": 1.7635185047984123,
      "epoch": 0.057069389216489366,
      "grad_norm": 3.2729835510253906,
      "learning_rate": 8.916497142303595e-06,
      "loss": 0.514,
      "mean_token_accuracy": 0.8288957670331001,
      "num_tokens": 22105037.0,
      "step": 18410
    },
    {
      "entropy": 1.84332295358181,
      "epoch": 0.05710038834153906,
      "grad_norm": 11.345158576965332,
      "learning_rate": 8.921340695534245e-06,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8165907859802246,
      "num_tokens": 22117072.0,
      "step": 18420
    },
    {
      "entropy": 1.8966827645897866,
      "epoch": 0.05713138746658875,
      "grad_norm": 12.612526893615723,
      "learning_rate": 8.926184248764893e-06,
      "loss": 0.6189,
      "mean_token_accuracy": 0.8158532097935677,
      "num_tokens": 22127952.0,
      "step": 18430
    },
    {
      "entropy": 1.8704052582383155,
      "epoch": 0.05716238659163845,
      "grad_norm": 5.032566070556641,
      "learning_rate": 8.931027801995545e-06,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8137423068284988,
      "num_tokens": 22139709.0,
      "step": 18440
    },
    {
      "entropy": 1.820935921370983,
      "epoch": 0.057193385716688146,
      "grad_norm": 10.067622184753418,
      "learning_rate": 8.935871355226195e-06,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8368907868862152,
      "num_tokens": 22152377.0,
      "step": 18450
    },
    {
      "entropy": 1.896417185664177,
      "epoch": 0.05722438484173784,
      "grad_norm": 12.381418228149414,
      "learning_rate": 8.940714908456845e-06,
      "loss": 0.6965,
      "mean_token_accuracy": 0.8038837254047394,
      "num_tokens": 22163577.0,
      "step": 18460
    },
    {
      "entropy": 1.8900171846151352,
      "epoch": 0.05725538396678754,
      "grad_norm": 11.906807899475098,
      "learning_rate": 8.945558461687495e-06,
      "loss": 0.6449,
      "mean_token_accuracy": 0.8201625868678093,
      "num_tokens": 22174708.0,
      "step": 18470
    },
    {
      "entropy": 1.9111595183610917,
      "epoch": 0.057286383091837235,
      "grad_norm": 11.19156265258789,
      "learning_rate": 8.950402014918145e-06,
      "loss": 0.6555,
      "mean_token_accuracy": 0.802647915482521,
      "num_tokens": 22186789.0,
      "step": 18480
    },
    {
      "entropy": 1.9203668639063836,
      "epoch": 0.057317382216886925,
      "grad_norm": 11.311783790588379,
      "learning_rate": 8.955245568148794e-06,
      "loss": 0.6169,
      "mean_token_accuracy": 0.8133427366614342,
      "num_tokens": 22198408.0,
      "step": 18490
    },
    {
      "entropy": 1.9032350957393647,
      "epoch": 0.05734838134193662,
      "grad_norm": 13.073493957519531,
      "learning_rate": 8.960089121379444e-06,
      "loss": 0.6344,
      "mean_token_accuracy": 0.80216343998909,
      "num_tokens": 22210305.0,
      "step": 18500
    },
    {
      "entropy": 1.8219037756323815,
      "epoch": 0.05737938046698632,
      "grad_norm": 9.571296691894531,
      "learning_rate": 8.964932674610096e-06,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8286261260509491,
      "num_tokens": 22222972.0,
      "step": 18510
    },
    {
      "entropy": 1.8522532120347024,
      "epoch": 0.057410379592036015,
      "grad_norm": 5.730122089385986,
      "learning_rate": 8.969776227840746e-06,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8103675618767738,
      "num_tokens": 22235073.0,
      "step": 18520
    },
    {
      "entropy": 1.851724323630333,
      "epoch": 0.05744137871708571,
      "grad_norm": 9.87784194946289,
      "learning_rate": 8.974619781071394e-06,
      "loss": 0.5931,
      "mean_token_accuracy": 0.8163824722170829,
      "num_tokens": 22247790.0,
      "step": 18530
    },
    {
      "entropy": 1.8631928369402886,
      "epoch": 0.05747237784213541,
      "grad_norm": 5.00623893737793,
      "learning_rate": 8.979463334302044e-06,
      "loss": 0.559,
      "mean_token_accuracy": 0.8131022065877914,
      "num_tokens": 22260240.0,
      "step": 18540
    },
    {
      "entropy": 1.910364383459091,
      "epoch": 0.0575033769671851,
      "grad_norm": 11.106468200683594,
      "learning_rate": 8.984306887532694e-06,
      "loss": 0.6496,
      "mean_token_accuracy": 0.8063996240496636,
      "num_tokens": 22272059.0,
      "step": 18550
    },
    {
      "entropy": 1.9144617825746537,
      "epoch": 0.057534376092234794,
      "grad_norm": 10.869990348815918,
      "learning_rate": 8.989150440763345e-06,
      "loss": 0.6416,
      "mean_token_accuracy": 0.8138428092002868,
      "num_tokens": 22283072.0,
      "step": 18560
    },
    {
      "entropy": 1.9242763727903367,
      "epoch": 0.05756537521728449,
      "grad_norm": 10.430070877075195,
      "learning_rate": 8.993993993993995e-06,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8113959297537804,
      "num_tokens": 22294697.0,
      "step": 18570
    },
    {
      "entropy": 1.9588421791791917,
      "epoch": 0.05759637434233419,
      "grad_norm": 12.86335277557373,
      "learning_rate": 8.998837547224645e-06,
      "loss": 0.6676,
      "mean_token_accuracy": 0.80771906375885,
      "num_tokens": 22305842.0,
      "step": 18580
    },
    {
      "entropy": 1.8290845677256584,
      "epoch": 0.057627373467383884,
      "grad_norm": 11.009787559509277,
      "learning_rate": 9.003681100455295e-06,
      "loss": 0.5679,
      "mean_token_accuracy": 0.819172665476799,
      "num_tokens": 22319673.0,
      "step": 18590
    },
    {
      "entropy": 1.8799395859241486,
      "epoch": 0.05765837259243358,
      "grad_norm": 5.414979457855225,
      "learning_rate": 9.008524653685945e-06,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8109808668494225,
      "num_tokens": 22332734.0,
      "step": 18600
    },
    {
      "entropy": 1.9298516079783439,
      "epoch": 0.05768937171748328,
      "grad_norm": 4.051270961761475,
      "learning_rate": 9.013368206916595e-06,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8061932161450386,
      "num_tokens": 22344081.0,
      "step": 18610
    },
    {
      "entropy": 1.9745507806539535,
      "epoch": 0.05772037084253297,
      "grad_norm": 10.910243034362793,
      "learning_rate": 9.018211760147245e-06,
      "loss": 0.7182,
      "mean_token_accuracy": 0.7929084002971649,
      "num_tokens": 22355535.0,
      "step": 18620
    },
    {
      "entropy": 1.880644341558218,
      "epoch": 0.057751369967582664,
      "grad_norm": 6.737995624542236,
      "learning_rate": 9.023055313377894e-06,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8097090050578117,
      "num_tokens": 22368050.0,
      "step": 18630
    },
    {
      "entropy": 1.9336914867162704,
      "epoch": 0.05778236909263236,
      "grad_norm": 9.826704978942871,
      "learning_rate": 9.027898866608544e-06,
      "loss": 0.6545,
      "mean_token_accuracy": 0.8081561967730522,
      "num_tokens": 22379637.0,
      "step": 18640
    },
    {
      "entropy": 1.9320360347628593,
      "epoch": 0.05781336821768206,
      "grad_norm": 5.833460330963135,
      "learning_rate": 9.032742419839194e-06,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8115993320941925,
      "num_tokens": 22391313.0,
      "step": 18650
    },
    {
      "entropy": 1.952963298559189,
      "epoch": 0.05784436734273175,
      "grad_norm": 5.079139709472656,
      "learning_rate": 9.037585973069844e-06,
      "loss": 0.6632,
      "mean_token_accuracy": 0.8034544378519058,
      "num_tokens": 22402538.0,
      "step": 18660
    },
    {
      "entropy": 1.7565592139959336,
      "epoch": 0.05787536646778145,
      "grad_norm": 4.124721050262451,
      "learning_rate": 9.042429526300494e-06,
      "loss": 0.5241,
      "mean_token_accuracy": 0.8233913615345955,
      "num_tokens": 22416764.0,
      "step": 18670
    },
    {
      "entropy": 1.900513444840908,
      "epoch": 0.05790636559283114,
      "grad_norm": 9.634764671325684,
      "learning_rate": 9.047273079531144e-06,
      "loss": 0.573,
      "mean_token_accuracy": 0.8233014553785324,
      "num_tokens": 22428670.0,
      "step": 18680
    },
    {
      "entropy": 1.9495866417884826,
      "epoch": 0.057937364717880836,
      "grad_norm": Infinity,
      "learning_rate": 9.052116632761795e-06,
      "loss": 0.6543,
      "mean_token_accuracy": 0.8013565137982368,
      "num_tokens": 22439439.0,
      "step": 18690
    },
    {
      "entropy": 1.8713518410921097,
      "epoch": 0.05796836384293053,
      "grad_norm": 9.244678497314453,
      "learning_rate": 9.056960185992445e-06,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8261402577161789,
      "num_tokens": 22450954.0,
      "step": 18700
    },
    {
      "entropy": 1.867144940048456,
      "epoch": 0.05799936296798023,
      "grad_norm": 5.15731143951416,
      "learning_rate": 9.061803739223095e-06,
      "loss": 0.5672,
      "mean_token_accuracy": 0.814095051586628,
      "num_tokens": 22463821.0,
      "step": 18710
    },
    {
      "entropy": 1.8838742166757583,
      "epoch": 0.058030362093029926,
      "grad_norm": 10.645661354064941,
      "learning_rate": 9.066647292453745e-06,
      "loss": 0.5763,
      "mean_token_accuracy": 0.82436051517725,
      "num_tokens": 22475600.0,
      "step": 18720
    },
    {
      "entropy": 1.8638369679450988,
      "epoch": 0.05806136121807962,
      "grad_norm": 14.198400497436523,
      "learning_rate": 9.071490845684395e-06,
      "loss": 0.6381,
      "mean_token_accuracy": 0.8040808498859405,
      "num_tokens": 22487874.0,
      "step": 18730
    },
    {
      "entropy": 1.8371347174048425,
      "epoch": 0.05809236034312931,
      "grad_norm": 5.825725555419922,
      "learning_rate": 9.076334398915045e-06,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8231277003884315,
      "num_tokens": 22500865.0,
      "step": 18740
    },
    {
      "entropy": 1.9247653126716613,
      "epoch": 0.05812335946817901,
      "grad_norm": 12.100114822387695,
      "learning_rate": 9.081177952145695e-06,
      "loss": 0.6471,
      "mean_token_accuracy": 0.8075411379337311,
      "num_tokens": 22511426.0,
      "step": 18750
    },
    {
      "entropy": 1.8342102706432342,
      "epoch": 0.058154358593228705,
      "grad_norm": 9.929917335510254,
      "learning_rate": 9.086021505376345e-06,
      "loss": 0.6312,
      "mean_token_accuracy": 0.7977308183908463,
      "num_tokens": 22524288.0,
      "step": 18760
    },
    {
      "entropy": 1.7855632066726685,
      "epoch": 0.0581853577182784,
      "grad_norm": 5.584056854248047,
      "learning_rate": 9.090865058606995e-06,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8210761070251464,
      "num_tokens": 22537475.0,
      "step": 18770
    },
    {
      "entropy": 1.8851278007030488,
      "epoch": 0.0582163568433281,
      "grad_norm": 10.225055694580078,
      "learning_rate": 9.095708611837644e-06,
      "loss": 0.6466,
      "mean_token_accuracy": 0.8099891990423203,
      "num_tokens": 22548279.0,
      "step": 18780
    },
    {
      "entropy": 1.841400173306465,
      "epoch": 0.058247355968377795,
      "grad_norm": 9.732316017150879,
      "learning_rate": 9.100552165068294e-06,
      "loss": 0.5658,
      "mean_token_accuracy": 0.820391620695591,
      "num_tokens": 22560249.0,
      "step": 18790
    },
    {
      "entropy": 1.7526367500424385,
      "epoch": 0.058278355093427485,
      "grad_norm": 10.728232383728027,
      "learning_rate": 9.105395718298944e-06,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8198321342468262,
      "num_tokens": 22573834.0,
      "step": 18800
    },
    {
      "entropy": 1.8276120245456695,
      "epoch": 0.05830935421847718,
      "grad_norm": 11.885202407836914,
      "learning_rate": 9.110239271529596e-06,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8176509469747544,
      "num_tokens": 22586688.0,
      "step": 18810
    },
    {
      "entropy": 1.8741453379392623,
      "epoch": 0.05834035334352688,
      "grad_norm": 12.088459968566895,
      "learning_rate": 9.115082824760246e-06,
      "loss": 0.6949,
      "mean_token_accuracy": 0.7987207651138306,
      "num_tokens": 22598358.0,
      "step": 18820
    },
    {
      "entropy": 1.8715686663985251,
      "epoch": 0.058371352468576575,
      "grad_norm": 10.86611270904541,
      "learning_rate": 9.119926377990896e-06,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8120866730809212,
      "num_tokens": 22609875.0,
      "step": 18830
    },
    {
      "entropy": 1.844074723124504,
      "epoch": 0.05840235159362627,
      "grad_norm": 10.710127830505371,
      "learning_rate": 9.124769931221545e-06,
      "loss": 0.6134,
      "mean_token_accuracy": 0.8119207620620728,
      "num_tokens": 22621314.0,
      "step": 18840
    },
    {
      "entropy": 1.8183260425925254,
      "epoch": 0.05843335071867597,
      "grad_norm": 9.820539474487305,
      "learning_rate": 9.129613484452194e-06,
      "loss": 0.6007,
      "mean_token_accuracy": 0.809662489593029,
      "num_tokens": 22633297.0,
      "step": 18850
    },
    {
      "entropy": 1.8575424775481224,
      "epoch": 0.05846434984372566,
      "grad_norm": 9.236701011657715,
      "learning_rate": 9.134457037682845e-06,
      "loss": 0.6365,
      "mean_token_accuracy": 0.8071271240711212,
      "num_tokens": 22645167.0,
      "step": 18860
    },
    {
      "entropy": 1.8752565145492555,
      "epoch": 0.058495348968775354,
      "grad_norm": 5.298765659332275,
      "learning_rate": 9.139300590913495e-06,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8160139411687851,
      "num_tokens": 22656745.0,
      "step": 18870
    },
    {
      "entropy": 1.8490721896290778,
      "epoch": 0.05852634809382505,
      "grad_norm": 10.962682723999023,
      "learning_rate": 9.144144144144145e-06,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8067515045404434,
      "num_tokens": 22669776.0,
      "step": 18880
    },
    {
      "entropy": 1.882719586789608,
      "epoch": 0.05855734721887475,
      "grad_norm": 10.184673309326172,
      "learning_rate": 9.148987697374795e-06,
      "loss": 0.6346,
      "mean_token_accuracy": 0.8162581637501717,
      "num_tokens": 22681064.0,
      "step": 18890
    },
    {
      "entropy": 1.7815841138362885,
      "epoch": 0.058588346343924444,
      "grad_norm": 9.868931770324707,
      "learning_rate": 9.153831250605445e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8312078416347504,
      "num_tokens": 22694642.0,
      "step": 18900
    },
    {
      "entropy": 1.8093548193573952,
      "epoch": 0.05861934546897414,
      "grad_norm": 10.930723190307617,
      "learning_rate": 9.158674803836095e-06,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8200191363692284,
      "num_tokens": 22707548.0,
      "step": 18910
    },
    {
      "entropy": 1.8189376056194306,
      "epoch": 0.05865034459402383,
      "grad_norm": 9.895159721374512,
      "learning_rate": 9.163518357066745e-06,
      "loss": 0.6132,
      "mean_token_accuracy": 0.8140119895339012,
      "num_tokens": 22720201.0,
      "step": 18920
    },
    {
      "entropy": 1.8482043415307998,
      "epoch": 0.05868134371907353,
      "grad_norm": 10.34995174407959,
      "learning_rate": 9.168361910297396e-06,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8194187000393868,
      "num_tokens": 22732586.0,
      "step": 18930
    },
    {
      "entropy": 1.8547095090150834,
      "epoch": 0.05871234284412322,
      "grad_norm": 9.405129432678223,
      "learning_rate": 9.173205463528046e-06,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8073139742016793,
      "num_tokens": 22743872.0,
      "step": 18940
    },
    {
      "entropy": 1.7496973037719727,
      "epoch": 0.05874334196917292,
      "grad_norm": 11.039433479309082,
      "learning_rate": 9.178049016758694e-06,
      "loss": 0.537,
      "mean_token_accuracy": 0.8219654500484467,
      "num_tokens": 22756915.0,
      "step": 18950
    },
    {
      "entropy": 1.8445199981331826,
      "epoch": 0.05877434109422262,
      "grad_norm": 10.18836498260498,
      "learning_rate": 9.182892569989344e-06,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8294688493013382,
      "num_tokens": 22768399.0,
      "step": 18960
    },
    {
      "entropy": 1.8422922030091287,
      "epoch": 0.05880534021927231,
      "grad_norm": 10.27499008178711,
      "learning_rate": 9.187736123219994e-06,
      "loss": 0.6632,
      "mean_token_accuracy": 0.815483920276165,
      "num_tokens": 22780846.0,
      "step": 18970
    },
    {
      "entropy": 1.9067127719521522,
      "epoch": 0.05883633934432201,
      "grad_norm": 9.372925758361816,
      "learning_rate": 9.192579676450646e-06,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8227426365017891,
      "num_tokens": 22792760.0,
      "step": 18980
    },
    {
      "entropy": 1.9074788480997085,
      "epoch": 0.0588673384693717,
      "grad_norm": 5.241424560546875,
      "learning_rate": 9.197423229681295e-06,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8211500853300094,
      "num_tokens": 22804075.0,
      "step": 18990
    },
    {
      "entropy": 1.911788022518158,
      "epoch": 0.058898337594421396,
      "grad_norm": 12.606864929199219,
      "learning_rate": 9.202266782911945e-06,
      "loss": 0.6369,
      "mean_token_accuracy": 0.8175441965460777,
      "num_tokens": 22814931.0,
      "step": 19000
    },
    {
      "entropy": 1.902238203585148,
      "epoch": 0.05892933671947109,
      "grad_norm": 5.6409831047058105,
      "learning_rate": 9.207110336142595e-06,
      "loss": 0.6392,
      "mean_token_accuracy": 0.8043625161051751,
      "num_tokens": 22826140.0,
      "step": 19010
    },
    {
      "entropy": 1.8769164964556695,
      "epoch": 0.05896033584452079,
      "grad_norm": 5.731686592102051,
      "learning_rate": 9.211953889373245e-06,
      "loss": 0.6231,
      "mean_token_accuracy": 0.8021165639162063,
      "num_tokens": 22837728.0,
      "step": 19020
    },
    {
      "entropy": 1.7606040745973588,
      "epoch": 0.058991334969570486,
      "grad_norm": 12.572894096374512,
      "learning_rate": 9.216797442603895e-06,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8280571162700653,
      "num_tokens": 22851334.0,
      "step": 19030
    },
    {
      "entropy": 1.8437473088502885,
      "epoch": 0.05902233409462018,
      "grad_norm": 10.523193359375,
      "learning_rate": 9.221640995834545e-06,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8213176369667053,
      "num_tokens": 22864325.0,
      "step": 19040
    },
    {
      "entropy": 1.9376160144805907,
      "epoch": 0.05905333321966987,
      "grad_norm": 9.312682151794434,
      "learning_rate": 9.226484549065195e-06,
      "loss": 0.6229,
      "mean_token_accuracy": 0.8095207557082176,
      "num_tokens": 22876271.0,
      "step": 19050
    },
    {
      "entropy": 1.8676683530211449,
      "epoch": 0.05908433234471957,
      "grad_norm": 10.493414878845215,
      "learning_rate": 9.231328102295845e-06,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8218972474336624,
      "num_tokens": 22888674.0,
      "step": 19060
    },
    {
      "entropy": 1.9024165108799935,
      "epoch": 0.059115331469769265,
      "grad_norm": 11.642742156982422,
      "learning_rate": 9.236171655526494e-06,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8063879519701004,
      "num_tokens": 22901132.0,
      "step": 19070
    },
    {
      "entropy": 1.8537215188145637,
      "epoch": 0.05914633059481896,
      "grad_norm": 10.604762077331543,
      "learning_rate": 9.241015208757144e-06,
      "loss": 0.5936,
      "mean_token_accuracy": 0.8061328649520874,
      "num_tokens": 22913819.0,
      "step": 19080
    },
    {
      "entropy": 1.9175572365522384,
      "epoch": 0.05917732971986866,
      "grad_norm": 5.126205921173096,
      "learning_rate": 9.245858761987794e-06,
      "loss": 0.6629,
      "mean_token_accuracy": 0.8014189884066582,
      "num_tokens": 22924724.0,
      "step": 19090
    },
    {
      "entropy": 1.9114294454455376,
      "epoch": 0.059208328844918355,
      "grad_norm": 10.768912315368652,
      "learning_rate": 9.250702315218446e-06,
      "loss": 0.6447,
      "mean_token_accuracy": 0.8088298216462135,
      "num_tokens": 22936371.0,
      "step": 19100
    },
    {
      "entropy": 1.9175557851791383,
      "epoch": 0.059239327969968045,
      "grad_norm": 9.46970272064209,
      "learning_rate": 9.255545868449096e-06,
      "loss": 0.6537,
      "mean_token_accuracy": 0.810999846458435,
      "num_tokens": 22948269.0,
      "step": 19110
    },
    {
      "entropy": 1.80474793612957,
      "epoch": 0.05927032709501774,
      "grad_norm": 7.381761074066162,
      "learning_rate": 9.260389421679746e-06,
      "loss": 0.559,
      "mean_token_accuracy": 0.8227289646863938,
      "num_tokens": 22961310.0,
      "step": 19120
    },
    {
      "entropy": 1.868805430829525,
      "epoch": 0.05930132622006744,
      "grad_norm": 10.440200805664062,
      "learning_rate": 9.265232974910395e-06,
      "loss": 0.548,
      "mean_token_accuracy": 0.8252261653542519,
      "num_tokens": 22973534.0,
      "step": 19130
    },
    {
      "entropy": 1.9009343415498734,
      "epoch": 0.059332325345117135,
      "grad_norm": 10.150739669799805,
      "learning_rate": 9.270076528141045e-06,
      "loss": 0.6062,
      "mean_token_accuracy": 0.8130706340074539,
      "num_tokens": 22984921.0,
      "step": 19140
    },
    {
      "entropy": 1.9186480671167374,
      "epoch": 0.05936332447016683,
      "grad_norm": 9.625990867614746,
      "learning_rate": 9.274920081371695e-06,
      "loss": 0.6403,
      "mean_token_accuracy": 0.8057694777846336,
      "num_tokens": 22996985.0,
      "step": 19150
    },
    {
      "entropy": 1.8890679344534873,
      "epoch": 0.05939432359521653,
      "grad_norm": 10.16703987121582,
      "learning_rate": 9.279763634602345e-06,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8147070035338402,
      "num_tokens": 23008337.0,
      "step": 19160
    },
    {
      "entropy": 1.9546454161405564,
      "epoch": 0.05942532272026622,
      "grad_norm": 13.017064094543457,
      "learning_rate": 9.284607187832995e-06,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7958240196108818,
      "num_tokens": 23019347.0,
      "step": 19170
    },
    {
      "entropy": 1.8713775292038917,
      "epoch": 0.059456321845315914,
      "grad_norm": 12.783992767333984,
      "learning_rate": 9.289450741063645e-06,
      "loss": 0.5894,
      "mean_token_accuracy": 0.8084977343678474,
      "num_tokens": 23031934.0,
      "step": 19180
    },
    {
      "entropy": 1.9044471591711045,
      "epoch": 0.05948732097036561,
      "grad_norm": 11.211695671081543,
      "learning_rate": 9.294294294294295e-06,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8160697892308235,
      "num_tokens": 23043989.0,
      "step": 19190
    },
    {
      "entropy": 1.8023822262883187,
      "epoch": 0.05951832009541531,
      "grad_norm": 10.612577438354492,
      "learning_rate": 9.299137847524945e-06,
      "loss": 0.6033,
      "mean_token_accuracy": 0.8159924641251564,
      "num_tokens": 23058046.0,
      "step": 19200
    },
    {
      "entropy": 1.9452301010489463,
      "epoch": 0.059549319220465004,
      "grad_norm": 11.767992973327637,
      "learning_rate": 9.303981400755595e-06,
      "loss": 0.68,
      "mean_token_accuracy": 0.803168785572052,
      "num_tokens": 23069178.0,
      "step": 19210
    },
    {
      "entropy": 1.918524721264839,
      "epoch": 0.0595803183455147,
      "grad_norm": 10.314653396606445,
      "learning_rate": 9.308824953986244e-06,
      "loss": 0.6834,
      "mean_token_accuracy": 0.8066515281796456,
      "num_tokens": 23080400.0,
      "step": 19220
    },
    {
      "entropy": 1.9129308730363845,
      "epoch": 0.05961131747056439,
      "grad_norm": 12.398180961608887,
      "learning_rate": 9.313668507216896e-06,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8124301135540009,
      "num_tokens": 23091734.0,
      "step": 19230
    },
    {
      "entropy": 1.9237210959196092,
      "epoch": 0.05964231659561409,
      "grad_norm": 10.278820037841797,
      "learning_rate": 9.318512060447546e-06,
      "loss": 0.6262,
      "mean_token_accuracy": 0.8095857813954354,
      "num_tokens": 23103397.0,
      "step": 19240
    },
    {
      "entropy": 1.8663773894309998,
      "epoch": 0.05967331572066378,
      "grad_norm": 10.0914888381958,
      "learning_rate": 9.323355613678196e-06,
      "loss": 0.5701,
      "mean_token_accuracy": 0.821681647002697,
      "num_tokens": 23115814.0,
      "step": 19250
    },
    {
      "entropy": 1.7783602967858314,
      "epoch": 0.05970431484571348,
      "grad_norm": 3.7728965282440186,
      "learning_rate": 9.328199166908846e-06,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8306898340582848,
      "num_tokens": 23128899.0,
      "step": 19260
    },
    {
      "entropy": 1.894435779750347,
      "epoch": 0.059735313970763176,
      "grad_norm": 11.809426307678223,
      "learning_rate": 9.333042720139494e-06,
      "loss": 0.6599,
      "mean_token_accuracy": 0.8009930282831192,
      "num_tokens": 23140688.0,
      "step": 19270
    },
    {
      "entropy": 1.8965399265289307,
      "epoch": 0.05976631309581287,
      "grad_norm": 10.043713569641113,
      "learning_rate": 9.337886273370145e-06,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8076204761862755,
      "num_tokens": 23152768.0,
      "step": 19280
    },
    {
      "entropy": 1.9386232048273087,
      "epoch": 0.05979731222086257,
      "grad_norm": 9.522479057312012,
      "learning_rate": 9.342729826600795e-06,
      "loss": 0.6379,
      "mean_token_accuracy": 0.8103454813361168,
      "num_tokens": 23164363.0,
      "step": 19290
    },
    {
      "entropy": 1.9121969774365426,
      "epoch": 0.05982831134591226,
      "grad_norm": 5.025160789489746,
      "learning_rate": 9.347573379831445e-06,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8114672183990479,
      "num_tokens": 23176305.0,
      "step": 19300
    },
    {
      "entropy": 1.8917111858725548,
      "epoch": 0.059859310470961956,
      "grad_norm": 4.935334205627441,
      "learning_rate": 9.352416933062095e-06,
      "loss": 0.599,
      "mean_token_accuracy": 0.8276031494140625,
      "num_tokens": 23187205.0,
      "step": 19310
    },
    {
      "entropy": 1.96262284219265,
      "epoch": 0.05989030959601165,
      "grad_norm": 9.109956741333008,
      "learning_rate": 9.357260486292745e-06,
      "loss": 0.646,
      "mean_token_accuracy": 0.8043781608343125,
      "num_tokens": 23198670.0,
      "step": 19320
    },
    {
      "entropy": 1.9255134493112565,
      "epoch": 0.05992130872106135,
      "grad_norm": 9.983922958374023,
      "learning_rate": 9.362104039523395e-06,
      "loss": 0.6009,
      "mean_token_accuracy": 0.8162341311573982,
      "num_tokens": 23210792.0,
      "step": 19330
    },
    {
      "entropy": 1.9093994736671447,
      "epoch": 0.059952307846111046,
      "grad_norm": 9.770051002502441,
      "learning_rate": 9.366947592754045e-06,
      "loss": 0.663,
      "mean_token_accuracy": 0.8095744162797928,
      "num_tokens": 23222187.0,
      "step": 19340
    },
    {
      "entropy": 1.8983681246638298,
      "epoch": 0.05998330697116074,
      "grad_norm": 13.514256477355957,
      "learning_rate": 9.371791145984696e-06,
      "loss": 0.6243,
      "mean_token_accuracy": 0.7971692577004432,
      "num_tokens": 23233365.0,
      "step": 19350
    },
    {
      "entropy": 1.9183629781007767,
      "epoch": 0.06001430609621043,
      "grad_norm": 10.821701049804688,
      "learning_rate": 9.376634699215346e-06,
      "loss": 0.6502,
      "mean_token_accuracy": 0.8030115276575088,
      "num_tokens": 23243672.0,
      "step": 19360
    },
    {
      "entropy": 1.807152123749256,
      "epoch": 0.06004530522126013,
      "grad_norm": 4.91693639755249,
      "learning_rate": 9.381478252445994e-06,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8239471137523651,
      "num_tokens": 23256613.0,
      "step": 19370
    },
    {
      "entropy": 1.869692163169384,
      "epoch": 0.060076304346309825,
      "grad_norm": 8.64077377319336,
      "learning_rate": 9.386321805676644e-06,
      "loss": 0.6068,
      "mean_token_accuracy": 0.8187321960926056,
      "num_tokens": 23268668.0,
      "step": 19380
    },
    {
      "entropy": 1.857068532705307,
      "epoch": 0.06010730347135952,
      "grad_norm": 11.18220329284668,
      "learning_rate": 9.391165358907294e-06,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8069566667079926,
      "num_tokens": 23281042.0,
      "step": 19390
    },
    {
      "entropy": 1.7603770941495895,
      "epoch": 0.06013830259640922,
      "grad_norm": 4.09375524520874,
      "learning_rate": 9.396008912137946e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8220126062631607,
      "num_tokens": 23295383.0,
      "step": 19400
    },
    {
      "entropy": 1.8189956784248351,
      "epoch": 0.060169301721458915,
      "grad_norm": 13.177933692932129,
      "learning_rate": 9.400852465368596e-06,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8236996352672576,
      "num_tokens": 23307499.0,
      "step": 19410
    },
    {
      "entropy": 1.7352744668722153,
      "epoch": 0.060200300846508605,
      "grad_norm": 10.09691333770752,
      "learning_rate": 9.405696018599245e-06,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8295902729034423,
      "num_tokens": 23320862.0,
      "step": 19420
    },
    {
      "entropy": 1.8696271255612373,
      "epoch": 0.0602312999715583,
      "grad_norm": 11.082074165344238,
      "learning_rate": 9.410539571829895e-06,
      "loss": 0.5574,
      "mean_token_accuracy": 0.8225132539868355,
      "num_tokens": 23332219.0,
      "step": 19430
    },
    {
      "entropy": 1.8461137875914573,
      "epoch": 0.060262299096608,
      "grad_norm": 10.549184799194336,
      "learning_rate": 9.415383125060545e-06,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8168609410524368,
      "num_tokens": 23343965.0,
      "step": 19440
    },
    {
      "entropy": 1.920869068801403,
      "epoch": 0.060293298221657694,
      "grad_norm": 4.992405891418457,
      "learning_rate": 9.420226678291195e-06,
      "loss": 0.6075,
      "mean_token_accuracy": 0.8106775835156441,
      "num_tokens": 23355438.0,
      "step": 19450
    },
    {
      "entropy": 1.8544103041291238,
      "epoch": 0.06032429734670739,
      "grad_norm": 10.945703506469727,
      "learning_rate": 9.425070231521845e-06,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8255405530333519,
      "num_tokens": 23368404.0,
      "step": 19460
    },
    {
      "entropy": 1.841069608926773,
      "epoch": 0.06035529647175709,
      "grad_norm": 11.322453498840332,
      "learning_rate": 9.429913784752495e-06,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8234514787793159,
      "num_tokens": 23381143.0,
      "step": 19470
    },
    {
      "entropy": 1.8469559505581856,
      "epoch": 0.06038629559680678,
      "grad_norm": 3.433302164077759,
      "learning_rate": 9.434757337983145e-06,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8232303768396377,
      "num_tokens": 23393701.0,
      "step": 19480
    },
    {
      "entropy": 1.965156337618828,
      "epoch": 0.060417294721856474,
      "grad_norm": 11.884709358215332,
      "learning_rate": 9.439600891213795e-06,
      "loss": 0.6732,
      "mean_token_accuracy": 0.8007081165909767,
      "num_tokens": 23404106.0,
      "step": 19490
    },
    {
      "entropy": 1.9104407191276551,
      "epoch": 0.06044829384690617,
      "grad_norm": 9.23556900024414,
      "learning_rate": 9.444444444444445e-06,
      "loss": 0.6494,
      "mean_token_accuracy": 0.8021333515644073,
      "num_tokens": 23415825.0,
      "step": 19500
    },
    {
      "entropy": 1.885290040075779,
      "epoch": 0.06047929297195587,
      "grad_norm": 10.439990043640137,
      "learning_rate": 9.449287997675094e-06,
      "loss": 0.5873,
      "mean_token_accuracy": 0.8280000373721123,
      "num_tokens": 23426924.0,
      "step": 19510
    },
    {
      "entropy": 1.8925482839345933,
      "epoch": 0.060510292097005564,
      "grad_norm": 10.890645027160645,
      "learning_rate": 9.454131550905746e-06,
      "loss": 0.5907,
      "mean_token_accuracy": 0.8103524506092071,
      "num_tokens": 23439002.0,
      "step": 19520
    },
    {
      "entropy": 1.9133965462446212,
      "epoch": 0.06054129122205526,
      "grad_norm": 11.458105087280273,
      "learning_rate": 9.458975104136396e-06,
      "loss": 0.6338,
      "mean_token_accuracy": 0.8142278388142585,
      "num_tokens": 23450330.0,
      "step": 19530
    },
    {
      "entropy": 1.9326462358236314,
      "epoch": 0.06057229034710495,
      "grad_norm": 9.607535362243652,
      "learning_rate": 9.463818657367046e-06,
      "loss": 0.686,
      "mean_token_accuracy": 0.7996918767690658,
      "num_tokens": 23461629.0,
      "step": 19540
    },
    {
      "entropy": 1.876935575157404,
      "epoch": 0.060603289472154646,
      "grad_norm": 3.0272324085235596,
      "learning_rate": 9.468662210597696e-06,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8154638946056366,
      "num_tokens": 23474503.0,
      "step": 19550
    },
    {
      "entropy": 1.9294023901224135,
      "epoch": 0.06063428859720434,
      "grad_norm": 9.758378982543945,
      "learning_rate": 9.473505763828346e-06,
      "loss": 0.6173,
      "mean_token_accuracy": 0.8036470055580139,
      "num_tokens": 23486313.0,
      "step": 19560
    },
    {
      "entropy": 1.9118182629346847,
      "epoch": 0.06066528772225404,
      "grad_norm": 8.618282318115234,
      "learning_rate": 9.478349317058995e-06,
      "loss": 0.6425,
      "mean_token_accuracy": 0.8147132098674774,
      "num_tokens": 23497631.0,
      "step": 19570
    },
    {
      "entropy": 1.8255969345569611,
      "epoch": 0.060696286847303736,
      "grad_norm": 4.955978870391846,
      "learning_rate": 9.483192870289645e-06,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8196621656417846,
      "num_tokens": 23511011.0,
      "step": 19580
    },
    {
      "entropy": 1.7804076254367829,
      "epoch": 0.06072728597235343,
      "grad_norm": 9.806567192077637,
      "learning_rate": 9.488036423520295e-06,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8275730326771736,
      "num_tokens": 23523377.0,
      "step": 19590
    },
    {
      "entropy": 1.7767413407564163,
      "epoch": 0.06075828509740312,
      "grad_norm": 5.016587257385254,
      "learning_rate": 9.492879976750945e-06,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8214924201369286,
      "num_tokens": 23537122.0,
      "step": 19600
    },
    {
      "entropy": 1.8715672463178634,
      "epoch": 0.06078928422245282,
      "grad_norm": 6.352642059326172,
      "learning_rate": 9.497723529981595e-06,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8125947907567024,
      "num_tokens": 23548737.0,
      "step": 19610
    },
    {
      "entropy": 1.7911435902118682,
      "epoch": 0.060820283347502516,
      "grad_norm": 11.929505348205566,
      "learning_rate": 9.502567083212245e-06,
      "loss": 0.546,
      "mean_token_accuracy": 0.8234824985265732,
      "num_tokens": 23561524.0,
      "step": 19620
    },
    {
      "entropy": 1.8728584364056586,
      "epoch": 0.06085128247255221,
      "grad_norm": 9.866854667663574,
      "learning_rate": 9.507410636442895e-06,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8147120550274849,
      "num_tokens": 23573298.0,
      "step": 19630
    },
    {
      "entropy": 1.7628555655479432,
      "epoch": 0.06088228159760191,
      "grad_norm": 4.673979759216309,
      "learning_rate": 9.512254189673545e-06,
      "loss": 0.551,
      "mean_token_accuracy": 0.823185084760189,
      "num_tokens": 23585490.0,
      "step": 19640
    },
    {
      "entropy": 1.9162281841039657,
      "epoch": 0.060913280722651605,
      "grad_norm": 9.883894920349121,
      "learning_rate": 9.517097742904196e-06,
      "loss": 0.6322,
      "mean_token_accuracy": 0.8094197094440461,
      "num_tokens": 23596459.0,
      "step": 19650
    },
    {
      "entropy": 1.7629586443305016,
      "epoch": 0.0609442798477013,
      "grad_norm": 9.179272651672363,
      "learning_rate": 9.521941296134846e-06,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8276288509368896,
      "num_tokens": 23609459.0,
      "step": 19660
    },
    {
      "entropy": 1.8585124626755714,
      "epoch": 0.06097527897275099,
      "grad_norm": 10.178589820861816,
      "learning_rate": 9.526784849365496e-06,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8119170770049096,
      "num_tokens": 23622078.0,
      "step": 19670
    },
    {
      "entropy": 1.8904533594846726,
      "epoch": 0.06100627809780069,
      "grad_norm": 9.675638198852539,
      "learning_rate": 9.531628402596146e-06,
      "loss": 0.6203,
      "mean_token_accuracy": 0.8164917901158333,
      "num_tokens": 23632839.0,
      "step": 19680
    },
    {
      "entropy": 1.8115496248006822,
      "epoch": 0.061037277222850385,
      "grad_norm": 12.059412956237793,
      "learning_rate": 9.536471955826794e-06,
      "loss": 0.5708,
      "mean_token_accuracy": 0.8198250159621239,
      "num_tokens": 23644698.0,
      "step": 19690
    },
    {
      "entropy": 1.8150453560054303,
      "epoch": 0.06106827634790008,
      "grad_norm": 10.134939193725586,
      "learning_rate": 9.541315509057446e-06,
      "loss": 0.6001,
      "mean_token_accuracy": 0.815633225440979,
      "num_tokens": 23657908.0,
      "step": 19700
    },
    {
      "entropy": 1.7574081301689148,
      "epoch": 0.06109927547294978,
      "grad_norm": 5.226987838745117,
      "learning_rate": 9.546159062288096e-06,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8180340453982353,
      "num_tokens": 23671715.0,
      "step": 19710
    },
    {
      "entropy": 1.8007687643170356,
      "epoch": 0.061130274597999475,
      "grad_norm": 9.646768569946289,
      "learning_rate": 9.551002615518745e-06,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8148997142910958,
      "num_tokens": 23684332.0,
      "step": 19720
    },
    {
      "entropy": 1.836924096941948,
      "epoch": 0.061161273723049164,
      "grad_norm": 9.189767837524414,
      "learning_rate": 9.555846168749395e-06,
      "loss": 0.568,
      "mean_token_accuracy": 0.817747424542904,
      "num_tokens": 23696452.0,
      "step": 19730
    },
    {
      "entropy": 1.8786251202225686,
      "epoch": 0.06119227284809886,
      "grad_norm": 12.79636287689209,
      "learning_rate": 9.560689721980045e-06,
      "loss": 0.6144,
      "mean_token_accuracy": 0.8185794189572334,
      "num_tokens": 23707815.0,
      "step": 19740
    },
    {
      "entropy": 1.8025352910161019,
      "epoch": 0.06122327197314856,
      "grad_norm": 11.08669376373291,
      "learning_rate": 9.565533275210695e-06,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8157659009099006,
      "num_tokens": 23720644.0,
      "step": 19750
    },
    {
      "entropy": 1.9162618890404701,
      "epoch": 0.061254271098198254,
      "grad_norm": 10.929869651794434,
      "learning_rate": 9.570376828441345e-06,
      "loss": 0.6926,
      "mean_token_accuracy": 0.8083868056535721,
      "num_tokens": 23731943.0,
      "step": 19760
    },
    {
      "entropy": 1.8641342878341676,
      "epoch": 0.06128527022324795,
      "grad_norm": 9.528030395507812,
      "learning_rate": 9.575220381671997e-06,
      "loss": 0.5878,
      "mean_token_accuracy": 0.8118364483118057,
      "num_tokens": 23744425.0,
      "step": 19770
    },
    {
      "entropy": 1.8606672033667564,
      "epoch": 0.06131626934829765,
      "grad_norm": 12.20190715789795,
      "learning_rate": 9.580063934902646e-06,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8095232203602791,
      "num_tokens": 23755827.0,
      "step": 19780
    },
    {
      "entropy": 1.9077710419893266,
      "epoch": 0.06134726847334734,
      "grad_norm": 9.677120208740234,
      "learning_rate": 9.584907488133295e-06,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8066776558756829,
      "num_tokens": 23767808.0,
      "step": 19790
    },
    {
      "entropy": 1.8880801230669022,
      "epoch": 0.061378267598397034,
      "grad_norm": 10.036674499511719,
      "learning_rate": 9.589751041363944e-06,
      "loss": 0.5946,
      "mean_token_accuracy": 0.8205381706357002,
      "num_tokens": 23779742.0,
      "step": 19800
    },
    {
      "entropy": 1.9239039212465285,
      "epoch": 0.06140926672344673,
      "grad_norm": 10.33579158782959,
      "learning_rate": 9.594594594594594e-06,
      "loss": 0.6357,
      "mean_token_accuracy": 0.8081588789820671,
      "num_tokens": 23790560.0,
      "step": 19810
    },
    {
      "entropy": 1.8777620539069175,
      "epoch": 0.06144026584849643,
      "grad_norm": 9.710867881774902,
      "learning_rate": 9.599438147825246e-06,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8112061858177185,
      "num_tokens": 23802654.0,
      "step": 19820
    },
    {
      "entropy": 1.8921529993414878,
      "epoch": 0.06147126497354612,
      "grad_norm": 10.47066879272461,
      "learning_rate": 9.604281701055896e-06,
      "loss": 0.6444,
      "mean_token_accuracy": 0.8064588844776154,
      "num_tokens": 23813660.0,
      "step": 19830
    },
    {
      "entropy": 1.8871847927570342,
      "epoch": 0.06150226409859582,
      "grad_norm": 10.314284324645996,
      "learning_rate": 9.609125254286546e-06,
      "loss": 0.5919,
      "mean_token_accuracy": 0.8220043256878853,
      "num_tokens": 23825033.0,
      "step": 19840
    },
    {
      "entropy": 1.8442516967654228,
      "epoch": 0.06153326322364551,
      "grad_norm": 2.9218690395355225,
      "learning_rate": 9.613968807517196e-06,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8329377219080925,
      "num_tokens": 23837886.0,
      "step": 19850
    },
    {
      "entropy": 1.9648214638233186,
      "epoch": 0.061564262348695206,
      "grad_norm": 10.593524932861328,
      "learning_rate": 9.618812360747845e-06,
      "loss": 0.6819,
      "mean_token_accuracy": 0.8005811557173729,
      "num_tokens": 23848949.0,
      "step": 19860
    },
    {
      "entropy": 1.9761517822742463,
      "epoch": 0.0615952614737449,
      "grad_norm": 11.34875774383545,
      "learning_rate": 9.623655913978495e-06,
      "loss": 0.6411,
      "mean_token_accuracy": 0.8136752307415008,
      "num_tokens": 23860198.0,
      "step": 19870
    },
    {
      "entropy": 1.935739828646183,
      "epoch": 0.0616262605987946,
      "grad_norm": 10.864781379699707,
      "learning_rate": 9.628499467209145e-06,
      "loss": 0.6602,
      "mean_token_accuracy": 0.8089005783200264,
      "num_tokens": 23872637.0,
      "step": 19880
    },
    {
      "entropy": 1.8924452632665634,
      "epoch": 0.061657259723844296,
      "grad_norm": 10.361483573913574,
      "learning_rate": 9.633343020439797e-06,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8306096389889717,
      "num_tokens": 23885123.0,
      "step": 19890
    },
    {
      "entropy": 1.9373007863759995,
      "epoch": 0.06168825884889399,
      "grad_norm": 9.866893768310547,
      "learning_rate": 9.638186573670445e-06,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8127657011151314,
      "num_tokens": 23897519.0,
      "step": 19900
    },
    {
      "entropy": 1.9933653771877289,
      "epoch": 0.06171925797394368,
      "grad_norm": 10.006440162658691,
      "learning_rate": 9.643030126901095e-06,
      "loss": 0.6684,
      "mean_token_accuracy": 0.810961103439331,
      "num_tokens": 23908829.0,
      "step": 19910
    },
    {
      "entropy": 1.9257026076316834,
      "epoch": 0.06175025709899338,
      "grad_norm": 10.654083251953125,
      "learning_rate": 9.647873680131745e-06,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8109946802258492,
      "num_tokens": 23920767.0,
      "step": 19920
    },
    {
      "entropy": 1.956542044878006,
      "epoch": 0.061781256224043075,
      "grad_norm": 11.350798606872559,
      "learning_rate": 9.652717233362395e-06,
      "loss": 0.6668,
      "mean_token_accuracy": 0.8026480153203011,
      "num_tokens": 23932164.0,
      "step": 19930
    },
    {
      "entropy": 2.0048277229070663,
      "epoch": 0.06181225534909277,
      "grad_norm": 11.242584228515625,
      "learning_rate": 9.657560786593046e-06,
      "loss": 0.6457,
      "mean_token_accuracy": 0.8089631497859955,
      "num_tokens": 23942822.0,
      "step": 19940
    },
    {
      "entropy": 1.9254214867949486,
      "epoch": 0.06184325447414247,
      "grad_norm": 10.975452423095703,
      "learning_rate": 9.662404339823696e-06,
      "loss": 0.6117,
      "mean_token_accuracy": 0.8136309564113617,
      "num_tokens": 23954872.0,
      "step": 19950
    },
    {
      "entropy": 1.8969217911362648,
      "epoch": 0.061874253599192165,
      "grad_norm": 9.684366226196289,
      "learning_rate": 9.667247893054346e-06,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8263826936483383,
      "num_tokens": 23967198.0,
      "step": 19960
    },
    {
      "entropy": 1.9214407287538051,
      "epoch": 0.061905252724241855,
      "grad_norm": 10.67751693725586,
      "learning_rate": 9.672091446284996e-06,
      "loss": 0.582,
      "mean_token_accuracy": 0.8223884150385856,
      "num_tokens": 23979194.0,
      "step": 19970
    },
    {
      "entropy": 1.8657239809632302,
      "epoch": 0.06193625184929155,
      "grad_norm": 9.526717185974121,
      "learning_rate": 9.676934999515646e-06,
      "loss": 0.538,
      "mean_token_accuracy": 0.8274316728115082,
      "num_tokens": 23991363.0,
      "step": 19980
    },
    {
      "entropy": 1.9447234928607942,
      "epoch": 0.06196725097434125,
      "grad_norm": 11.365060806274414,
      "learning_rate": 9.681778552746296e-06,
      "loss": 0.6874,
      "mean_token_accuracy": 0.8004522323608398,
      "num_tokens": 24002188.0,
      "step": 19990
    },
    {
      "entropy": 1.8590040877461433,
      "epoch": 0.061998250099390945,
      "grad_norm": 6.042444705963135,
      "learning_rate": 9.686622105976946e-06,
      "loss": 0.5923,
      "mean_token_accuracy": 0.814041730761528,
      "num_tokens": 24015280.0,
      "step": 20000
    },
    {
      "entropy": 1.9083186939358712,
      "epoch": 0.06202924922444064,
      "grad_norm": 10.619732856750488,
      "learning_rate": 9.691465659207595e-06,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8160168409347535,
      "num_tokens": 24027453.0,
      "step": 20010
    },
    {
      "entropy": 1.922300359606743,
      "epoch": 0.06206024834949034,
      "grad_norm": 12.882160186767578,
      "learning_rate": 9.696309212438245e-06,
      "loss": 0.649,
      "mean_token_accuracy": 0.7888608485460281,
      "num_tokens": 24038637.0,
      "step": 20020
    },
    {
      "entropy": 1.9450786724686622,
      "epoch": 0.062091247474540034,
      "grad_norm": 9.845813751220703,
      "learning_rate": 9.701152765668895e-06,
      "loss": 0.645,
      "mean_token_accuracy": 0.8010739460587502,
      "num_tokens": 24050162.0,
      "step": 20030
    },
    {
      "entropy": 1.9187857955694199,
      "epoch": 0.062122246599589724,
      "grad_norm": 9.616793632507324,
      "learning_rate": 9.705996318899545e-06,
      "loss": 0.6361,
      "mean_token_accuracy": 0.8148136526346207,
      "num_tokens": 24061516.0,
      "step": 20040
    },
    {
      "entropy": 1.90914705991745,
      "epoch": 0.06215324572463942,
      "grad_norm": 12.597700119018555,
      "learning_rate": 9.710839872130195e-06,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8118062347173691,
      "num_tokens": 24072040.0,
      "step": 20050
    },
    {
      "entropy": 1.7790238752961158,
      "epoch": 0.06218424484968912,
      "grad_norm": 14.254045486450195,
      "learning_rate": 9.715683425360845e-06,
      "loss": 0.5597,
      "mean_token_accuracy": 0.824961057305336,
      "num_tokens": 24084516.0,
      "step": 20060
    },
    {
      "entropy": 1.8951515421271323,
      "epoch": 0.062215243974738814,
      "grad_norm": 10.398776054382324,
      "learning_rate": 9.720526978591496e-06,
      "loss": 0.6339,
      "mean_token_accuracy": 0.8136864483356476,
      "num_tokens": 24096174.0,
      "step": 20070
    },
    {
      "entropy": 1.8427141726016998,
      "epoch": 0.06224624309978851,
      "grad_norm": 9.789461135864258,
      "learning_rate": 9.725370531822146e-06,
      "loss": 0.5943,
      "mean_token_accuracy": 0.8145849913358688,
      "num_tokens": 24107922.0,
      "step": 20080
    },
    {
      "entropy": 1.9320272147655486,
      "epoch": 0.06227724222483821,
      "grad_norm": 11.617203712463379,
      "learning_rate": 9.730214085052796e-06,
      "loss": 0.668,
      "mean_token_accuracy": 0.802889634668827,
      "num_tokens": 24119219.0,
      "step": 20090
    },
    {
      "entropy": 1.8822171539068222,
      "epoch": 0.0623082413498879,
      "grad_norm": 10.93297004699707,
      "learning_rate": 9.735057638283446e-06,
      "loss": 0.5582,
      "mean_token_accuracy": 0.8173771098256111,
      "num_tokens": 24131388.0,
      "step": 20100
    },
    {
      "entropy": 1.9237861156463623,
      "epoch": 0.06233924047493759,
      "grad_norm": 10.885433197021484,
      "learning_rate": 9.739901191514094e-06,
      "loss": 0.6236,
      "mean_token_accuracy": 0.8068635702133179,
      "num_tokens": 24143423.0,
      "step": 20110
    },
    {
      "entropy": 1.8605427652597428,
      "epoch": 0.06237023959998729,
      "grad_norm": 10.937981605529785,
      "learning_rate": 9.744744744744746e-06,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8124633759260178,
      "num_tokens": 24155500.0,
      "step": 20120
    },
    {
      "entropy": 1.921093289554119,
      "epoch": 0.06240123872503699,
      "grad_norm": 4.816215515136719,
      "learning_rate": 9.749588297975396e-06,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8212477296590805,
      "num_tokens": 24167119.0,
      "step": 20130
    },
    {
      "entropy": 1.822756864130497,
      "epoch": 0.06243223785008668,
      "grad_norm": 10.295876502990723,
      "learning_rate": 9.754431851206046e-06,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8297939330339432,
      "num_tokens": 24179336.0,
      "step": 20140
    },
    {
      "entropy": 1.8902019761502742,
      "epoch": 0.06246323697513638,
      "grad_norm": 9.802702903747559,
      "learning_rate": 9.759275404436695e-06,
      "loss": 0.592,
      "mean_token_accuracy": 0.8155349537730217,
      "num_tokens": 24191945.0,
      "step": 20150
    },
    {
      "entropy": 1.9736180931329728,
      "epoch": 0.06249423610018607,
      "grad_norm": 10.913537979125977,
      "learning_rate": 9.764118957667345e-06,
      "loss": 0.646,
      "mean_token_accuracy": 0.8018572524189949,
      "num_tokens": 24203220.0,
      "step": 20160
    },
    {
      "entropy": 1.8472725585103036,
      "epoch": 0.06252523522523577,
      "grad_norm": 11.09897518157959,
      "learning_rate": 9.768962510897995e-06,
      "loss": 0.6617,
      "mean_token_accuracy": 0.8151746213436126,
      "num_tokens": 24216924.0,
      "step": 20170
    },
    {
      "entropy": 1.8363325014710425,
      "epoch": 0.06255623435028547,
      "grad_norm": 9.55675220489502,
      "learning_rate": 9.773806064128645e-06,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8332865908741951,
      "num_tokens": 24229749.0,
      "step": 20180
    },
    {
      "entropy": 1.8290324866771699,
      "epoch": 0.06258723347533515,
      "grad_norm": 12.252378463745117,
      "learning_rate": 9.778649617359297e-06,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8365808725357056,
      "num_tokens": 24242819.0,
      "step": 20190
    },
    {
      "entropy": 1.9081189304590225,
      "epoch": 0.06261823260038485,
      "grad_norm": 10.117230415344238,
      "learning_rate": 9.783493170589947e-06,
      "loss": 0.6745,
      "mean_token_accuracy": 0.8055532008409501,
      "num_tokens": 24253826.0,
      "step": 20200
    },
    {
      "entropy": 1.8748064145445824,
      "epoch": 0.06264923172543455,
      "grad_norm": 9.556116104125977,
      "learning_rate": 9.788336723820595e-06,
      "loss": 0.6262,
      "mean_token_accuracy": 0.8184110090136528,
      "num_tokens": 24264932.0,
      "step": 20210
    },
    {
      "entropy": 1.8884572684764862,
      "epoch": 0.06268023085048424,
      "grad_norm": 4.309450149536133,
      "learning_rate": 9.793180277051245e-06,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8134237229824066,
      "num_tokens": 24277309.0,
      "step": 20220
    },
    {
      "entropy": 1.946840487420559,
      "epoch": 0.06271122997553394,
      "grad_norm": 10.026110649108887,
      "learning_rate": 9.798023830281895e-06,
      "loss": 0.6271,
      "mean_token_accuracy": 0.8044777557253837,
      "num_tokens": 24288780.0,
      "step": 20230
    },
    {
      "entropy": 2.013564696907997,
      "epoch": 0.06274222910058364,
      "grad_norm": 9.077942848205566,
      "learning_rate": 9.802867383512546e-06,
      "loss": 0.6781,
      "mean_token_accuracy": 0.800484599173069,
      "num_tokens": 24299775.0,
      "step": 20240
    },
    {
      "entropy": 1.8675778850913047,
      "epoch": 0.06277322822563333,
      "grad_norm": 9.851303100585938,
      "learning_rate": 9.807710936743196e-06,
      "loss": 0.5568,
      "mean_token_accuracy": 0.8174749821424484,
      "num_tokens": 24312128.0,
      "step": 20250
    },
    {
      "entropy": 1.9423525124788283,
      "epoch": 0.06280422735068303,
      "grad_norm": 7.952591419219971,
      "learning_rate": 9.812554489973846e-06,
      "loss": 0.6073,
      "mean_token_accuracy": 0.8145397856831551,
      "num_tokens": 24323096.0,
      "step": 20260
    },
    {
      "entropy": 1.7513898000121118,
      "epoch": 0.06283522647573273,
      "grad_norm": 5.437192916870117,
      "learning_rate": 9.817398043204496e-06,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8165362849831581,
      "num_tokens": 24337460.0,
      "step": 20270
    },
    {
      "entropy": 1.907211183011532,
      "epoch": 0.06286622560078242,
      "grad_norm": 5.400768756866455,
      "learning_rate": 9.822241596435146e-06,
      "loss": 0.6175,
      "mean_token_accuracy": 0.8163802400231361,
      "num_tokens": 24349174.0,
      "step": 20280
    },
    {
      "entropy": 1.93508680164814,
      "epoch": 0.06289722472583212,
      "grad_norm": 10.331210136413574,
      "learning_rate": 9.827085149665796e-06,
      "loss": 0.6279,
      "mean_token_accuracy": 0.8010757058858872,
      "num_tokens": 24361061.0,
      "step": 20290
    },
    {
      "entropy": 1.9249700546264648,
      "epoch": 0.06292822385088181,
      "grad_norm": 9.38033390045166,
      "learning_rate": 9.831928702896445e-06,
      "loss": 0.608,
      "mean_token_accuracy": 0.8222104609012604,
      "num_tokens": 24372277.0,
      "step": 20300
    },
    {
      "entropy": 1.8913897737860679,
      "epoch": 0.0629592229759315,
      "grad_norm": 4.759194374084473,
      "learning_rate": 9.836772256127097e-06,
      "loss": 0.6044,
      "mean_token_accuracy": 0.8175428286194801,
      "num_tokens": 24384034.0,
      "step": 20310
    },
    {
      "entropy": 1.8855891197919845,
      "epoch": 0.0629902221009812,
      "grad_norm": 4.686741828918457,
      "learning_rate": 9.841615809357745e-06,
      "loss": 0.6091,
      "mean_token_accuracy": 0.8151167094707489,
      "num_tokens": 24395926.0,
      "step": 20320
    },
    {
      "entropy": 1.9041914016008377,
      "epoch": 0.06302122122603089,
      "grad_norm": 4.964253902435303,
      "learning_rate": 9.846459362588395e-06,
      "loss": 0.6008,
      "mean_token_accuracy": 0.8038349270820617,
      "num_tokens": 24407924.0,
      "step": 20330
    },
    {
      "entropy": 1.8939720645546914,
      "epoch": 0.06305222035108059,
      "grad_norm": 10.10213851928711,
      "learning_rate": 9.851302915819045e-06,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8080206617712975,
      "num_tokens": 24419254.0,
      "step": 20340
    },
    {
      "entropy": 1.8911166504025458,
      "epoch": 0.06308321947613028,
      "grad_norm": 4.227776050567627,
      "learning_rate": 9.856146469049695e-06,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8054097011685372,
      "num_tokens": 24431388.0,
      "step": 20350
    },
    {
      "entropy": 1.8960236981511116,
      "epoch": 0.06311421860117998,
      "grad_norm": 12.084880828857422,
      "learning_rate": 9.860990022280346e-06,
      "loss": 0.601,
      "mean_token_accuracy": 0.8084520295262336,
      "num_tokens": 24443385.0,
      "step": 20360
    },
    {
      "entropy": 1.871785145998001,
      "epoch": 0.06314521772622968,
      "grad_norm": 10.692331314086914,
      "learning_rate": 9.865833575510996e-06,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8134460240602494,
      "num_tokens": 24454988.0,
      "step": 20370
    },
    {
      "entropy": 1.9693672031164169,
      "epoch": 0.06317621685127937,
      "grad_norm": 10.991315841674805,
      "learning_rate": 9.870677128741646e-06,
      "loss": 0.7165,
      "mean_token_accuracy": 0.7969113975763321,
      "num_tokens": 24465991.0,
      "step": 20380
    },
    {
      "entropy": 1.9215772941708564,
      "epoch": 0.06320721597632907,
      "grad_norm": 11.294339179992676,
      "learning_rate": 9.875520681972296e-06,
      "loss": 0.6661,
      "mean_token_accuracy": 0.7995204910635948,
      "num_tokens": 24478784.0,
      "step": 20390
    },
    {
      "entropy": 1.750515715777874,
      "epoch": 0.06323821510137877,
      "grad_norm": 9.189525604248047,
      "learning_rate": 9.880364235202946e-06,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8383323296904563,
      "num_tokens": 24492223.0,
      "step": 20400
    },
    {
      "entropy": 1.8770269468426704,
      "epoch": 0.06326921422642846,
      "grad_norm": 9.287210464477539,
      "learning_rate": 9.885207788433596e-06,
      "loss": 0.6492,
      "mean_token_accuracy": 0.8062692806124687,
      "num_tokens": 24503912.0,
      "step": 20410
    },
    {
      "entropy": 1.8530636951327324,
      "epoch": 0.06330021335147816,
      "grad_norm": 7.70045280456543,
      "learning_rate": 9.890051341664246e-06,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8267319366335869,
      "num_tokens": 24515605.0,
      "step": 20420
    },
    {
      "entropy": 1.9133968889713286,
      "epoch": 0.06333121247652786,
      "grad_norm": 10.080760955810547,
      "learning_rate": 9.894894894894896e-06,
      "loss": 0.6142,
      "mean_token_accuracy": 0.8130926743149758,
      "num_tokens": 24527065.0,
      "step": 20430
    },
    {
      "entropy": 1.876038283109665,
      "epoch": 0.06336221160157754,
      "grad_norm": 9.016529083251953,
      "learning_rate": 9.899738448125546e-06,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8140005797147751,
      "num_tokens": 24538739.0,
      "step": 20440
    },
    {
      "entropy": 1.8368092089891435,
      "epoch": 0.06339321072662724,
      "grad_norm": 9.778054237365723,
      "learning_rate": 9.904582001356195e-06,
      "loss": 0.6006,
      "mean_token_accuracy": 0.81819748878479,
      "num_tokens": 24551215.0,
      "step": 20450
    },
    {
      "entropy": 1.7856343001127244,
      "epoch": 0.06342420985167693,
      "grad_norm": 4.523138523101807,
      "learning_rate": 9.909425554586845e-06,
      "loss": 0.509,
      "mean_token_accuracy": 0.8145459353923797,
      "num_tokens": 24564545.0,
      "step": 20460
    },
    {
      "entropy": 1.8864993780851365,
      "epoch": 0.06345520897672663,
      "grad_norm": 9.66295051574707,
      "learning_rate": 9.914269107817495e-06,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8107953786849975,
      "num_tokens": 24576201.0,
      "step": 20470
    },
    {
      "entropy": 1.8306699201464653,
      "epoch": 0.06348620810177633,
      "grad_norm": 8.954712867736816,
      "learning_rate": 9.919112661048145e-06,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8272392004728317,
      "num_tokens": 24588462.0,
      "step": 20480
    },
    {
      "entropy": 1.8582230091094971,
      "epoch": 0.06351720722682602,
      "grad_norm": 5.900847911834717,
      "learning_rate": 9.923956214278797e-06,
      "loss": 0.6333,
      "mean_token_accuracy": 0.8117007941007615,
      "num_tokens": 24601048.0,
      "step": 20490
    },
    {
      "entropy": 1.9099309206008912,
      "epoch": 0.06354820635187572,
      "grad_norm": 10.398176193237305,
      "learning_rate": 9.928799767509447e-06,
      "loss": 0.6342,
      "mean_token_accuracy": 0.8157493248581886,
      "num_tokens": 24612082.0,
      "step": 20500
    },
    {
      "entropy": 1.8192737758159638,
      "epoch": 0.06357920547692542,
      "grad_norm": 4.901766300201416,
      "learning_rate": 9.933643320740096e-06,
      "loss": 0.5962,
      "mean_token_accuracy": 0.8077544063329697,
      "num_tokens": 24624244.0,
      "step": 20510
    },
    {
      "entropy": 1.9380394339561462,
      "epoch": 0.06361020460197511,
      "grad_norm": 9.72773551940918,
      "learning_rate": 9.938486873970746e-06,
      "loss": 0.6725,
      "mean_token_accuracy": 0.8050031334161758,
      "num_tokens": 24635227.0,
      "step": 20520
    },
    {
      "entropy": 1.8201100319623946,
      "epoch": 0.06364120372702481,
      "grad_norm": 5.236706256866455,
      "learning_rate": 9.943330427201394e-06,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8271597489714623,
      "num_tokens": 24647349.0,
      "step": 20530
    },
    {
      "entropy": 1.955614548921585,
      "epoch": 0.0636722028520745,
      "grad_norm": 13.138460159301758,
      "learning_rate": 9.948173980432046e-06,
      "loss": 0.7057,
      "mean_token_accuracy": 0.796806488931179,
      "num_tokens": 24658536.0,
      "step": 20540
    },
    {
      "entropy": 1.8944236859679222,
      "epoch": 0.0637032019771242,
      "grad_norm": 7.109180927276611,
      "learning_rate": 9.953017533662696e-06,
      "loss": 0.6351,
      "mean_token_accuracy": 0.8019287914037705,
      "num_tokens": 24670364.0,
      "step": 20550
    },
    {
      "entropy": 1.8803847134113312,
      "epoch": 0.06373420110217388,
      "grad_norm": 9.590316772460938,
      "learning_rate": 9.957861086893346e-06,
      "loss": 0.6256,
      "mean_token_accuracy": 0.8110292464494705,
      "num_tokens": 24682351.0,
      "step": 20560
    },
    {
      "entropy": 1.9103557452559472,
      "epoch": 0.06376520022722358,
      "grad_norm": 10.402158737182617,
      "learning_rate": 9.962704640123996e-06,
      "loss": 0.6655,
      "mean_token_accuracy": 0.7996529176831245,
      "num_tokens": 24693410.0,
      "step": 20570
    },
    {
      "entropy": 1.8738579228520393,
      "epoch": 0.06379619935227328,
      "grad_norm": 8.206809997558594,
      "learning_rate": 9.967548193354646e-06,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8173448964953423,
      "num_tokens": 24705775.0,
      "step": 20580
    },
    {
      "entropy": 1.8843411058187485,
      "epoch": 0.06382719847732297,
      "grad_norm": 11.01453971862793,
      "learning_rate": 9.972391746585295e-06,
      "loss": 0.6117,
      "mean_token_accuracy": 0.808614219725132,
      "num_tokens": 24717660.0,
      "step": 20590
    },
    {
      "entropy": 1.8635393604636192,
      "epoch": 0.06385819760237267,
      "grad_norm": 10.689123153686523,
      "learning_rate": 9.977235299815945e-06,
      "loss": 0.6426,
      "mean_token_accuracy": 0.8036311730742455,
      "num_tokens": 24729686.0,
      "step": 20600
    },
    {
      "entropy": 1.9377967566251755,
      "epoch": 0.06388919672742237,
      "grad_norm": 11.218433380126953,
      "learning_rate": 9.982078853046597e-06,
      "loss": 0.6821,
      "mean_token_accuracy": 0.8068268105387688,
      "num_tokens": 24739992.0,
      "step": 20610
    },
    {
      "entropy": 1.8930637896060944,
      "epoch": 0.06392019585247206,
      "grad_norm": 10.52428150177002,
      "learning_rate": 9.986922406277247e-06,
      "loss": 0.6418,
      "mean_token_accuracy": 0.802625036239624,
      "num_tokens": 24751815.0,
      "step": 20620
    },
    {
      "entropy": 1.9082047209143638,
      "epoch": 0.06395119497752176,
      "grad_norm": 9.473612785339355,
      "learning_rate": 9.991765959507895e-06,
      "loss": 0.612,
      "mean_token_accuracy": 0.8163775265216827,
      "num_tokens": 24763588.0,
      "step": 20630
    },
    {
      "entropy": 1.8783060640096665,
      "epoch": 0.06398219410257146,
      "grad_norm": 10.244823455810547,
      "learning_rate": 9.996609512738545e-06,
      "loss": 0.6235,
      "mean_token_accuracy": 0.8123615190386773,
      "num_tokens": 24776091.0,
      "step": 20640
    },
    {
      "entropy": 1.8689485356211661,
      "epoch": 0.06401319322762115,
      "grad_norm": 9.722251892089844,
      "learning_rate": 9.999273546183343e-06,
      "loss": 0.5782,
      "mean_token_accuracy": 0.8126691862940788,
      "num_tokens": 24787537.0,
      "step": 20650
    },
    {
      "entropy": 1.869950420409441,
      "epoch": 0.06404419235267085,
      "grad_norm": 9.074020385742188,
      "learning_rate": 9.99685317639837e-06,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8139035999774933,
      "num_tokens": 24799896.0,
      "step": 20660
    },
    {
      "entropy": 1.8778914123773576,
      "epoch": 0.06407519147772055,
      "grad_norm": 11.21394157409668,
      "learning_rate": 9.99443456334761e-06,
      "loss": 0.543,
      "mean_token_accuracy": 0.8123081102967262,
      "num_tokens": 24811525.0,
      "step": 20670
    },
    {
      "entropy": 1.9493975609540939,
      "epoch": 0.06410619060277023,
      "grad_norm": 13.748980522155762,
      "learning_rate": 9.992017704906994e-06,
      "loss": 0.647,
      "mean_token_accuracy": 0.8159363105893135,
      "num_tokens": 24822680.0,
      "step": 20680
    },
    {
      "entropy": 1.8906102269887923,
      "epoch": 0.06413718972781993,
      "grad_norm": 8.22630786895752,
      "learning_rate": 9.989602598956046e-06,
      "loss": 0.6267,
      "mean_token_accuracy": 0.8106621414422989,
      "num_tokens": 24834924.0,
      "step": 20690
    },
    {
      "entropy": 1.7157309882342815,
      "epoch": 0.06416818885286962,
      "grad_norm": 7.910831928253174,
      "learning_rate": 9.987189243377873e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8414781808853149,
      "num_tokens": 24849152.0,
      "step": 20700
    },
    {
      "entropy": 1.9214319348335267,
      "epoch": 0.06419918797791932,
      "grad_norm": 10.313517570495605,
      "learning_rate": 9.984777636059161e-06,
      "loss": 0.7036,
      "mean_token_accuracy": 0.7919689938426018,
      "num_tokens": 24860357.0,
      "step": 20710
    },
    {
      "entropy": 1.8385179117321968,
      "epoch": 0.06423018710296902,
      "grad_norm": 12.135076522827148,
      "learning_rate": 9.98236777489017e-06,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8177133709192276,
      "num_tokens": 24872095.0,
      "step": 20720
    },
    {
      "entropy": 1.777934755384922,
      "epoch": 0.06426118622801871,
      "grad_norm": 4.797874450683594,
      "learning_rate": 9.979959657764716e-06,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8323575839400291,
      "num_tokens": 24884741.0,
      "step": 20730
    },
    {
      "entropy": 1.8482744053006173,
      "epoch": 0.06429218535306841,
      "grad_norm": 9.592241287231445,
      "learning_rate": 9.977553282580177e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8222592979669571,
      "num_tokens": 24897187.0,
      "step": 20740
    },
    {
      "entropy": 1.8500724270939828,
      "epoch": 0.0643231844781181,
      "grad_norm": 12.197331428527832,
      "learning_rate": 9.975148647237474e-06,
      "loss": 0.6756,
      "mean_token_accuracy": 0.7995992138981819,
      "num_tokens": 24908381.0,
      "step": 20750
    },
    {
      "entropy": 1.75242570489645,
      "epoch": 0.0643541836031678,
      "grad_norm": 10.444792747497559,
      "learning_rate": 9.972745749641067e-06,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8243147745728493,
      "num_tokens": 24920923.0,
      "step": 20760
    },
    {
      "entropy": 1.8526248589158059,
      "epoch": 0.0643851827282175,
      "grad_norm": 9.078262329101562,
      "learning_rate": 9.97034458769895e-06,
      "loss": 0.6349,
      "mean_token_accuracy": 0.8068570986390113,
      "num_tokens": 24932578.0,
      "step": 20770
    },
    {
      "entropy": 1.9244740456342697,
      "epoch": 0.0644161818532672,
      "grad_norm": 10.673161506652832,
      "learning_rate": 9.967945159322642e-06,
      "loss": 0.687,
      "mean_token_accuracy": 0.7971266448497772,
      "num_tokens": 24942881.0,
      "step": 20780
    },
    {
      "entropy": 1.8756011351943016,
      "epoch": 0.06444718097831689,
      "grad_norm": 9.983181953430176,
      "learning_rate": 9.965547462427177e-06,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8142784401774407,
      "num_tokens": 24954564.0,
      "step": 20790
    },
    {
      "entropy": 1.8296712294220925,
      "epoch": 0.06447818010336659,
      "grad_norm": 10.166830062866211,
      "learning_rate": 9.963151494931094e-06,
      "loss": 0.6427,
      "mean_token_accuracy": 0.8073472276329994,
      "num_tokens": 24967386.0,
      "step": 20800
    },
    {
      "entropy": 1.8981720179319381,
      "epoch": 0.06450917922841627,
      "grad_norm": 11.244156837463379,
      "learning_rate": 9.960757254756438e-06,
      "loss": 0.669,
      "mean_token_accuracy": 0.8046011224389076,
      "num_tokens": 24978303.0,
      "step": 20810
    },
    {
      "entropy": 1.7593111276626587,
      "epoch": 0.06454017835346597,
      "grad_norm": 11.301054954528809,
      "learning_rate": 9.958364739828752e-06,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8244746640324593,
      "num_tokens": 24991174.0,
      "step": 20820
    },
    {
      "entropy": 1.7012663453817367,
      "epoch": 0.06457117747851567,
      "grad_norm": 11.489786148071289,
      "learning_rate": 9.955973948077055e-06,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8441385194659233,
      "num_tokens": 25003945.0,
      "step": 20830
    },
    {
      "entropy": 1.8124509736895562,
      "epoch": 0.06460217660356536,
      "grad_norm": 10.668350219726562,
      "learning_rate": 9.953584877433851e-06,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8089659824967385,
      "num_tokens": 25016118.0,
      "step": 20840
    },
    {
      "entropy": 1.745204885303974,
      "epoch": 0.06463317572861506,
      "grad_norm": 10.901769638061523,
      "learning_rate": 9.951197525835119e-06,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8224120557308197,
      "num_tokens": 25029471.0,
      "step": 20850
    },
    {
      "entropy": 1.835092043876648,
      "epoch": 0.06466417485366475,
      "grad_norm": 10.224356651306152,
      "learning_rate": 9.94881189122029e-06,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8145354777574539,
      "num_tokens": 25041531.0,
      "step": 20860
    },
    {
      "entropy": 1.7869165703654288,
      "epoch": 0.06469517397871445,
      "grad_norm": 12.071456909179688,
      "learning_rate": 9.946427971532263e-06,
      "loss": 0.6134,
      "mean_token_accuracy": 0.8095423832535744,
      "num_tokens": 25054431.0,
      "step": 20870
    },
    {
      "entropy": 1.8208570063114167,
      "epoch": 0.06472617310376415,
      "grad_norm": 10.90512466430664,
      "learning_rate": 9.944045764717379e-06,
      "loss": 0.6059,
      "mean_token_accuracy": 0.813726843893528,
      "num_tokens": 25066679.0,
      "step": 20880
    },
    {
      "entropy": 1.7698472633957862,
      "epoch": 0.06475717222881384,
      "grad_norm": 4.871018409729004,
      "learning_rate": 9.941665268725422e-06,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8185109004378319,
      "num_tokens": 25078924.0,
      "step": 20890
    },
    {
      "entropy": 1.7922344714403153,
      "epoch": 0.06478817135386354,
      "grad_norm": Infinity,
      "learning_rate": 9.939286481509611e-06,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8192875310778618,
      "num_tokens": 25091264.0,
      "step": 20900
    },
    {
      "entropy": 1.873502266407013,
      "epoch": 0.06481917047891324,
      "grad_norm": 10.781414985656738,
      "learning_rate": 9.93690940102659e-06,
      "loss": 0.6287,
      "mean_token_accuracy": 0.8078038066625595,
      "num_tokens": 25102391.0,
      "step": 20910
    },
    {
      "entropy": 1.8319077044725418,
      "epoch": 0.06485016960396293,
      "grad_norm": 5.344398021697998,
      "learning_rate": 9.934534025236426e-06,
      "loss": 0.586,
      "mean_token_accuracy": 0.8086917147040367,
      "num_tokens": 25114388.0,
      "step": 20920
    },
    {
      "entropy": 1.8269810006022453,
      "epoch": 0.06488116872901262,
      "grad_norm": 10.4851655960083,
      "learning_rate": 9.93216035210259e-06,
      "loss": 0.6322,
      "mean_token_accuracy": 0.8202467620372772,
      "num_tokens": 25126354.0,
      "step": 20930
    },
    {
      "entropy": 1.842079259455204,
      "epoch": 0.06491216785406231,
      "grad_norm": 9.595065116882324,
      "learning_rate": 9.929788379591967e-06,
      "loss": 0.6143,
      "mean_token_accuracy": 0.81163260191679,
      "num_tokens": 25138345.0,
      "step": 20940
    },
    {
      "entropy": 1.8531109601259232,
      "epoch": 0.06494316697911201,
      "grad_norm": 10.80106258392334,
      "learning_rate": 9.92741810567483e-06,
      "loss": 0.5955,
      "mean_token_accuracy": 0.8235780015587807,
      "num_tokens": 25149452.0,
      "step": 20950
    },
    {
      "entropy": 1.7707980051636696,
      "epoch": 0.0649741661041617,
      "grad_norm": 9.250375747680664,
      "learning_rate": 9.925049528324852e-06,
      "loss": 0.5531,
      "mean_token_accuracy": 0.825781124830246,
      "num_tokens": 25162412.0,
      "step": 20960
    },
    {
      "entropy": 1.7839957982301713,
      "epoch": 0.0650051652292114,
      "grad_norm": 11.089311599731445,
      "learning_rate": 9.922682645519076e-06,
      "loss": 0.535,
      "mean_token_accuracy": 0.8361950904130936,
      "num_tokens": 25174698.0,
      "step": 20970
    },
    {
      "entropy": 1.8392813488841058,
      "epoch": 0.0650361643542611,
      "grad_norm": 9.709254264831543,
      "learning_rate": 9.920317455237932e-06,
      "loss": 0.6204,
      "mean_token_accuracy": 0.8103036060929298,
      "num_tokens": 25187089.0,
      "step": 20980
    },
    {
      "entropy": 1.89919556081295,
      "epoch": 0.0650671634793108,
      "grad_norm": 10.977981567382812,
      "learning_rate": 9.917953955465215e-06,
      "loss": 0.69,
      "mean_token_accuracy": 0.7884634464979172,
      "num_tokens": 25198974.0,
      "step": 20990
    },
    {
      "entropy": 1.8912572488188744,
      "epoch": 0.0650981626043605,
      "grad_norm": 11.527606964111328,
      "learning_rate": 9.915592144188078e-06,
      "loss": 0.6459,
      "mean_token_accuracy": 0.8056445106863975,
      "num_tokens": 25209971.0,
      "step": 21000
    },
    {
      "entropy": 1.8872391551733017,
      "epoch": 0.06512916172941019,
      "grad_norm": 10.813753128051758,
      "learning_rate": 9.913232019397025e-06,
      "loss": 0.7104,
      "mean_token_accuracy": 0.7989979892969131,
      "num_tokens": 25220791.0,
      "step": 21010
    },
    {
      "entropy": 1.8582836613059044,
      "epoch": 0.06516016085445989,
      "grad_norm": 9.749269485473633,
      "learning_rate": 9.910873579085914e-06,
      "loss": 0.6316,
      "mean_token_accuracy": 0.8038420900702477,
      "num_tokens": 25233097.0,
      "step": 21020
    },
    {
      "entropy": 1.9419476687908173,
      "epoch": 0.06519115997950958,
      "grad_norm": 10.637131690979004,
      "learning_rate": 9.908516821251943e-06,
      "loss": 0.6322,
      "mean_token_accuracy": 0.811190040409565,
      "num_tokens": 25244036.0,
      "step": 21030
    },
    {
      "entropy": 1.8555924728512765,
      "epoch": 0.06522215910455928,
      "grad_norm": 11.770044326782227,
      "learning_rate": 9.906161743895632e-06,
      "loss": 0.6444,
      "mean_token_accuracy": 0.8096898928284645,
      "num_tokens": 25255598.0,
      "step": 21040
    },
    {
      "entropy": 1.8110544815659524,
      "epoch": 0.06525315822960896,
      "grad_norm": 9.333264350891113,
      "learning_rate": 9.903808345020833e-06,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8238018527626991,
      "num_tokens": 25268260.0,
      "step": 21050
    },
    {
      "entropy": 1.8922196328639984,
      "epoch": 0.06528415735465866,
      "grad_norm": 10.881926536560059,
      "learning_rate": 9.901456622634717e-06,
      "loss": 0.6427,
      "mean_token_accuracy": 0.8115344852209091,
      "num_tokens": 25279383.0,
      "step": 21060
    },
    {
      "entropy": 1.8002661630511283,
      "epoch": 0.06531515647970836,
      "grad_norm": 11.18062686920166,
      "learning_rate": 9.899106574747767e-06,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8290015637874604,
      "num_tokens": 25292228.0,
      "step": 21070
    },
    {
      "entropy": 1.8496287435293197,
      "epoch": 0.06534615560475805,
      "grad_norm": 8.944875717163086,
      "learning_rate": 9.896758199373761e-06,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8110730588436127,
      "num_tokens": 25304999.0,
      "step": 21080
    },
    {
      "entropy": 1.887920793890953,
      "epoch": 0.06537715472980775,
      "grad_norm": 5.810867786407471,
      "learning_rate": 9.894411494529786e-06,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8183253973722457,
      "num_tokens": 25316919.0,
      "step": 21090
    },
    {
      "entropy": 1.7966985195875167,
      "epoch": 0.06540815385485745,
      "grad_norm": 10.770297050476074,
      "learning_rate": 9.892066458236207e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8241160228848458,
      "num_tokens": 25329663.0,
      "step": 21100
    },
    {
      "entropy": 1.936796745657921,
      "epoch": 0.06543915297990714,
      "grad_norm": 10.986133575439453,
      "learning_rate": 9.88972308851668e-06,
      "loss": 0.6237,
      "mean_token_accuracy": 0.7997463896870614,
      "num_tokens": 25341721.0,
      "step": 21110
    },
    {
      "entropy": 1.917391985654831,
      "epoch": 0.06547015210495684,
      "grad_norm": 11.787429809570312,
      "learning_rate": 9.887381383398138e-06,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8246586143970489,
      "num_tokens": 25353957.0,
      "step": 21120
    },
    {
      "entropy": 1.8626626536250115,
      "epoch": 0.06550115123000654,
      "grad_norm": 10.935876846313477,
      "learning_rate": 9.885041340910771e-06,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8301349982619286,
      "num_tokens": 25366548.0,
      "step": 21130
    },
    {
      "entropy": 1.9012475624680518,
      "epoch": 0.06553215035505623,
      "grad_norm": 11.089349746704102,
      "learning_rate": 9.882702959088042e-06,
      "loss": 0.6256,
      "mean_token_accuracy": 0.8115785628557205,
      "num_tokens": 25378165.0,
      "step": 21140
    },
    {
      "entropy": 1.9432383313775063,
      "epoch": 0.06556314948010593,
      "grad_norm": 6.057243824005127,
      "learning_rate": 9.880366235966667e-06,
      "loss": 0.6253,
      "mean_token_accuracy": 0.8162486657500267,
      "num_tokens": 25389601.0,
      "step": 21150
    },
    {
      "entropy": 1.8184719279408454,
      "epoch": 0.06559414860515562,
      "grad_norm": 9.748690605163574,
      "learning_rate": 9.878031169586607e-06,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8212730750441551,
      "num_tokens": 25402464.0,
      "step": 21160
    },
    {
      "entropy": 1.8200145453214644,
      "epoch": 0.06562514773020532,
      "grad_norm": 10.591629981994629,
      "learning_rate": 9.875697757991068e-06,
      "loss": 0.5807,
      "mean_token_accuracy": 0.8190927669405937,
      "num_tokens": 25415261.0,
      "step": 21170
    },
    {
      "entropy": 1.911136743426323,
      "epoch": 0.065656146855255,
      "grad_norm": 9.886494636535645,
      "learning_rate": 9.87336599922648e-06,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8133341312408447,
      "num_tokens": 25426211.0,
      "step": 21180
    },
    {
      "entropy": 1.859015080332756,
      "epoch": 0.0656871459803047,
      "grad_norm": 11.79246711730957,
      "learning_rate": 9.871035891342516e-06,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8148143604397774,
      "num_tokens": 25438384.0,
      "step": 21190
    },
    {
      "entropy": 1.845600600540638,
      "epoch": 0.0657181451053544,
      "grad_norm": 11.987141609191895,
      "learning_rate": 9.868707432392061e-06,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8216294556856155,
      "num_tokens": 25451164.0,
      "step": 21200
    },
    {
      "entropy": 1.970012903213501,
      "epoch": 0.0657491442304041,
      "grad_norm": 10.808934211730957,
      "learning_rate": 9.866380620431211e-06,
      "loss": 0.674,
      "mean_token_accuracy": 0.8027332067489624,
      "num_tokens": 25461708.0,
      "step": 21210
    },
    {
      "entropy": 1.8186683028936386,
      "epoch": 0.06578014335545379,
      "grad_norm": 5.0801897048950195,
      "learning_rate": 9.86405545351927e-06,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8230162873864174,
      "num_tokens": 25475658.0,
      "step": 21220
    },
    {
      "entropy": 1.9227222502231598,
      "epoch": 0.06581114248050349,
      "grad_norm": 10.78120231628418,
      "learning_rate": 9.861731929718746e-06,
      "loss": 0.6277,
      "mean_token_accuracy": 0.816400445997715,
      "num_tokens": 25487133.0,
      "step": 21230
    },
    {
      "entropy": 1.921273510158062,
      "epoch": 0.06584214160555318,
      "grad_norm": 11.297560691833496,
      "learning_rate": 9.859410047095337e-06,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8111256033182144,
      "num_tokens": 25498732.0,
      "step": 21240
    },
    {
      "entropy": 1.8860565900802613,
      "epoch": 0.06587314073060288,
      "grad_norm": 12.347557067871094,
      "learning_rate": 9.857089803717928e-06,
      "loss": 0.5763,
      "mean_token_accuracy": 0.814181549847126,
      "num_tokens": 25510149.0,
      "step": 21250
    },
    {
      "entropy": 1.8615178689360619,
      "epoch": 0.06590413985565258,
      "grad_norm": 13.131210327148438,
      "learning_rate": 9.854771197658584e-06,
      "loss": 0.6138,
      "mean_token_accuracy": 0.8099151000380516,
      "num_tokens": 25522335.0,
      "step": 21260
    },
    {
      "entropy": 1.9241080656647682,
      "epoch": 0.06593513898070227,
      "grad_norm": 4.803411960601807,
      "learning_rate": 9.852454226992548e-06,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8204170763492584,
      "num_tokens": 25533747.0,
      "step": 21270
    },
    {
      "entropy": 1.889383627474308,
      "epoch": 0.06596613810575197,
      "grad_norm": 10.4091796875,
      "learning_rate": 9.850138889798216e-06,
      "loss": 0.6602,
      "mean_token_accuracy": 0.8120569944381714,
      "num_tokens": 25545456.0,
      "step": 21280
    },
    {
      "entropy": 1.9471757873892783,
      "epoch": 0.06599713723080167,
      "grad_norm": 9.847548484802246,
      "learning_rate": 9.847825184157157e-06,
      "loss": 0.6242,
      "mean_token_accuracy": 0.808134414255619,
      "num_tokens": 25556403.0,
      "step": 21290
    },
    {
      "entropy": 1.8040209040045738,
      "epoch": 0.06602813635585135,
      "grad_norm": 9.907560348510742,
      "learning_rate": 9.845513108154088e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8381869062781334,
      "num_tokens": 25569237.0,
      "step": 21300
    },
    {
      "entropy": 1.8439020454883575,
      "epoch": 0.06605913548090105,
      "grad_norm": 10.755191802978516,
      "learning_rate": 9.843202659876867e-06,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8121224045753479,
      "num_tokens": 25581127.0,
      "step": 21310
    },
    {
      "entropy": 1.9237766891717911,
      "epoch": 0.06609013460595074,
      "grad_norm": 9.386770248413086,
      "learning_rate": 9.8408938374165e-06,
      "loss": 0.6702,
      "mean_token_accuracy": 0.8007935658097267,
      "num_tokens": 25592005.0,
      "step": 21320
    },
    {
      "entropy": 1.9236200451850891,
      "epoch": 0.06612113373100044,
      "grad_norm": 11.110610961914062,
      "learning_rate": 9.838586638867122e-06,
      "loss": 0.6142,
      "mean_token_accuracy": 0.8141868889331818,
      "num_tokens": 25603901.0,
      "step": 21330
    },
    {
      "entropy": 1.8876248642802238,
      "epoch": 0.06615213285605014,
      "grad_norm": 11.1986665725708,
      "learning_rate": 9.836281062325994e-06,
      "loss": 0.5891,
      "mean_token_accuracy": 0.8205958276987075,
      "num_tokens": 25615461.0,
      "step": 21340
    },
    {
      "entropy": 1.8958319693803787,
      "epoch": 0.06618313198109983,
      "grad_norm": 10.968427658081055,
      "learning_rate": 9.833977105893499e-06,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8049457132816314,
      "num_tokens": 25627052.0,
      "step": 21350
    },
    {
      "entropy": 1.7655820041894912,
      "epoch": 0.06621413110614953,
      "grad_norm": 11.009693145751953,
      "learning_rate": 9.831674767673128e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8243117600679397,
      "num_tokens": 25640796.0,
      "step": 21360
    },
    {
      "entropy": 1.9570647537708283,
      "epoch": 0.06624513023119923,
      "grad_norm": 4.816678524017334,
      "learning_rate": 9.829374045771485e-06,
      "loss": 0.6455,
      "mean_token_accuracy": 0.8167396351695061,
      "num_tokens": 25652130.0,
      "step": 21370
    },
    {
      "entropy": 1.8768548294901848,
      "epoch": 0.06627612935624892,
      "grad_norm": 9.173911094665527,
      "learning_rate": 9.827074938298272e-06,
      "loss": 0.563,
      "mean_token_accuracy": 0.8292611509561538,
      "num_tokens": 25664530.0,
      "step": 21380
    },
    {
      "entropy": 1.836962741613388,
      "epoch": 0.06630712848129862,
      "grad_norm": 11.056595802307129,
      "learning_rate": 9.824777443366282e-06,
      "loss": 0.5817,
      "mean_token_accuracy": 0.8175490334630012,
      "num_tokens": 25676479.0,
      "step": 21390
    },
    {
      "entropy": 1.862608587741852,
      "epoch": 0.06633812760634832,
      "grad_norm": 10.60038948059082,
      "learning_rate": 9.822481559091401e-06,
      "loss": 0.6626,
      "mean_token_accuracy": 0.8113431319594383,
      "num_tokens": 25687935.0,
      "step": 21400
    },
    {
      "entropy": 1.8528019905090332,
      "epoch": 0.06636912673139801,
      "grad_norm": 13.311474800109863,
      "learning_rate": 9.820187283592584e-06,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8175876498222351,
      "num_tokens": 25700014.0,
      "step": 21410
    },
    {
      "entropy": 1.903902080655098,
      "epoch": 0.0664001258564477,
      "grad_norm": 9.098358154296875,
      "learning_rate": 9.817894614991875e-06,
      "loss": 0.6484,
      "mean_token_accuracy": 0.8050385147333146,
      "num_tokens": 25711005.0,
      "step": 21420
    },
    {
      "entropy": 1.8798560991883277,
      "epoch": 0.06643112498149739,
      "grad_norm": 10.936999320983887,
      "learning_rate": 9.815603551414374e-06,
      "loss": 0.5749,
      "mean_token_accuracy": 0.822943688929081,
      "num_tokens": 25722335.0,
      "step": 21430
    },
    {
      "entropy": 1.7517326176166534,
      "epoch": 0.06646212410654709,
      "grad_norm": 10.7169189453125,
      "learning_rate": 9.813314090988247e-06,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8164523139595985,
      "num_tokens": 25736104.0,
      "step": 21440
    },
    {
      "entropy": 1.9384674742817878,
      "epoch": 0.06649312323159678,
      "grad_norm": 10.496194839477539,
      "learning_rate": 9.811026231844714e-06,
      "loss": 0.6222,
      "mean_token_accuracy": 0.8058070495724678,
      "num_tokens": 25747103.0,
      "step": 21450
    },
    {
      "entropy": 1.9147820815443992,
      "epoch": 0.06652412235664648,
      "grad_norm": 11.332496643066406,
      "learning_rate": 9.808739972118045e-06,
      "loss": 0.6591,
      "mean_token_accuracy": 0.7969366267323494,
      "num_tokens": 25759932.0,
      "step": 21460
    },
    {
      "entropy": 1.9171302869915963,
      "epoch": 0.06655512148169618,
      "grad_norm": 10.07258129119873,
      "learning_rate": 9.806455309945553e-06,
      "loss": 0.596,
      "mean_token_accuracy": 0.8198638454079628,
      "num_tokens": 25771193.0,
      "step": 21470
    },
    {
      "entropy": 1.663824899494648,
      "epoch": 0.06658612060674587,
      "grad_norm": 4.567380428314209,
      "learning_rate": 9.804172243467576e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8362728267908096,
      "num_tokens": 25785955.0,
      "step": 21480
    },
    {
      "entropy": 1.819168321788311,
      "epoch": 0.06661711973179557,
      "grad_norm": 6.046937465667725,
      "learning_rate": 9.8018907708275e-06,
      "loss": 0.6468,
      "mean_token_accuracy": 0.8072507992386818,
      "num_tokens": 25798694.0,
      "step": 21490
    },
    {
      "entropy": 1.8685800537467003,
      "epoch": 0.06664811885684527,
      "grad_norm": 9.6455659866333,
      "learning_rate": 9.799610890171714e-06,
      "loss": 0.5722,
      "mean_token_accuracy": 0.8169274970889091,
      "num_tokens": 25810540.0,
      "step": 21500
    },
    {
      "entropy": 1.894263032078743,
      "epoch": 0.06667911798189496,
      "grad_norm": 11.558295249938965,
      "learning_rate": 9.797332599649637e-06,
      "loss": 0.636,
      "mean_token_accuracy": 0.8053765878081321,
      "num_tokens": 25822467.0,
      "step": 21510
    },
    {
      "entropy": 1.9177772477269173,
      "epoch": 0.06671011710694466,
      "grad_norm": 10.580260276794434,
      "learning_rate": 9.795055897413697e-06,
      "loss": 0.684,
      "mean_token_accuracy": 0.800214584171772,
      "num_tokens": 25834010.0,
      "step": 21520
    },
    {
      "entropy": 1.831970725953579,
      "epoch": 0.06674111623199436,
      "grad_norm": 10.084224700927734,
      "learning_rate": 9.792780781619318e-06,
      "loss": 0.5683,
      "mean_token_accuracy": 0.821027934551239,
      "num_tokens": 25846769.0,
      "step": 21530
    },
    {
      "entropy": 1.9271451473236083,
      "epoch": 0.06677211535704405,
      "grad_norm": 11.508028030395508,
      "learning_rate": 9.790507250424926e-06,
      "loss": 0.643,
      "mean_token_accuracy": 0.7898375526070595,
      "num_tokens": 25857730.0,
      "step": 21540
    },
    {
      "entropy": 1.9292464286088944,
      "epoch": 0.06680311448209374,
      "grad_norm": 9.729171752929688,
      "learning_rate": 9.788235301991947e-06,
      "loss": 0.6296,
      "mean_token_accuracy": 0.8045027181506157,
      "num_tokens": 25869739.0,
      "step": 21550
    },
    {
      "entropy": 1.9347792953252791,
      "epoch": 0.06683411360714343,
      "grad_norm": 9.912041664123535,
      "learning_rate": 9.785964934484776e-06,
      "loss": 0.7101,
      "mean_token_accuracy": 0.7909634590148926,
      "num_tokens": 25880786.0,
      "step": 21560
    },
    {
      "entropy": 1.8601661220192909,
      "epoch": 0.06686511273219313,
      "grad_norm": 10.346519470214844,
      "learning_rate": 9.783696146070801e-06,
      "loss": 0.5925,
      "mean_token_accuracy": 0.8147589430212975,
      "num_tokens": 25892319.0,
      "step": 21570
    },
    {
      "entropy": 1.7954956993460656,
      "epoch": 0.06689611185724283,
      "grad_norm": 4.684342384338379,
      "learning_rate": 9.781428934920377e-06,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8224958389997482,
      "num_tokens": 25905020.0,
      "step": 21580
    },
    {
      "entropy": 1.8782492965459823,
      "epoch": 0.06692711098229252,
      "grad_norm": 12.385150909423828,
      "learning_rate": 9.77916329920682e-06,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8114234715700149,
      "num_tokens": 25916570.0,
      "step": 21590
    },
    {
      "entropy": 1.8336504146456718,
      "epoch": 0.06695811010734222,
      "grad_norm": 3.51712965965271,
      "learning_rate": 9.776899237106418e-06,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8270602032542229,
      "num_tokens": 25929984.0,
      "step": 21600
    },
    {
      "entropy": 1.8505462616682054,
      "epoch": 0.06698910923239192,
      "grad_norm": 9.539434432983398,
      "learning_rate": 9.774636746798405e-06,
      "loss": 0.635,
      "mean_token_accuracy": 0.8051048040390014,
      "num_tokens": 25941877.0,
      "step": 21610
    },
    {
      "entropy": 1.7775013580918313,
      "epoch": 0.06702010835744161,
      "grad_norm": 11.277313232421875,
      "learning_rate": 9.77237582646497e-06,
      "loss": 0.5501,
      "mean_token_accuracy": 0.823443454504013,
      "num_tokens": 25955271.0,
      "step": 21620
    },
    {
      "entropy": 1.8188173368573188,
      "epoch": 0.06705110748249131,
      "grad_norm": 5.36696720123291,
      "learning_rate": 9.770116474291232e-06,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8193252950906753,
      "num_tokens": 25967242.0,
      "step": 21630
    },
    {
      "entropy": 1.819683888554573,
      "epoch": 0.067082106607541,
      "grad_norm": 10.795523643493652,
      "learning_rate": 9.767858688465254e-06,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8111403912305832,
      "num_tokens": 25979320.0,
      "step": 21640
    },
    {
      "entropy": 1.7906116724014283,
      "epoch": 0.0671131057325907,
      "grad_norm": 9.500042915344238,
      "learning_rate": 9.765602467178033e-06,
      "loss": 0.5362,
      "mean_token_accuracy": 0.822031632065773,
      "num_tokens": 25992253.0,
      "step": 21650
    },
    {
      "entropy": 1.8245372220873832,
      "epoch": 0.0671441048576404,
      "grad_norm": 8.064600944519043,
      "learning_rate": 9.763347808623481e-06,
      "loss": 0.5656,
      "mean_token_accuracy": 0.817147271335125,
      "num_tokens": 26003957.0,
      "step": 21660
    },
    {
      "entropy": 1.8516656443476678,
      "epoch": 0.06717510398269008,
      "grad_norm": 10.404606819152832,
      "learning_rate": 9.761094710998432e-06,
      "loss": 0.5836,
      "mean_token_accuracy": 0.8104123935103417,
      "num_tokens": 26015881.0,
      "step": 21670
    },
    {
      "entropy": 1.827203567326069,
      "epoch": 0.06720610310773978,
      "grad_norm": 13.826399803161621,
      "learning_rate": 9.75884317250263e-06,
      "loss": 0.5807,
      "mean_token_accuracy": 0.8180100679397583,
      "num_tokens": 26028572.0,
      "step": 21680
    },
    {
      "entropy": 1.7154892578721046,
      "epoch": 0.06723710223278948,
      "grad_norm": 4.944169044494629,
      "learning_rate": 9.756593191338725e-06,
      "loss": 0.3884,
      "mean_token_accuracy": 0.848066033422947,
      "num_tokens": 26043181.0,
      "step": 21690
    },
    {
      "entropy": 1.8305002465844153,
      "epoch": 0.06726810135783917,
      "grad_norm": 11.294840812683105,
      "learning_rate": 9.754344765712266e-06,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8094679519534111,
      "num_tokens": 26055431.0,
      "step": 21700
    },
    {
      "entropy": 1.8822925835847855,
      "epoch": 0.06729910048288887,
      "grad_norm": 11.073433876037598,
      "learning_rate": 9.752097893831698e-06,
      "loss": 0.6668,
      "mean_token_accuracy": 0.8047523692250251,
      "num_tokens": 26067040.0,
      "step": 21710
    },
    {
      "entropy": 1.8371146380901338,
      "epoch": 0.06733009960793856,
      "grad_norm": 9.113365173339844,
      "learning_rate": 9.749852573908346e-06,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8240691289305687,
      "num_tokens": 26080151.0,
      "step": 21720
    },
    {
      "entropy": 1.8499283462762832,
      "epoch": 0.06736109873298826,
      "grad_norm": 11.711502075195312,
      "learning_rate": 9.747608804156427e-06,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8179016202688217,
      "num_tokens": 26092394.0,
      "step": 21730
    },
    {
      "entropy": 1.7742941290140153,
      "epoch": 0.06739209785803796,
      "grad_norm": 10.995613098144531,
      "learning_rate": 9.745366582793027e-06,
      "loss": 0.5917,
      "mean_token_accuracy": 0.8143564939498902,
      "num_tokens": 26105790.0,
      "step": 21740
    },
    {
      "entropy": 1.727923959493637,
      "epoch": 0.06742309698308765,
      "grad_norm": 3.2797532081604004,
      "learning_rate": 9.7431259080381e-06,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8418777331709861,
      "num_tokens": 26119446.0,
      "step": 21750
    },
    {
      "entropy": 1.8147297531366349,
      "epoch": 0.06745409610813735,
      "grad_norm": 8.432580947875977,
      "learning_rate": 9.740886778114467e-06,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8339284613728524,
      "num_tokens": 26131348.0,
      "step": 21760
    },
    {
      "entropy": 1.8820328325033189,
      "epoch": 0.06748509523318705,
      "grad_norm": 5.698887825012207,
      "learning_rate": 9.738649191247806e-06,
      "loss": 0.6172,
      "mean_token_accuracy": 0.809808611869812,
      "num_tokens": 26143041.0,
      "step": 21770
    },
    {
      "entropy": 1.8815367594361305,
      "epoch": 0.06751609435823674,
      "grad_norm": 8.804140090942383,
      "learning_rate": 9.736413145666649e-06,
      "loss": 0.594,
      "mean_token_accuracy": 0.820245711505413,
      "num_tokens": 26154978.0,
      "step": 21780
    },
    {
      "entropy": 1.9202604204416276,
      "epoch": 0.06754709348328644,
      "grad_norm": 9.845060348510742,
      "learning_rate": 9.734178639602368e-06,
      "loss": 0.6266,
      "mean_token_accuracy": 0.8070004045963287,
      "num_tokens": 26166500.0,
      "step": 21790
    },
    {
      "entropy": 1.8778617054224014,
      "epoch": 0.06757809260833612,
      "grad_norm": 9.954373359680176,
      "learning_rate": 9.731945671289185e-06,
      "loss": 0.6406,
      "mean_token_accuracy": 0.8081085130572319,
      "num_tokens": 26178330.0,
      "step": 21800
    },
    {
      "entropy": 1.9139564037322998,
      "epoch": 0.06760909173338582,
      "grad_norm": 10.78117847442627,
      "learning_rate": 9.72971423896414e-06,
      "loss": 0.6322,
      "mean_token_accuracy": 0.8143807515501976,
      "num_tokens": 26189448.0,
      "step": 21810
    },
    {
      "entropy": 1.8785874828696252,
      "epoch": 0.06764009085843552,
      "grad_norm": 5.0894389152526855,
      "learning_rate": 9.727484340867116e-06,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8152860775589943,
      "num_tokens": 26201634.0,
      "step": 21820
    },
    {
      "entropy": 1.8347378268837928,
      "epoch": 0.06767108998348521,
      "grad_norm": 4.857226371765137,
      "learning_rate": 9.725255975240813e-06,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8315959751605988,
      "num_tokens": 26213558.0,
      "step": 21830
    },
    {
      "entropy": 1.845048761367798,
      "epoch": 0.06770208910853491,
      "grad_norm": 11.329482078552246,
      "learning_rate": 9.723029140330748e-06,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8252388536930084,
      "num_tokens": 26225961.0,
      "step": 21840
    },
    {
      "entropy": 1.8393208265304566,
      "epoch": 0.0677330882335846,
      "grad_norm": 10.937715530395508,
      "learning_rate": 9.72080383438525e-06,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8155218675732613,
      "num_tokens": 26238570.0,
      "step": 21850
    },
    {
      "entropy": 1.835605874657631,
      "epoch": 0.0677640873586343,
      "grad_norm": 9.71265983581543,
      "learning_rate": 9.718580055655447e-06,
      "loss": 0.5487,
      "mean_token_accuracy": 0.8284683138132095,
      "num_tokens": 26249804.0,
      "step": 21860
    },
    {
      "entropy": 1.9305734172463418,
      "epoch": 0.067795086483684,
      "grad_norm": 9.470818519592285,
      "learning_rate": 9.716357802395276e-06,
      "loss": 0.6421,
      "mean_token_accuracy": 0.8088799849152565,
      "num_tokens": 26261058.0,
      "step": 21870
    },
    {
      "entropy": 1.8729921713471414,
      "epoch": 0.0678260856087337,
      "grad_norm": 11.346515655517578,
      "learning_rate": 9.714137072861461e-06,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8200590804219245,
      "num_tokens": 26273172.0,
      "step": 21880
    },
    {
      "entropy": 1.8179200991988183,
      "epoch": 0.0678570847337834,
      "grad_norm": 9.96755313873291,
      "learning_rate": 9.711917865313517e-06,
      "loss": 0.5983,
      "mean_token_accuracy": 0.813332186639309,
      "num_tokens": 26286190.0,
      "step": 21890
    },
    {
      "entropy": 1.807481935620308,
      "epoch": 0.06788808385883309,
      "grad_norm": 9.412008285522461,
      "learning_rate": 9.709700178013736e-06,
      "loss": 0.5088,
      "mean_token_accuracy": 0.823253333568573,
      "num_tokens": 26299950.0,
      "step": 21900
    },
    {
      "entropy": 1.8429307714104652,
      "epoch": 0.06791908298388279,
      "grad_norm": 10.125129699707031,
      "learning_rate": 9.707484009227192e-06,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8161771893501282,
      "num_tokens": 26312676.0,
      "step": 21910
    },
    {
      "entropy": 1.932323306798935,
      "epoch": 0.06795008210893247,
      "grad_norm": 10.303938865661621,
      "learning_rate": 9.705269357221728e-06,
      "loss": 0.6293,
      "mean_token_accuracy": 0.8122246876358986,
      "num_tokens": 26324476.0,
      "step": 21920
    },
    {
      "entropy": 1.8644244894385338,
      "epoch": 0.06798108123398217,
      "grad_norm": 9.893197059631348,
      "learning_rate": 9.703056220267948e-06,
      "loss": 0.5923,
      "mean_token_accuracy": 0.8110254645347595,
      "num_tokens": 26336986.0,
      "step": 21930
    },
    {
      "entropy": 1.9435497313737868,
      "epoch": 0.06801208035903186,
      "grad_norm": 10.371121406555176,
      "learning_rate": 9.700844596639224e-06,
      "loss": 0.6676,
      "mean_token_accuracy": 0.8060028344392777,
      "num_tokens": 26348032.0,
      "step": 21940
    },
    {
      "entropy": 1.818027876317501,
      "epoch": 0.06804307948408156,
      "grad_norm": 10.99317741394043,
      "learning_rate": 9.698634484611671e-06,
      "loss": 0.4973,
      "mean_token_accuracy": 0.8351524174213409,
      "num_tokens": 26360903.0,
      "step": 21950
    },
    {
      "entropy": 2.044380483031273,
      "epoch": 0.06807407860913126,
      "grad_norm": 11.019384384155273,
      "learning_rate": 9.696425882464162e-06,
      "loss": 0.7252,
      "mean_token_accuracy": 0.7958701580762864,
      "num_tokens": 26371825.0,
      "step": 21960
    },
    {
      "entropy": 1.9928947612643242,
      "epoch": 0.06810507773418095,
      "grad_norm": 9.34325122833252,
      "learning_rate": 9.694218788478302e-06,
      "loss": 0.6251,
      "mean_token_accuracy": 0.8070891216397286,
      "num_tokens": 26382919.0,
      "step": 21970
    },
    {
      "entropy": 1.8768664389848708,
      "epoch": 0.06813607685923065,
      "grad_norm": 9.70384407043457,
      "learning_rate": 9.692013200938443e-06,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8396727994084359,
      "num_tokens": 26395459.0,
      "step": 21980
    },
    {
      "entropy": 1.8494936734437943,
      "epoch": 0.06816707598428035,
      "grad_norm": 11.369257926940918,
      "learning_rate": 9.689809118131661e-06,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8196944579482078,
      "num_tokens": 26407374.0,
      "step": 21990
    },
    {
      "entropy": 1.8978412017226218,
      "epoch": 0.06819807510933004,
      "grad_norm": 11.591387748718262,
      "learning_rate": 9.68760653834776e-06,
      "loss": 0.6132,
      "mean_token_accuracy": 0.8112616434693336,
      "num_tokens": 26419627.0,
      "step": 22000
    },
    {
      "entropy": 1.8775578573346139,
      "epoch": 0.06822907423437974,
      "grad_norm": 5.471284866333008,
      "learning_rate": 9.685405459879265e-06,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8241677179932594,
      "num_tokens": 26431836.0,
      "step": 22010
    },
    {
      "entropy": 1.9307914853096009,
      "epoch": 0.06826007335942944,
      "grad_norm": 9.451096534729004,
      "learning_rate": 9.683205881021414e-06,
      "loss": 0.6088,
      "mean_token_accuracy": 0.8177047148346901,
      "num_tokens": 26442888.0,
      "step": 22020
    },
    {
      "entropy": 1.915931698679924,
      "epoch": 0.06829107248447913,
      "grad_norm": 10.326228141784668,
      "learning_rate": 9.681007800072153e-06,
      "loss": 0.6463,
      "mean_token_accuracy": 0.8002750858664512,
      "num_tokens": 26454901.0,
      "step": 22030
    },
    {
      "entropy": 1.9367796674370765,
      "epoch": 0.06832207160952881,
      "grad_norm": 10.724263191223145,
      "learning_rate": 9.678811215332136e-06,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8125025644898415,
      "num_tokens": 26466644.0,
      "step": 22040
    },
    {
      "entropy": 1.8923877745866775,
      "epoch": 0.06835307073457851,
      "grad_norm": 9.895198822021484,
      "learning_rate": 9.676616125104707e-06,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8232117027044297,
      "num_tokens": 26478470.0,
      "step": 22050
    },
    {
      "entropy": 1.9004954114556312,
      "epoch": 0.06838406985962821,
      "grad_norm": 9.4879150390625,
      "learning_rate": 9.674422527695905e-06,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8110221117734909,
      "num_tokens": 26490224.0,
      "step": 22060
    },
    {
      "entropy": 1.9001964166760446,
      "epoch": 0.0684150689846779,
      "grad_norm": 5.616433620452881,
      "learning_rate": 9.672230421414466e-06,
      "loss": 0.625,
      "mean_token_accuracy": 0.8154096305370331,
      "num_tokens": 26501366.0,
      "step": 22070
    },
    {
      "entropy": 1.9321230724453926,
      "epoch": 0.0684460681097276,
      "grad_norm": 9.737174987792969,
      "learning_rate": 9.670039804571791e-06,
      "loss": 0.6395,
      "mean_token_accuracy": 0.8030791565775871,
      "num_tokens": 26512519.0,
      "step": 22080
    },
    {
      "entropy": 1.9279805243015289,
      "epoch": 0.0684770672347773,
      "grad_norm": 9.636087417602539,
      "learning_rate": 9.667850675481966e-06,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8268870130181313,
      "num_tokens": 26524253.0,
      "step": 22090
    },
    {
      "entropy": 1.8550891995429992,
      "epoch": 0.068508066359827,
      "grad_norm": 4.487429141998291,
      "learning_rate": 9.665663032461747e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8270553275942802,
      "num_tokens": 26536731.0,
      "step": 22100
    },
    {
      "entropy": 1.9424635365605354,
      "epoch": 0.06853906548487669,
      "grad_norm": 10.442728996276855,
      "learning_rate": 9.663476873830555e-06,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8206922665238381,
      "num_tokens": 26547996.0,
      "step": 22110
    },
    {
      "entropy": 1.9067403554916382,
      "epoch": 0.06857006460992639,
      "grad_norm": 9.885056495666504,
      "learning_rate": 9.661292197910468e-06,
      "loss": 0.626,
      "mean_token_accuracy": 0.8120665445923805,
      "num_tokens": 26560444.0,
      "step": 22120
    },
    {
      "entropy": 1.9578855365514756,
      "epoch": 0.06860106373497608,
      "grad_norm": 8.84913158416748,
      "learning_rate": 9.659109003026222e-06,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8090391054749488,
      "num_tokens": 26572313.0,
      "step": 22130
    },
    {
      "entropy": 2.025720348954201,
      "epoch": 0.06863206286002578,
      "grad_norm": 9.901925086975098,
      "learning_rate": 9.656927287505196e-06,
      "loss": 0.7,
      "mean_token_accuracy": 0.8067919075489044,
      "num_tokens": 26583097.0,
      "step": 22140
    },
    {
      "entropy": 2.0064867049455644,
      "epoch": 0.06866306198507548,
      "grad_norm": 10.104426383972168,
      "learning_rate": 9.654747049677416e-06,
      "loss": 0.6419,
      "mean_token_accuracy": 0.8101355582475662,
      "num_tokens": 26594726.0,
      "step": 22150
    },
    {
      "entropy": 1.9431262537837029,
      "epoch": 0.06869406111012517,
      "grad_norm": 10.839469909667969,
      "learning_rate": 9.652568287875552e-06,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8269416391849518,
      "num_tokens": 26606536.0,
      "step": 22160
    },
    {
      "entropy": 2.012579309940338,
      "epoch": 0.06872506023517486,
      "grad_norm": 12.621667861938477,
      "learning_rate": 9.650391000434892e-06,
      "loss": 0.6522,
      "mean_token_accuracy": 0.8193601354956627,
      "num_tokens": 26617239.0,
      "step": 22170
    },
    {
      "entropy": 1.8841126412153244,
      "epoch": 0.06875605936022455,
      "grad_norm": 5.069724082946777,
      "learning_rate": 9.648215185693367e-06,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8241742476820946,
      "num_tokens": 26629571.0,
      "step": 22180
    },
    {
      "entropy": 1.938437856733799,
      "epoch": 0.06878705848527425,
      "grad_norm": 10.184552192687988,
      "learning_rate": 9.646040841991519e-06,
      "loss": 0.6364,
      "mean_token_accuracy": 0.812780536711216,
      "num_tokens": 26640715.0,
      "step": 22190
    },
    {
      "entropy": 1.8110738307237626,
      "epoch": 0.06881805761032395,
      "grad_norm": 10.346161842346191,
      "learning_rate": 9.64386796767251e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8277497068047523,
      "num_tokens": 26655037.0,
      "step": 22200
    },
    {
      "entropy": 1.9208206847310065,
      "epoch": 0.06884905673537364,
      "grad_norm": 5.2261762619018555,
      "learning_rate": 9.641696561082117e-06,
      "loss": 0.5982,
      "mean_token_accuracy": 0.8199071541428566,
      "num_tokens": 26667082.0,
      "step": 22210
    },
    {
      "entropy": 1.8494412750005722,
      "epoch": 0.06888005586042334,
      "grad_norm": 10.328657150268555,
      "learning_rate": 9.639526620568718e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.8268266186118126,
      "num_tokens": 26680702.0,
      "step": 22220
    },
    {
      "entropy": 2.001448130607605,
      "epoch": 0.06891105498547304,
      "grad_norm": 10.208023071289062,
      "learning_rate": 9.637358144483292e-06,
      "loss": 0.6457,
      "mean_token_accuracy": 0.8080255687236786,
      "num_tokens": 26692418.0,
      "step": 22230
    },
    {
      "entropy": 1.9539388507604598,
      "epoch": 0.06894205411052273,
      "grad_norm": 5.077010631561279,
      "learning_rate": 9.635191131179423e-06,
      "loss": 0.5923,
      "mean_token_accuracy": 0.8029374584555626,
      "num_tokens": 26705460.0,
      "step": 22240
    },
    {
      "entropy": 2.013536052405834,
      "epoch": 0.06897305323557243,
      "grad_norm": 8.475672721862793,
      "learning_rate": 9.633025579013265e-06,
      "loss": 0.6455,
      "mean_token_accuracy": 0.8068231120705605,
      "num_tokens": 26716216.0,
      "step": 22250
    },
    {
      "entropy": 1.9996852800250053,
      "epoch": 0.06900405236062213,
      "grad_norm": 11.207576751708984,
      "learning_rate": 9.630861486343582e-06,
      "loss": 0.6351,
      "mean_token_accuracy": 0.8085829988121986,
      "num_tokens": 26726768.0,
      "step": 22260
    },
    {
      "entropy": 1.8975500375032426,
      "epoch": 0.06903505148567182,
      "grad_norm": 9.219078063964844,
      "learning_rate": 9.628698851531698e-06,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8229989439249039,
      "num_tokens": 26739895.0,
      "step": 22270
    },
    {
      "entropy": 1.9055659562349319,
      "epoch": 0.06906605061072152,
      "grad_norm": 9.515271186828613,
      "learning_rate": 9.62653767294152e-06,
      "loss": 0.5738,
      "mean_token_accuracy": 0.8302743345499038,
      "num_tokens": 26751846.0,
      "step": 22280
    },
    {
      "entropy": 1.928154693543911,
      "epoch": 0.0690970497357712,
      "grad_norm": 4.5626654624938965,
      "learning_rate": 9.624377948939526e-06,
      "loss": 0.578,
      "mean_token_accuracy": 0.818497559428215,
      "num_tokens": 26764244.0,
      "step": 22290
    },
    {
      "entropy": 1.9365676030516625,
      "epoch": 0.0691280488608209,
      "grad_norm": 10.669965744018555,
      "learning_rate": 9.622219677894753e-06,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8216531693935394,
      "num_tokens": 26777119.0,
      "step": 22300
    },
    {
      "entropy": 1.9759273901581764,
      "epoch": 0.0691590479858706,
      "grad_norm": 9.490134239196777,
      "learning_rate": 9.620062858178804e-06,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8160349875688553,
      "num_tokens": 26789166.0,
      "step": 22310
    },
    {
      "entropy": 1.930388794839382,
      "epoch": 0.06919004711092029,
      "grad_norm": 11.474687576293945,
      "learning_rate": 9.617907488165825e-06,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8070300281047821,
      "num_tokens": 26800730.0,
      "step": 22320
    },
    {
      "entropy": 1.9698352724313737,
      "epoch": 0.06922104623596999,
      "grad_norm": 10.189593315124512,
      "learning_rate": 9.615753566232525e-06,
      "loss": 0.6158,
      "mean_token_accuracy": 0.8105901405215263,
      "num_tokens": 26812034.0,
      "step": 22330
    },
    {
      "entropy": 1.8571903064846993,
      "epoch": 0.06925204536101968,
      "grad_norm": 12.529641151428223,
      "learning_rate": 9.613601090758144e-06,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8161864310503006,
      "num_tokens": 26825051.0,
      "step": 22340
    },
    {
      "entropy": 1.9772468566894532,
      "epoch": 0.06928304448606938,
      "grad_norm": 4.139663219451904,
      "learning_rate": 9.611450060124465e-06,
      "loss": 0.6634,
      "mean_token_accuracy": 0.8000595793128014,
      "num_tokens": 26836544.0,
      "step": 22350
    },
    {
      "entropy": 1.8732447177171707,
      "epoch": 0.06931404361111908,
      "grad_norm": 11.927010536193848,
      "learning_rate": 9.609300472715811e-06,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8216081500053406,
      "num_tokens": 26849615.0,
      "step": 22360
    },
    {
      "entropy": 1.93919685035944,
      "epoch": 0.06934504273616877,
      "grad_norm": 10.041327476501465,
      "learning_rate": 9.607152326919017e-06,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8051743671298027,
      "num_tokens": 26861482.0,
      "step": 22370
    },
    {
      "entropy": 1.970766744017601,
      "epoch": 0.06937604186121847,
      "grad_norm": 10.42013931274414,
      "learning_rate": 9.605005621123464e-06,
      "loss": 0.6534,
      "mean_token_accuracy": 0.8068610474467277,
      "num_tokens": 26873126.0,
      "step": 22380
    },
    {
      "entropy": 1.9374214485287666,
      "epoch": 0.06940704098626817,
      "grad_norm": 9.681081771850586,
      "learning_rate": 9.602860353721028e-06,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8023192837834359,
      "num_tokens": 26885153.0,
      "step": 22390
    },
    {
      "entropy": 1.9398547366261483,
      "epoch": 0.06943804011131786,
      "grad_norm": 9.45241928100586,
      "learning_rate": 9.600716523106113e-06,
      "loss": 0.6011,
      "mean_token_accuracy": 0.82165118008852,
      "num_tokens": 26896187.0,
      "step": 22400
    },
    {
      "entropy": 1.9944514393806458,
      "epoch": 0.06946903923636755,
      "grad_norm": 9.26956844329834,
      "learning_rate": 9.598574127675626e-06,
      "loss": 0.6097,
      "mean_token_accuracy": 0.8134081959724426,
      "num_tokens": 26907111.0,
      "step": 22410
    },
    {
      "entropy": 1.9747971966862679,
      "epoch": 0.06950003836141724,
      "grad_norm": 10.627933502197266,
      "learning_rate": 9.59643316582898e-06,
      "loss": 0.6869,
      "mean_token_accuracy": 0.7921556413173676,
      "num_tokens": 26918762.0,
      "step": 22420
    },
    {
      "entropy": 1.975240734219551,
      "epoch": 0.06953103748646694,
      "grad_norm": 11.592079162597656,
      "learning_rate": 9.594293635968081e-06,
      "loss": 0.6539,
      "mean_token_accuracy": 0.800179885327816,
      "num_tokens": 26930609.0,
      "step": 22430
    },
    {
      "entropy": 1.9276101559400558,
      "epoch": 0.06956203661151664,
      "grad_norm": 10.43637466430664,
      "learning_rate": 9.59215553649733e-06,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8026261925697327,
      "num_tokens": 26942087.0,
      "step": 22440
    },
    {
      "entropy": 1.9058784544467926,
      "epoch": 0.06959303573656633,
      "grad_norm": 10.7579984664917,
      "learning_rate": 9.590018865823617e-06,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8225520506501198,
      "num_tokens": 26954256.0,
      "step": 22450
    },
    {
      "entropy": 1.928253923356533,
      "epoch": 0.06962403486161603,
      "grad_norm": 9.777754783630371,
      "learning_rate": 9.587883622356315e-06,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8160154640674591,
      "num_tokens": 26966403.0,
      "step": 22460
    },
    {
      "entropy": 1.8594375014305116,
      "epoch": 0.06965503398666573,
      "grad_norm": 4.670237064361572,
      "learning_rate": 9.585749804507275e-06,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8113547876477242,
      "num_tokens": 26979420.0,
      "step": 22470
    },
    {
      "entropy": 1.8346465826034546,
      "epoch": 0.06968603311171542,
      "grad_norm": 10.510963439941406,
      "learning_rate": 9.58361741069082e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8267633840441704,
      "num_tokens": 26991602.0,
      "step": 22480
    },
    {
      "entropy": 1.855713202059269,
      "epoch": 0.06971703223676512,
      "grad_norm": 10.206911087036133,
      "learning_rate": 9.581486439323741e-06,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8163058936595917,
      "num_tokens": 27003931.0,
      "step": 22490
    },
    {
      "entropy": 1.9139684349298478,
      "epoch": 0.06974803136181482,
      "grad_norm": 10.377299308776855,
      "learning_rate": 9.579356888825293e-06,
      "loss": 0.622,
      "mean_token_accuracy": 0.8105499014258385,
      "num_tokens": 27016144.0,
      "step": 22500
    },
    {
      "entropy": 1.8956247389316558,
      "epoch": 0.06977903048686451,
      "grad_norm": 11.713562965393066,
      "learning_rate": 9.57722875761719e-06,
      "loss": 0.5887,
      "mean_token_accuracy": 0.8171533569693565,
      "num_tokens": 27028209.0,
      "step": 22510
    },
    {
      "entropy": 1.8996628910303115,
      "epoch": 0.06981002961191421,
      "grad_norm": 10.300981521606445,
      "learning_rate": 9.5751020441236e-06,
      "loss": 0.6273,
      "mean_token_accuracy": 0.816479966044426,
      "num_tokens": 27039827.0,
      "step": 22520
    },
    {
      "entropy": 1.8548725992441177,
      "epoch": 0.0698410287369639,
      "grad_norm": 9.6876220703125,
      "learning_rate": 9.572976746771132e-06,
      "loss": 0.529,
      "mean_token_accuracy": 0.8360526219010354,
      "num_tokens": 27051785.0,
      "step": 22530
    },
    {
      "entropy": 1.8720781534910202,
      "epoch": 0.06987202786201359,
      "grad_norm": 11.248559951782227,
      "learning_rate": 9.570852863988847e-06,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8292858317494393,
      "num_tokens": 27063846.0,
      "step": 22540
    },
    {
      "entropy": 1.928021316230297,
      "epoch": 0.06990302698706329,
      "grad_norm": 12.513522148132324,
      "learning_rate": 9.568730394208245e-06,
      "loss": 0.6212,
      "mean_token_accuracy": 0.8208240970969201,
      "num_tokens": 27075131.0,
      "step": 22550
    },
    {
      "entropy": 1.9080385088920593,
      "epoch": 0.06993402611211298,
      "grad_norm": 13.877976417541504,
      "learning_rate": 9.566609335863253e-06,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8138008996844291,
      "num_tokens": 27088488.0,
      "step": 22560
    },
    {
      "entropy": 1.8622148722410201,
      "epoch": 0.06996502523716268,
      "grad_norm": 9.615792274475098,
      "learning_rate": 9.564489687390232e-06,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8177649453282356,
      "num_tokens": 27101651.0,
      "step": 22570
    },
    {
      "entropy": 1.9193992719054223,
      "epoch": 0.06999602436221238,
      "grad_norm": 10.330147743225098,
      "learning_rate": 9.562371447227966e-06,
      "loss": 0.6293,
      "mean_token_accuracy": 0.8139034882187843,
      "num_tokens": 27112881.0,
      "step": 22580
    },
    {
      "entropy": 1.8690322995185853,
      "epoch": 0.07002702348726207,
      "grad_norm": 9.160511016845703,
      "learning_rate": 9.560254613817653e-06,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8230489745736123,
      "num_tokens": 27124763.0,
      "step": 22590
    },
    {
      "entropy": 1.8515427842736245,
      "epoch": 0.07005802261231177,
      "grad_norm": 9.302170753479004,
      "learning_rate": 9.558139185602919e-06,
      "loss": 0.6005,
      "mean_token_accuracy": 0.8178195133805275,
      "num_tokens": 27136756.0,
      "step": 22600
    },
    {
      "entropy": 1.9153072491288186,
      "epoch": 0.07008902173736146,
      "grad_norm": 4.208265781402588,
      "learning_rate": 9.556025161029786e-06,
      "loss": 0.5976,
      "mean_token_accuracy": 0.816322073340416,
      "num_tokens": 27148060.0,
      "step": 22610
    },
    {
      "entropy": 1.958570632338524,
      "epoch": 0.07012002086241116,
      "grad_norm": 11.623967170715332,
      "learning_rate": 9.553912538546687e-06,
      "loss": 0.663,
      "mean_token_accuracy": 0.8184066906571388,
      "num_tokens": 27158765.0,
      "step": 22620
    },
    {
      "entropy": 1.8749097064137459,
      "epoch": 0.07015101998746086,
      "grad_norm": 10.970122337341309,
      "learning_rate": 9.55180131660445e-06,
      "loss": 0.5946,
      "mean_token_accuracy": 0.81072598695755,
      "num_tokens": 27170806.0,
      "step": 22630
    },
    {
      "entropy": 1.831792925298214,
      "epoch": 0.07018201911251055,
      "grad_norm": 11.485851287841797,
      "learning_rate": 9.54969149365631e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8216899335384369,
      "num_tokens": 27184157.0,
      "step": 22640
    },
    {
      "entropy": 1.9119883999228477,
      "epoch": 0.07021301823756025,
      "grad_norm": 11.170741081237793,
      "learning_rate": 9.547583068157877e-06,
      "loss": 0.6163,
      "mean_token_accuracy": 0.8011090368032455,
      "num_tokens": 27196415.0,
      "step": 22650
    },
    {
      "entropy": 1.8825739532709123,
      "epoch": 0.07024401736260993,
      "grad_norm": 8.26751708984375,
      "learning_rate": 9.54547603856716e-06,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8156488195061684,
      "num_tokens": 27209042.0,
      "step": 22660
    },
    {
      "entropy": 1.9585013464093208,
      "epoch": 0.07027501648765963,
      "grad_norm": 13.17714786529541,
      "learning_rate": 9.54337040334454e-06,
      "loss": 0.66,
      "mean_token_accuracy": 0.8024971842765808,
      "num_tokens": 27219980.0,
      "step": 22670
    },
    {
      "entropy": 1.9092792928218842,
      "epoch": 0.07030601561270933,
      "grad_norm": 8.935144424438477,
      "learning_rate": 9.54126616095278e-06,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8157609686255455,
      "num_tokens": 27231294.0,
      "step": 22680
    },
    {
      "entropy": 1.8577600359916686,
      "epoch": 0.07033701473775902,
      "grad_norm": 11.60800838470459,
      "learning_rate": 9.539163309857014e-06,
      "loss": 0.559,
      "mean_token_accuracy": 0.8194443017244339,
      "num_tokens": 27243917.0,
      "step": 22690
    },
    {
      "entropy": 1.9067883223295212,
      "epoch": 0.07036801386280872,
      "grad_norm": 9.336089134216309,
      "learning_rate": 9.537061848524734e-06,
      "loss": 0.6081,
      "mean_token_accuracy": 0.8156686320900917,
      "num_tokens": 27255668.0,
      "step": 22700
    },
    {
      "entropy": 1.8172156438231468,
      "epoch": 0.07039901298785842,
      "grad_norm": 5.0443644523620605,
      "learning_rate": 9.534961775425811e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8269337728619576,
      "num_tokens": 27269533.0,
      "step": 22710
    },
    {
      "entropy": 1.8444980800151825,
      "epoch": 0.07043001211290811,
      "grad_norm": 9.690264701843262,
      "learning_rate": 9.532863089032457e-06,
      "loss": 0.554,
      "mean_token_accuracy": 0.8164748609066009,
      "num_tokens": 27282775.0,
      "step": 22720
    },
    {
      "entropy": 2.0001364931464196,
      "epoch": 0.07046101123795781,
      "grad_norm": 11.46723747253418,
      "learning_rate": 9.530765787819247e-06,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8077256262302399,
      "num_tokens": 27294314.0,
      "step": 22730
    },
    {
      "entropy": 2.0072741538286207,
      "epoch": 0.0704920103630075,
      "grad_norm": 9.978995323181152,
      "learning_rate": 9.528669870263103e-06,
      "loss": 0.6748,
      "mean_token_accuracy": 0.7956588104367256,
      "num_tokens": 27305909.0,
      "step": 22740
    },
    {
      "entropy": 2.031547136604786,
      "epoch": 0.0705230094880572,
      "grad_norm": 9.277776718139648,
      "learning_rate": 9.526575334843284e-06,
      "loss": 0.6595,
      "mean_token_accuracy": 0.8040603443980217,
      "num_tokens": 27317574.0,
      "step": 22750
    },
    {
      "entropy": 1.887120671570301,
      "epoch": 0.0705540086131069,
      "grad_norm": 10.967029571533203,
      "learning_rate": 9.524482180041396e-06,
      "loss": 0.5722,
      "mean_token_accuracy": 0.809663151204586,
      "num_tokens": 27330161.0,
      "step": 22760
    },
    {
      "entropy": 1.8155884474515915,
      "epoch": 0.0705850077381566,
      "grad_norm": 10.948394775390625,
      "learning_rate": 9.522390404341375e-06,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8259642213582993,
      "num_tokens": 27343602.0,
      "step": 22770
    },
    {
      "entropy": 1.9294879660010338,
      "epoch": 0.07061600686320628,
      "grad_norm": 9.418436050415039,
      "learning_rate": 9.52030000622949e-06,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8182932585477829,
      "num_tokens": 27355422.0,
      "step": 22780
    },
    {
      "entropy": 1.9679240599274634,
      "epoch": 0.07064700598825598,
      "grad_norm": 5.147764682769775,
      "learning_rate": 9.51821098419433e-06,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8287713721394538,
      "num_tokens": 27367573.0,
      "step": 22790
    },
    {
      "entropy": 1.9291693687438964,
      "epoch": 0.07067800511330567,
      "grad_norm": 4.149558067321777,
      "learning_rate": 9.516123336726806e-06,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8070420354604722,
      "num_tokens": 27379593.0,
      "step": 22800
    },
    {
      "entropy": 2.0455012649297712,
      "epoch": 0.07070900423835537,
      "grad_norm": 11.943114280700684,
      "learning_rate": 9.514037062320148e-06,
      "loss": 0.7181,
      "mean_token_accuracy": 0.7937376946210861,
      "num_tokens": 27390417.0,
      "step": 22810
    },
    {
      "entropy": 1.928147019445896,
      "epoch": 0.07074000336340507,
      "grad_norm": 10.236185073852539,
      "learning_rate": 9.511952159469895e-06,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8198776602745056,
      "num_tokens": 27402550.0,
      "step": 22820
    },
    {
      "entropy": 2.0119119971990584,
      "epoch": 0.07077100248845476,
      "grad_norm": 8.378026962280273,
      "learning_rate": 9.509868626673897e-06,
      "loss": 0.6267,
      "mean_token_accuracy": 0.821306724846363,
      "num_tokens": 27413355.0,
      "step": 22830
    },
    {
      "entropy": 1.9993405610322952,
      "epoch": 0.07080200161350446,
      "grad_norm": 9.925045013427734,
      "learning_rate": 9.507786462432295e-06,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8195290580391884,
      "num_tokens": 27424504.0,
      "step": 22840
    },
    {
      "entropy": 1.9374178424477577,
      "epoch": 0.07083300073855416,
      "grad_norm": 4.789980888366699,
      "learning_rate": 9.505705665247544e-06,
      "loss": 0.6441,
      "mean_token_accuracy": 0.8104717791080475,
      "num_tokens": 27436791.0,
      "step": 22850
    },
    {
      "entropy": 1.9383593767881393,
      "epoch": 0.07086399986360385,
      "grad_norm": 10.263440132141113,
      "learning_rate": 9.503626233624376e-06,
      "loss": 0.627,
      "mean_token_accuracy": 0.8090912804007531,
      "num_tokens": 27448858.0,
      "step": 22860
    },
    {
      "entropy": 2.0473968207836153,
      "epoch": 0.07089499898865355,
      "grad_norm": 12.426820755004883,
      "learning_rate": 9.501548166069823e-06,
      "loss": 0.7145,
      "mean_token_accuracy": 0.7896327748894691,
      "num_tokens": 27459568.0,
      "step": 22870
    },
    {
      "entropy": 1.9629662334918976,
      "epoch": 0.07092599811370325,
      "grad_norm": 9.21473503112793,
      "learning_rate": 9.499471461093198e-06,
      "loss": 0.6637,
      "mean_token_accuracy": 0.8080029338598251,
      "num_tokens": 27472413.0,
      "step": 22880
    },
    {
      "entropy": 1.9620086148381233,
      "epoch": 0.07095699723875294,
      "grad_norm": 11.438567161560059,
      "learning_rate": 9.497396117206091e-06,
      "loss": 0.6037,
      "mean_token_accuracy": 0.8183909133076668,
      "num_tokens": 27484336.0,
      "step": 22890
    },
    {
      "entropy": 1.9684356123209,
      "epoch": 0.07098799636380264,
      "grad_norm": 10.009490966796875,
      "learning_rate": 9.49532213292237e-06,
      "loss": 0.6147,
      "mean_token_accuracy": 0.8186342790722847,
      "num_tokens": 27495646.0,
      "step": 22900
    },
    {
      "entropy": 1.8952196419239045,
      "epoch": 0.07101899548885232,
      "grad_norm": 10.987327575683594,
      "learning_rate": 9.493249506758174e-06,
      "loss": 0.5656,
      "mean_token_accuracy": 0.8256588101387023,
      "num_tokens": 27507145.0,
      "step": 22910
    },
    {
      "entropy": 1.9499731242656708,
      "epoch": 0.07104999461390202,
      "grad_norm": 12.571517944335938,
      "learning_rate": 9.491178237231904e-06,
      "loss": 0.6393,
      "mean_token_accuracy": 0.8107420906424523,
      "num_tokens": 27518536.0,
      "step": 22920
    },
    {
      "entropy": 1.9733731657266618,
      "epoch": 0.07108099373895171,
      "grad_norm": 11.42594051361084,
      "learning_rate": 9.48910832286423e-06,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8104237765073776,
      "num_tokens": 27529692.0,
      "step": 22930
    },
    {
      "entropy": 1.9351124539971352,
      "epoch": 0.07111199286400141,
      "grad_norm": 12.896414756774902,
      "learning_rate": 9.48703976217807e-06,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8083524718880654,
      "num_tokens": 27541473.0,
      "step": 22940
    },
    {
      "entropy": 2.000078111886978,
      "epoch": 0.07114299198905111,
      "grad_norm": 10.5455904006958,
      "learning_rate": 9.484972553698609e-06,
      "loss": 0.7026,
      "mean_token_accuracy": 0.7948870256543159,
      "num_tokens": 27552542.0,
      "step": 22950
    },
    {
      "entropy": 1.87677208930254,
      "epoch": 0.0711739911141008,
      "grad_norm": 11.3195161819458,
      "learning_rate": 9.482906695953262e-06,
      "loss": 0.5929,
      "mean_token_accuracy": 0.819133634865284,
      "num_tokens": 27565263.0,
      "step": 22960
    },
    {
      "entropy": 1.8816159687936307,
      "epoch": 0.0712049902391505,
      "grad_norm": 5.181549072265625,
      "learning_rate": 9.480842187471707e-06,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8206304356455802,
      "num_tokens": 27577618.0,
      "step": 22970
    },
    {
      "entropy": 1.8843361094594002,
      "epoch": 0.0712359893642002,
      "grad_norm": 9.86581039428711,
      "learning_rate": 9.478779026785849e-06,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8279527604579926,
      "num_tokens": 27589842.0,
      "step": 22980
    },
    {
      "entropy": 1.9283671468496322,
      "epoch": 0.0712669884892499,
      "grad_norm": 8.319514274597168,
      "learning_rate": 9.476717212429832e-06,
      "loss": 0.6132,
      "mean_token_accuracy": 0.8075576022267341,
      "num_tokens": 27601806.0,
      "step": 22990
    },
    {
      "entropy": 1.9603848546743392,
      "epoch": 0.07129798761429959,
      "grad_norm": 10.907440185546875,
      "learning_rate": 9.474656742940032e-06,
      "loss": 0.6049,
      "mean_token_accuracy": 0.817129735648632,
      "num_tokens": 27613900.0,
      "step": 23000
    },
    {
      "entropy": 1.923836286365986,
      "epoch": 0.07132898673934929,
      "grad_norm": 10.593782424926758,
      "learning_rate": 9.472597616855047e-06,
      "loss": 0.6263,
      "mean_token_accuracy": 0.8129079177975654,
      "num_tokens": 27626469.0,
      "step": 23010
    },
    {
      "entropy": 1.9737991467118263,
      "epoch": 0.07135998586439898,
      "grad_norm": 9.002423286437988,
      "learning_rate": 9.470539832715709e-06,
      "loss": 0.6858,
      "mean_token_accuracy": 0.8034425944089889,
      "num_tokens": 27638941.0,
      "step": 23020
    },
    {
      "entropy": 2.0140700817108153,
      "epoch": 0.07139098498944867,
      "grad_norm": 9.269657135009766,
      "learning_rate": 9.468483389065051e-06,
      "loss": 0.7441,
      "mean_token_accuracy": 0.7920542553067207,
      "num_tokens": 27650333.0,
      "step": 23030
    },
    {
      "entropy": 1.9881491467356682,
      "epoch": 0.07142198411449836,
      "grad_norm": 9.53425407409668,
      "learning_rate": 9.466428284448339e-06,
      "loss": 0.605,
      "mean_token_accuracy": 0.815938925743103,
      "num_tokens": 27661706.0,
      "step": 23040
    },
    {
      "entropy": 1.9040154725313188,
      "epoch": 0.07145298323954806,
      "grad_norm": 11.15352725982666,
      "learning_rate": 9.464374517413028e-06,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8235550940036773,
      "num_tokens": 27674006.0,
      "step": 23050
    },
    {
      "entropy": 1.8507097408175468,
      "epoch": 0.07148398236459776,
      "grad_norm": 4.413679599761963,
      "learning_rate": 9.462322086508796e-06,
      "loss": 0.613,
      "mean_token_accuracy": 0.822106608748436,
      "num_tokens": 27686803.0,
      "step": 23060
    },
    {
      "entropy": 1.8648405969142914,
      "epoch": 0.07151498148964745,
      "grad_norm": 8.988687515258789,
      "learning_rate": 9.460270990287506e-06,
      "loss": 0.623,
      "mean_token_accuracy": 0.8131396234035492,
      "num_tokens": 27699064.0,
      "step": 23070
    },
    {
      "entropy": 1.9689454659819603,
      "epoch": 0.07154598061469715,
      "grad_norm": 12.583850860595703,
      "learning_rate": 9.458221227303229e-06,
      "loss": 0.6461,
      "mean_token_accuracy": 0.8088790848851204,
      "num_tokens": 27710056.0,
      "step": 23080
    },
    {
      "entropy": 1.920731896162033,
      "epoch": 0.07157697973974685,
      "grad_norm": 7.8442254066467285,
      "learning_rate": 9.456172796112224e-06,
      "loss": 0.588,
      "mean_token_accuracy": 0.8172019869089127,
      "num_tokens": 27722078.0,
      "step": 23090
    },
    {
      "entropy": 1.8544889241456985,
      "epoch": 0.07160797886479654,
      "grad_norm": 4.247855186462402,
      "learning_rate": 9.454125695272939e-06,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8311522156000137,
      "num_tokens": 27734016.0,
      "step": 23100
    },
    {
      "entropy": 1.905607244372368,
      "epoch": 0.07163897798984624,
      "grad_norm": 11.669919967651367,
      "learning_rate": 9.452079923346001e-06,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8104119807481766,
      "num_tokens": 27745967.0,
      "step": 23110
    },
    {
      "entropy": 1.8860177457332612,
      "epoch": 0.07166997711489594,
      "grad_norm": 8.93431568145752,
      "learning_rate": 9.45003547889422e-06,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8315723642706871,
      "num_tokens": 27758582.0,
      "step": 23120
    },
    {
      "entropy": 1.9184095725417136,
      "epoch": 0.07170097623994563,
      "grad_norm": 9.845928192138672,
      "learning_rate": 9.447992360482587e-06,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8231678128242492,
      "num_tokens": 27770200.0,
      "step": 23130
    },
    {
      "entropy": 1.9659554988145829,
      "epoch": 0.07173197536499533,
      "grad_norm": 6.174232006072998,
      "learning_rate": 9.445950566678251e-06,
      "loss": 0.6704,
      "mean_token_accuracy": 0.8078634425997734,
      "num_tokens": 27781380.0,
      "step": 23140
    },
    {
      "entropy": 1.8729309558868408,
      "epoch": 0.07176297449004501,
      "grad_norm": 11.597721099853516,
      "learning_rate": 9.443910096050535e-06,
      "loss": 0.584,
      "mean_token_accuracy": 0.8244254574179649,
      "num_tokens": 27793522.0,
      "step": 23150
    },
    {
      "entropy": 1.9304191544651985,
      "epoch": 0.07179397361509471,
      "grad_norm": 4.9579596519470215,
      "learning_rate": 9.441870947170925e-06,
      "loss": 0.6446,
      "mean_token_accuracy": 0.812780550122261,
      "num_tokens": 27805347.0,
      "step": 23160
    },
    {
      "entropy": 1.9905481100082398,
      "epoch": 0.0718249727401444,
      "grad_norm": 8.615478515625,
      "learning_rate": 9.439833118613064e-06,
      "loss": 0.6513,
      "mean_token_accuracy": 0.8182663396000862,
      "num_tokens": 27816623.0,
      "step": 23170
    },
    {
      "entropy": 1.9626749917864799,
      "epoch": 0.0718559718651941,
      "grad_norm": 10.739019393920898,
      "learning_rate": 9.437796608952747e-06,
      "loss": 0.5874,
      "mean_token_accuracy": 0.8244615375995636,
      "num_tokens": 27828120.0,
      "step": 23180
    },
    {
      "entropy": 1.9498226195573807,
      "epoch": 0.0718869709902438,
      "grad_norm": 10.651679039001465,
      "learning_rate": 9.435761416767925e-06,
      "loss": 0.6655,
      "mean_token_accuracy": 0.8038348644971848,
      "num_tokens": 27839640.0,
      "step": 23190
    },
    {
      "entropy": 1.8966133877635003,
      "epoch": 0.0719179701152935,
      "grad_norm": 8.705821990966797,
      "learning_rate": 9.433727540638685e-06,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8108318716287612,
      "num_tokens": 27851113.0,
      "step": 23200
    },
    {
      "entropy": 1.8680199176073073,
      "epoch": 0.07194896924034319,
      "grad_norm": 5.520404815673828,
      "learning_rate": 9.431694979147263e-06,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8188697442412376,
      "num_tokens": 27862805.0,
      "step": 23210
    },
    {
      "entropy": 1.9000079199671744,
      "epoch": 0.07197996836539289,
      "grad_norm": 9.184073448181152,
      "learning_rate": 9.429663730878031e-06,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8288420543074608,
      "num_tokens": 27875156.0,
      "step": 23220
    },
    {
      "entropy": 1.8455702617764473,
      "epoch": 0.07201096749044258,
      "grad_norm": 8.574012756347656,
      "learning_rate": 9.427633794417493e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8355481848120689,
      "num_tokens": 27888621.0,
      "step": 23230
    },
    {
      "entropy": 1.9175620287656785,
      "epoch": 0.07204196661549228,
      "grad_norm": 9.760550498962402,
      "learning_rate": 9.425605168354284e-06,
      "loss": 0.6302,
      "mean_token_accuracy": 0.8198355168104172,
      "num_tokens": 27899563.0,
      "step": 23240
    },
    {
      "entropy": 1.8596885740756988,
      "epoch": 0.07207296574054198,
      "grad_norm": 5.949073791503906,
      "learning_rate": 9.423577851279158e-06,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8252129018306732,
      "num_tokens": 27913195.0,
      "step": 23250
    },
    {
      "entropy": 1.9171173721551895,
      "epoch": 0.07210396486559167,
      "grad_norm": 9.470051765441895,
      "learning_rate": 9.421551841784998e-06,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8131398126482964,
      "num_tokens": 27925430.0,
      "step": 23260
    },
    {
      "entropy": 1.9633556708693505,
      "epoch": 0.07213496399064137,
      "grad_norm": 8.595085144042969,
      "learning_rate": 9.4195271384668e-06,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8224080309271813,
      "num_tokens": 27936940.0,
      "step": 23270
    },
    {
      "entropy": 1.937751042842865,
      "epoch": 0.07216596311569105,
      "grad_norm": 5.916355609893799,
      "learning_rate": 9.417503739921671e-06,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8086767017841339,
      "num_tokens": 27948799.0,
      "step": 23280
    },
    {
      "entropy": 1.9640158087015152,
      "epoch": 0.07219696224074075,
      "grad_norm": 10.401774406433105,
      "learning_rate": 9.415481644748828e-06,
      "loss": 0.6664,
      "mean_token_accuracy": 0.8089988082647324,
      "num_tokens": 27959298.0,
      "step": 23290
    },
    {
      "entropy": 1.8880821034312247,
      "epoch": 0.07222796136579045,
      "grad_norm": 10.773529052734375,
      "learning_rate": 9.413460851549596e-06,
      "loss": 0.6015,
      "mean_token_accuracy": 0.808536796271801,
      "num_tokens": 27971177.0,
      "step": 23300
    },
    {
      "entropy": 1.9368318811058998,
      "epoch": 0.07225896049084014,
      "grad_norm": 8.663378715515137,
      "learning_rate": 9.411441358927392e-06,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8213057667016983,
      "num_tokens": 27983758.0,
      "step": 23310
    },
    {
      "entropy": 1.9660808324813843,
      "epoch": 0.07228995961588984,
      "grad_norm": 9.40843677520752,
      "learning_rate": 9.40942316548774e-06,
      "loss": 0.6292,
      "mean_token_accuracy": 0.8099093735218048,
      "num_tokens": 27994875.0,
      "step": 23320
    },
    {
      "entropy": 1.9370207205414771,
      "epoch": 0.07232095874093954,
      "grad_norm": 8.460911750793457,
      "learning_rate": 9.407406269838248e-06,
      "loss": 0.616,
      "mean_token_accuracy": 0.8174383148550988,
      "num_tokens": 28005892.0,
      "step": 23330
    },
    {
      "entropy": 1.9526633992791176,
      "epoch": 0.07235195786598923,
      "grad_norm": 9.294373512268066,
      "learning_rate": 9.405390670588613e-06,
      "loss": 0.6369,
      "mean_token_accuracy": 0.7995718091726303,
      "num_tokens": 28017932.0,
      "step": 23340
    },
    {
      "entropy": 1.9452801957726478,
      "epoch": 0.07238295699103893,
      "grad_norm": 9.387127876281738,
      "learning_rate": 9.403376366350623e-06,
      "loss": 0.6482,
      "mean_token_accuracy": 0.8082485362887383,
      "num_tokens": 28029626.0,
      "step": 23350
    },
    {
      "entropy": 1.8676959410309792,
      "epoch": 0.07241395611608863,
      "grad_norm": 9.851313591003418,
      "learning_rate": 9.401363355738139e-06,
      "loss": 0.5982,
      "mean_token_accuracy": 0.8089207410812378,
      "num_tokens": 28041602.0,
      "step": 23360
    },
    {
      "entropy": 1.8555365659296512,
      "epoch": 0.07244495524113832,
      "grad_norm": 9.005014419555664,
      "learning_rate": 9.399351637367101e-06,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8123884066939354,
      "num_tokens": 28054923.0,
      "step": 23370
    },
    {
      "entropy": 1.8651632323861123,
      "epoch": 0.07247595436618802,
      "grad_norm": 4.899734020233154,
      "learning_rate": 9.397341209855522e-06,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8213669985532761,
      "num_tokens": 28067132.0,
      "step": 23380
    },
    {
      "entropy": 1.9289941102266313,
      "epoch": 0.07250695349123772,
      "grad_norm": 5.1334228515625,
      "learning_rate": 9.395332071823485e-06,
      "loss": 0.6144,
      "mean_token_accuracy": 0.8181710079312324,
      "num_tokens": 28078505.0,
      "step": 23390
    },
    {
      "entropy": 1.8720559775829315,
      "epoch": 0.0725379526162874,
      "grad_norm": 10.077095031738281,
      "learning_rate": 9.393324221893131e-06,
      "loss": 0.5644,
      "mean_token_accuracy": 0.818347430229187,
      "num_tokens": 28090333.0,
      "step": 23400
    },
    {
      "entropy": 1.853566548228264,
      "epoch": 0.0725689517413371,
      "grad_norm": 4.725801467895508,
      "learning_rate": 9.391317658688664e-06,
      "loss": 0.5701,
      "mean_token_accuracy": 0.8122278556227684,
      "num_tokens": 28103819.0,
      "step": 23410
    },
    {
      "entropy": 1.9520633488893508,
      "epoch": 0.07259995086638679,
      "grad_norm": 9.427369117736816,
      "learning_rate": 9.389312380836351e-06,
      "loss": 0.6409,
      "mean_token_accuracy": 0.8200425282120705,
      "num_tokens": 28114738.0,
      "step": 23420
    },
    {
      "entropy": 1.8956316590309144,
      "epoch": 0.07263094999143649,
      "grad_norm": 7.797948360443115,
      "learning_rate": 9.3873083869645e-06,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8178161531686783,
      "num_tokens": 28127067.0,
      "step": 23430
    },
    {
      "entropy": 1.9396205574274064,
      "epoch": 0.07266194911648619,
      "grad_norm": 10.093338012695312,
      "learning_rate": 9.38530567570348e-06,
      "loss": 0.6226,
      "mean_token_accuracy": 0.8116014674305916,
      "num_tokens": 28138551.0,
      "step": 23440
    },
    {
      "entropy": 2.0014042764902116,
      "epoch": 0.07269294824153588,
      "grad_norm": 8.427999496459961,
      "learning_rate": 9.383304245685689e-06,
      "loss": 0.6412,
      "mean_token_accuracy": 0.8133781552314758,
      "num_tokens": 28149878.0,
      "step": 23450
    },
    {
      "entropy": 1.9623918294906617,
      "epoch": 0.07272394736658558,
      "grad_norm": 8.854625701904297,
      "learning_rate": 9.381304095545583e-06,
      "loss": 0.6584,
      "mean_token_accuracy": 0.8096672505140304,
      "num_tokens": 28160988.0,
      "step": 23460
    },
    {
      "entropy": 1.9225846633315087,
      "epoch": 0.07275494649163528,
      "grad_norm": 12.70853042602539,
      "learning_rate": 9.379305223919642e-06,
      "loss": 0.6568,
      "mean_token_accuracy": 0.8003104001283645,
      "num_tokens": 28172035.0,
      "step": 23470
    },
    {
      "entropy": 1.8282870531082154,
      "epoch": 0.07278594561668497,
      "grad_norm": 8.768390655517578,
      "learning_rate": 9.377307629446383e-06,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8221458449959755,
      "num_tokens": 28184093.0,
      "step": 23480
    },
    {
      "entropy": 1.8909123882651329,
      "epoch": 0.07281694474173467,
      "grad_norm": 20.700326919555664,
      "learning_rate": 9.375311310766353e-06,
      "loss": 0.5838,
      "mean_token_accuracy": 0.8204225182533265,
      "num_tokens": 28195961.0,
      "step": 23490
    },
    {
      "entropy": 1.8430864825844764,
      "epoch": 0.07284794386678436,
      "grad_norm": 5.11720609664917,
      "learning_rate": 9.373316266522123e-06,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8319470316171647,
      "num_tokens": 28208526.0,
      "step": 23500
    },
    {
      "entropy": 1.8916525810956955,
      "epoch": 0.07287894299183406,
      "grad_norm": 11.949532508850098,
      "learning_rate": 9.371322495358281e-06,
      "loss": 0.6055,
      "mean_token_accuracy": 0.8044298276305198,
      "num_tokens": 28220641.0,
      "step": 23510
    },
    {
      "entropy": 1.979685339331627,
      "epoch": 0.07290994211688374,
      "grad_norm": 10.18620777130127,
      "learning_rate": 9.369329995921444e-06,
      "loss": 0.6482,
      "mean_token_accuracy": 0.8135093718767166,
      "num_tokens": 28231519.0,
      "step": 23520
    },
    {
      "entropy": 1.9537073016166686,
      "epoch": 0.07294094124193344,
      "grad_norm": 9.836045265197754,
      "learning_rate": 9.36733876686023e-06,
      "loss": 0.6507,
      "mean_token_accuracy": 0.8180084466934204,
      "num_tokens": 28242672.0,
      "step": 23530
    },
    {
      "entropy": 1.9465522095561028,
      "epoch": 0.07297194036698314,
      "grad_norm": 4.765397071838379,
      "learning_rate": 9.365348806825274e-06,
      "loss": 0.6548,
      "mean_token_accuracy": 0.8062368303537368,
      "num_tokens": 28254108.0,
      "step": 23540
    },
    {
      "entropy": 1.9817388027906417,
      "epoch": 0.07300293949203283,
      "grad_norm": 9.245866775512695,
      "learning_rate": 9.36336011446921e-06,
      "loss": 0.6389,
      "mean_token_accuracy": 0.8076906755566597,
      "num_tokens": 28265018.0,
      "step": 23550
    },
    {
      "entropy": 1.8746733576059342,
      "epoch": 0.07303393861708253,
      "grad_norm": 10.122679710388184,
      "learning_rate": 9.36137268844668e-06,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8259296610951423,
      "num_tokens": 28277414.0,
      "step": 23560
    },
    {
      "entropy": 1.9215095147490502,
      "epoch": 0.07306493774213223,
      "grad_norm": 9.02968692779541,
      "learning_rate": 9.359386527414325e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8263942644000053,
      "num_tokens": 28289205.0,
      "step": 23570
    },
    {
      "entropy": 1.8805431053042412,
      "epoch": 0.07309593686718192,
      "grad_norm": 8.153913497924805,
      "learning_rate": 9.35740163003077e-06,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8194609686732293,
      "num_tokens": 28300813.0,
      "step": 23580
    },
    {
      "entropy": 1.8591163650155067,
      "epoch": 0.07312693599223162,
      "grad_norm": 11.29405403137207,
      "learning_rate": 9.355417994956647e-06,
      "loss": 0.6421,
      "mean_token_accuracy": 0.7990970879793167,
      "num_tokens": 28313650.0,
      "step": 23590
    },
    {
      "entropy": 1.887549029290676,
      "epoch": 0.07315793511728132,
      "grad_norm": 4.2539472579956055,
      "learning_rate": 9.353435620854559e-06,
      "loss": 0.539,
      "mean_token_accuracy": 0.8247372597455979,
      "num_tokens": 28325816.0,
      "step": 23600
    },
    {
      "entropy": 1.966456551849842,
      "epoch": 0.07318893424233101,
      "grad_norm": 9.739566802978516,
      "learning_rate": 9.3514545063891e-06,
      "loss": 0.6504,
      "mean_token_accuracy": 0.8081991970539093,
      "num_tokens": 28337088.0,
      "step": 23610
    },
    {
      "entropy": 1.9840751081705092,
      "epoch": 0.07321993336738071,
      "grad_norm": 11.179567337036133,
      "learning_rate": 9.349474650226844e-06,
      "loss": 0.6291,
      "mean_token_accuracy": 0.8135482758283615,
      "num_tokens": 28347375.0,
      "step": 23620
    },
    {
      "entropy": 1.9408112317323685,
      "epoch": 0.0732509324924304,
      "grad_norm": 9.25151538848877,
      "learning_rate": 9.347496051036333e-06,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8124508634209633,
      "num_tokens": 28358563.0,
      "step": 23630
    },
    {
      "entropy": 1.947693009674549,
      "epoch": 0.0732819316174801,
      "grad_norm": 9.259462356567383,
      "learning_rate": 9.345518707488087e-06,
      "loss": 0.6084,
      "mean_token_accuracy": 0.8231876268982887,
      "num_tokens": 28370399.0,
      "step": 23640
    },
    {
      "entropy": 1.846913254261017,
      "epoch": 0.07331293074252979,
      "grad_norm": 8.472787857055664,
      "learning_rate": 9.343542618254596e-06,
      "loss": 0.5402,
      "mean_token_accuracy": 0.822886273264885,
      "num_tokens": 28382814.0,
      "step": 23650
    },
    {
      "entropy": 1.848525333404541,
      "epoch": 0.07334392986757948,
      "grad_norm": 11.601837158203125,
      "learning_rate": 9.341567782010304e-06,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8181129440665245,
      "num_tokens": 28395944.0,
      "step": 23660
    },
    {
      "entropy": 1.778299406170845,
      "epoch": 0.07337492899262918,
      "grad_norm": 10.347676277160645,
      "learning_rate": 9.339594197431625e-06,
      "loss": 0.548,
      "mean_token_accuracy": 0.8195615544915199,
      "num_tokens": 28409215.0,
      "step": 23670
    },
    {
      "entropy": 1.8473999500274658,
      "epoch": 0.07340592811767888,
      "grad_norm": 11.07874584197998,
      "learning_rate": 9.337621863196925e-06,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8226366356015206,
      "num_tokens": 28421381.0,
      "step": 23680
    },
    {
      "entropy": 1.8470186293125153,
      "epoch": 0.07343692724272857,
      "grad_norm": 11.499205589294434,
      "learning_rate": 9.335650777986526e-06,
      "loss": 0.5876,
      "mean_token_accuracy": 0.824992710351944,
      "num_tokens": 28433681.0,
      "step": 23690
    },
    {
      "entropy": 1.9232969626784324,
      "epoch": 0.07346792636777827,
      "grad_norm": 11.156820297241211,
      "learning_rate": 9.333680940482696e-06,
      "loss": 0.5953,
      "mean_token_accuracy": 0.8178783282637596,
      "num_tokens": 28445840.0,
      "step": 23700
    },
    {
      "entropy": 1.9458463191986084,
      "epoch": 0.07349892549282797,
      "grad_norm": 5.818225860595703,
      "learning_rate": 9.331712349369646e-06,
      "loss": 0.6437,
      "mean_token_accuracy": 0.8080273911356926,
      "num_tokens": 28457345.0,
      "step": 23710
    },
    {
      "entropy": 1.9871521532535552,
      "epoch": 0.07352992461787766,
      "grad_norm": 9.459403991699219,
      "learning_rate": 9.329745003333538e-06,
      "loss": 0.6474,
      "mean_token_accuracy": 0.8077614203095436,
      "num_tokens": 28468578.0,
      "step": 23720
    },
    {
      "entropy": 1.9817648902535439,
      "epoch": 0.07356092374292736,
      "grad_norm": 8.348322868347168,
      "learning_rate": 9.327778901062466e-06,
      "loss": 0.6367,
      "mean_token_accuracy": 0.8168795883655549,
      "num_tokens": 28479788.0,
      "step": 23730
    },
    {
      "entropy": 1.9540872916579246,
      "epoch": 0.07359192286797706,
      "grad_norm": 8.266214370727539,
      "learning_rate": 9.325814041246455e-06,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8153672978281975,
      "num_tokens": 28491967.0,
      "step": 23740
    },
    {
      "entropy": 1.960209146142006,
      "epoch": 0.07362292199302675,
      "grad_norm": 11.05349349975586,
      "learning_rate": 9.323850422577469e-06,
      "loss": 0.6019,
      "mean_token_accuracy": 0.811015397310257,
      "num_tokens": 28503930.0,
      "step": 23750
    },
    {
      "entropy": 1.9480343893170358,
      "epoch": 0.07365392111807645,
      "grad_norm": 10.689523696899414,
      "learning_rate": 9.321888043749389e-06,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8198656231164932,
      "num_tokens": 28515451.0,
      "step": 23760
    },
    {
      "entropy": 1.994300290942192,
      "epoch": 0.07368492024312613,
      "grad_norm": 8.471319198608398,
      "learning_rate": 9.319926903458033e-06,
      "loss": 0.631,
      "mean_token_accuracy": 0.8167888432741165,
      "num_tokens": 28526353.0,
      "step": 23770
    },
    {
      "entropy": 1.9629247948527335,
      "epoch": 0.07371591936817583,
      "grad_norm": 10.409026145935059,
      "learning_rate": 9.317967000401127e-06,
      "loss": 0.6331,
      "mean_token_accuracy": 0.8050680428743362,
      "num_tokens": 28537484.0,
      "step": 23780
    },
    {
      "entropy": 1.8861139222979546,
      "epoch": 0.07374691849322552,
      "grad_norm": 8.942246437072754,
      "learning_rate": 9.316008333278319e-06,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8210712686181069,
      "num_tokens": 28550244.0,
      "step": 23790
    },
    {
      "entropy": 1.867449736595154,
      "epoch": 0.07377791761827522,
      "grad_norm": 12.16384506225586,
      "learning_rate": 9.314050900791163e-06,
      "loss": 0.5982,
      "mean_token_accuracy": 0.8131821945309639,
      "num_tokens": 28562527.0,
      "step": 23800
    },
    {
      "entropy": 1.8669994980096818,
      "epoch": 0.07380891674332492,
      "grad_norm": 9.032928466796875,
      "learning_rate": 9.312094701643134e-06,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8275477036833763,
      "num_tokens": 28575138.0,
      "step": 23810
    },
    {
      "entropy": 2.010842078924179,
      "epoch": 0.07383991586837461,
      "grad_norm": 9.27695369720459,
      "learning_rate": 9.3101397345396e-06,
      "loss": 0.635,
      "mean_token_accuracy": 0.8104411855340004,
      "num_tokens": 28586985.0,
      "step": 23820
    },
    {
      "entropy": 1.9500906214118003,
      "epoch": 0.07387091499342431,
      "grad_norm": 3.9518938064575195,
      "learning_rate": 9.30818599818784e-06,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8293194532394409,
      "num_tokens": 28598926.0,
      "step": 23830
    },
    {
      "entropy": 1.9337919741868972,
      "epoch": 0.07390191411847401,
      "grad_norm": 10.783734321594238,
      "learning_rate": 9.306233491297024e-06,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8271838411688804,
      "num_tokens": 28610265.0,
      "step": 23840
    },
    {
      "entropy": 1.8781434014439582,
      "epoch": 0.0739329132435237,
      "grad_norm": 4.472099781036377,
      "learning_rate": 9.30428221257822e-06,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8306781485676765,
      "num_tokens": 28622824.0,
      "step": 23850
    },
    {
      "entropy": 1.9807551354169846,
      "epoch": 0.0739639123685734,
      "grad_norm": 11.769026756286621,
      "learning_rate": 9.302332160744387e-06,
      "loss": 0.6261,
      "mean_token_accuracy": 0.8115956127643585,
      "num_tokens": 28633746.0,
      "step": 23860
    },
    {
      "entropy": 1.9240091010928153,
      "epoch": 0.0739949114936231,
      "grad_norm": 11.441720962524414,
      "learning_rate": 9.300383334510372e-06,
      "loss": 0.6157,
      "mean_token_accuracy": 0.8130579099059105,
      "num_tokens": 28646025.0,
      "step": 23870
    },
    {
      "entropy": 1.9416207402944565,
      "epoch": 0.0740259106186728,
      "grad_norm": 9.635127067565918,
      "learning_rate": 9.298435732592904e-06,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8167257070541382,
      "num_tokens": 28658461.0,
      "step": 23880
    },
    {
      "entropy": 1.9926012963056565,
      "epoch": 0.07405690974372248,
      "grad_norm": 9.58889102935791,
      "learning_rate": 9.296489353710593e-06,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8202344119548798,
      "num_tokens": 28669851.0,
      "step": 23890
    },
    {
      "entropy": 1.917009449005127,
      "epoch": 0.07408790886877217,
      "grad_norm": 12.779633522033691,
      "learning_rate": 9.294544196583929e-06,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8271616026759148,
      "num_tokens": 28682219.0,
      "step": 23900
    },
    {
      "entropy": 1.9119545385241508,
      "epoch": 0.07411890799382187,
      "grad_norm": 4.876058578491211,
      "learning_rate": 9.29260025993527e-06,
      "loss": 0.5712,
      "mean_token_accuracy": 0.8157782420516014,
      "num_tokens": 28694730.0,
      "step": 23910
    },
    {
      "entropy": 1.982969084382057,
      "epoch": 0.07414990711887157,
      "grad_norm": 4.22367525100708,
      "learning_rate": 9.290657542488846e-06,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8123695835471153,
      "num_tokens": 28706446.0,
      "step": 23920
    },
    {
      "entropy": 1.965246671438217,
      "epoch": 0.07418090624392126,
      "grad_norm": 11.446246147155762,
      "learning_rate": 9.28871604297075e-06,
      "loss": 0.6086,
      "mean_token_accuracy": 0.8144690200686455,
      "num_tokens": 28718411.0,
      "step": 23930
    },
    {
      "entropy": 1.9847482353448869,
      "epoch": 0.07421190536897096,
      "grad_norm": 11.718853950500488,
      "learning_rate": 9.28677576010895e-06,
      "loss": 0.6281,
      "mean_token_accuracy": 0.8073780536651611,
      "num_tokens": 28730646.0,
      "step": 23940
    },
    {
      "entropy": 1.9997831106185913,
      "epoch": 0.07424290449402066,
      "grad_norm": 9.053925514221191,
      "learning_rate": 9.284836692633257e-06,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8186374083161354,
      "num_tokens": 28741168.0,
      "step": 23950
    },
    {
      "entropy": 1.8356414943933488,
      "epoch": 0.07427390361907035,
      "grad_norm": 9.540495872497559,
      "learning_rate": 9.282898839275347e-06,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8257902413606644,
      "num_tokens": 28755165.0,
      "step": 23960
    },
    {
      "entropy": 1.9523360833525658,
      "epoch": 0.07430490274412005,
      "grad_norm": 11.081901550292969,
      "learning_rate": 9.280962198768745e-06,
      "loss": 0.6224,
      "mean_token_accuracy": 0.8144389301538467,
      "num_tokens": 28766577.0,
      "step": 23970
    },
    {
      "entropy": 1.962296548485756,
      "epoch": 0.07433590186916975,
      "grad_norm": 9.727673530578613,
      "learning_rate": 9.279026769848825e-06,
      "loss": 0.6022,
      "mean_token_accuracy": 0.8120512589812279,
      "num_tokens": 28777780.0,
      "step": 23980
    },
    {
      "entropy": 1.9339822575449943,
      "epoch": 0.07436690099421944,
      "grad_norm": 12.94129467010498,
      "learning_rate": 9.277092551252813e-06,
      "loss": 0.6079,
      "mean_token_accuracy": 0.8044057339429855,
      "num_tokens": 28789644.0,
      "step": 23990
    },
    {
      "entropy": 1.9828555971384048,
      "epoch": 0.07439790011926914,
      "grad_norm": 11.137704849243164,
      "learning_rate": 9.275159541719766e-06,
      "loss": 0.6334,
      "mean_token_accuracy": 0.8016148954629898,
      "num_tokens": 28800391.0,
      "step": 24000
    },
    {
      "entropy": 1.8854010567069053,
      "epoch": 0.07442889924431884,
      "grad_norm": 9.98965835571289,
      "learning_rate": 9.273227739990585e-06,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8219901576638222,
      "num_tokens": 28812899.0,
      "step": 24010
    },
    {
      "entropy": 1.9441875860095024,
      "epoch": 0.07445989836936852,
      "grad_norm": 9.923999786376953,
      "learning_rate": 9.271297144808003e-06,
      "loss": 0.637,
      "mean_token_accuracy": 0.8106958836317062,
      "num_tokens": 28824358.0,
      "step": 24020
    },
    {
      "entropy": 1.889090073108673,
      "epoch": 0.07449089749441822,
      "grad_norm": 8.807568550109863,
      "learning_rate": 9.269367754916594e-06,
      "loss": 0.5404,
      "mean_token_accuracy": 0.83165952116251,
      "num_tokens": 28835906.0,
      "step": 24030
    },
    {
      "entropy": 1.8731929019093514,
      "epoch": 0.07452189661946791,
      "grad_norm": 11.27502727508545,
      "learning_rate": 9.267439569062747e-06,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8130838423967361,
      "num_tokens": 28848120.0,
      "step": 24040
    },
    {
      "entropy": 1.9439046129584312,
      "epoch": 0.07455289574451761,
      "grad_norm": 10.14338207244873,
      "learning_rate": 9.26551258599468e-06,
      "loss": 0.5926,
      "mean_token_accuracy": 0.817124992609024,
      "num_tokens": 28860425.0,
      "step": 24050
    },
    {
      "entropy": 1.8763557612895965,
      "epoch": 0.0745838948695673,
      "grad_norm": 9.435752868652344,
      "learning_rate": 9.26358680446244e-06,
      "loss": 0.5757,
      "mean_token_accuracy": 0.82074084430933,
      "num_tokens": 28873156.0,
      "step": 24060
    },
    {
      "entropy": 1.8747549593448638,
      "epoch": 0.074614893994617,
      "grad_norm": 12.689223289489746,
      "learning_rate": 9.26166222321788e-06,
      "loss": 0.5661,
      "mean_token_accuracy": 0.817321227490902,
      "num_tokens": 28885693.0,
      "step": 24070
    },
    {
      "entropy": 1.7927760377526283,
      "epoch": 0.0746458931196667,
      "grad_norm": 8.860281944274902,
      "learning_rate": 9.25973884101468e-06,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8320321753621102,
      "num_tokens": 28898315.0,
      "step": 24080
    },
    {
      "entropy": 1.8106766402721406,
      "epoch": 0.0746768922447164,
      "grad_norm": 11.25741958618164,
      "learning_rate": 9.257816656608314e-06,
      "loss": 0.554,
      "mean_token_accuracy": 0.8215843066573143,
      "num_tokens": 28912120.0,
      "step": 24090
    },
    {
      "entropy": 1.951988846063614,
      "epoch": 0.07470789136976609,
      "grad_norm": 9.571369171142578,
      "learning_rate": 9.25589566875608e-06,
      "loss": 0.6411,
      "mean_token_accuracy": 0.8124495342373848,
      "num_tokens": 28924042.0,
      "step": 24100
    },
    {
      "entropy": 1.850806337594986,
      "epoch": 0.07473889049481579,
      "grad_norm": 9.427949905395508,
      "learning_rate": 9.253975876217073e-06,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8303135067224503,
      "num_tokens": 28936552.0,
      "step": 24110
    },
    {
      "entropy": 1.9311017155647279,
      "epoch": 0.07476988961986548,
      "grad_norm": 5.728914260864258,
      "learning_rate": 9.25205727775219e-06,
      "loss": 0.6528,
      "mean_token_accuracy": 0.8051609516143798,
      "num_tokens": 28948713.0,
      "step": 24120
    },
    {
      "entropy": 1.9310711935162543,
      "epoch": 0.07480088874491518,
      "grad_norm": 11.202091217041016,
      "learning_rate": 9.250139872124125e-06,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8168716624379158,
      "num_tokens": 28959606.0,
      "step": 24130
    },
    {
      "entropy": 1.8889696165919303,
      "epoch": 0.07483188786996486,
      "grad_norm": 3.5714454650878906,
      "learning_rate": 9.248223658097366e-06,
      "loss": 0.547,
      "mean_token_accuracy": 0.8295445218682289,
      "num_tokens": 28971784.0,
      "step": 24140
    },
    {
      "entropy": 1.9066083252429962,
      "epoch": 0.07486288699501456,
      "grad_norm": 15.336698532104492,
      "learning_rate": 9.246308634438193e-06,
      "loss": 0.6225,
      "mean_token_accuracy": 0.814862783253193,
      "num_tokens": 28983340.0,
      "step": 24150
    },
    {
      "entropy": 1.9325315952301025,
      "epoch": 0.07489388612006426,
      "grad_norm": 10.859710693359375,
      "learning_rate": 9.244394799914674e-06,
      "loss": 0.6306,
      "mean_token_accuracy": 0.8129459515213966,
      "num_tokens": 28995466.0,
      "step": 24160
    },
    {
      "entropy": 1.9299397438764572,
      "epoch": 0.07492488524511395,
      "grad_norm": 11.286776542663574,
      "learning_rate": 9.242482153296657e-06,
      "loss": 0.6152,
      "mean_token_accuracy": 0.8135070979595185,
      "num_tokens": 29007025.0,
      "step": 24170
    },
    {
      "entropy": 1.9263689830899238,
      "epoch": 0.07495588437016365,
      "grad_norm": 8.885149002075195,
      "learning_rate": 9.240570693355777e-06,
      "loss": 0.6203,
      "mean_token_accuracy": 0.8149414286017418,
      "num_tokens": 29019269.0,
      "step": 24180
    },
    {
      "entropy": 1.9538878142833709,
      "epoch": 0.07498688349521335,
      "grad_norm": 12.976593971252441,
      "learning_rate": 9.238660418865444e-06,
      "loss": 0.6581,
      "mean_token_accuracy": 0.8068120896816253,
      "num_tokens": 29030437.0,
      "step": 24190
    },
    {
      "entropy": 1.9112654134631157,
      "epoch": 0.07501788262026304,
      "grad_norm": 9.866148948669434,
      "learning_rate": 9.236751328600838e-06,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8209733456373215,
      "num_tokens": 29042237.0,
      "step": 24200
    },
    {
      "entropy": 2.0098152339458464,
      "epoch": 0.07504888174531274,
      "grad_norm": 9.860359191894531,
      "learning_rate": 9.234843421338919e-06,
      "loss": 0.691,
      "mean_token_accuracy": 0.7986131757497787,
      "num_tokens": 29053227.0,
      "step": 24210
    },
    {
      "entropy": 1.914099135249853,
      "epoch": 0.07507988087036244,
      "grad_norm": 3.684596300125122,
      "learning_rate": 9.232936695858406e-06,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8275950834155082,
      "num_tokens": 29066785.0,
      "step": 24220
    },
    {
      "entropy": 1.8749552696943284,
      "epoch": 0.07511087999541213,
      "grad_norm": 9.360295295715332,
      "learning_rate": 9.231031150939787e-06,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8239286541938782,
      "num_tokens": 29080125.0,
      "step": 24230
    },
    {
      "entropy": 1.9456191077828406,
      "epoch": 0.07514187912046183,
      "grad_norm": 9.981027603149414,
      "learning_rate": 9.229126785365307e-06,
      "loss": 0.6293,
      "mean_token_accuracy": 0.8097566932439804,
      "num_tokens": 29091466.0,
      "step": 24240
    },
    {
      "entropy": 1.9858911633491516,
      "epoch": 0.07517287824551153,
      "grad_norm": 10.95433521270752,
      "learning_rate": 9.227223597918977e-06,
      "loss": 0.6356,
      "mean_token_accuracy": 0.8043408066034317,
      "num_tokens": 29102658.0,
      "step": 24250
    },
    {
      "entropy": 1.9025256112217903,
      "epoch": 0.07520387737056121,
      "grad_norm": 4.23534631729126,
      "learning_rate": 9.225321587386555e-06,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8235425055027008,
      "num_tokens": 29115098.0,
      "step": 24260
    },
    {
      "entropy": 1.9474715083837508,
      "epoch": 0.0752348764956109,
      "grad_norm": 8.94451904296875,
      "learning_rate": 9.22342075255555e-06,
      "loss": 0.5524,
      "mean_token_accuracy": 0.82459936439991,
      "num_tokens": 29127073.0,
      "step": 24270
    },
    {
      "entropy": 2.0239467322826385,
      "epoch": 0.0752658756206606,
      "grad_norm": 9.617466926574707,
      "learning_rate": 9.221521092215226e-06,
      "loss": 0.6738,
      "mean_token_accuracy": 0.8019678488373756,
      "num_tokens": 29137792.0,
      "step": 24280
    },
    {
      "entropy": 1.9810176610946655,
      "epoch": 0.0752968747457103,
      "grad_norm": 9.926680564880371,
      "learning_rate": 9.219622605156583e-06,
      "loss": 0.6936,
      "mean_token_accuracy": 0.8082947731018066,
      "num_tokens": 29148654.0,
      "step": 24290
    },
    {
      "entropy": 1.9755667805671693,
      "epoch": 0.07532787387076,
      "grad_norm": 9.290923118591309,
      "learning_rate": 9.217725290172373e-06,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8143254667520523,
      "num_tokens": 29159751.0,
      "step": 24300
    },
    {
      "entropy": 1.9506172388792038,
      "epoch": 0.07535887299580969,
      "grad_norm": 4.7398362159729,
      "learning_rate": 9.215829146057074e-06,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8142923697829246,
      "num_tokens": 29171860.0,
      "step": 24310
    },
    {
      "entropy": 1.964981135725975,
      "epoch": 0.07538987212085939,
      "grad_norm": 10.409062385559082,
      "learning_rate": 9.213934171606911e-06,
      "loss": 0.585,
      "mean_token_accuracy": 0.8187711656093597,
      "num_tokens": 29183756.0,
      "step": 24320
    },
    {
      "entropy": 1.8626610353589057,
      "epoch": 0.07542087124590909,
      "grad_norm": 16.499589920043945,
      "learning_rate": 9.212040365619834e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8303863450884819,
      "num_tokens": 29196691.0,
      "step": 24330
    },
    {
      "entropy": 1.9400402843952178,
      "epoch": 0.07545187037095878,
      "grad_norm": 10.061908721923828,
      "learning_rate": 9.210147726895522e-06,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8127806216478348,
      "num_tokens": 29208279.0,
      "step": 24340
    },
    {
      "entropy": 1.9271167308092116,
      "epoch": 0.07548286949600848,
      "grad_norm": 4.295543193817139,
      "learning_rate": 9.208256254235383e-06,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8224424123764038,
      "num_tokens": 29219803.0,
      "step": 24350
    },
    {
      "entropy": 1.9891034051775933,
      "epoch": 0.07551386862105817,
      "grad_norm": 12.840479850769043,
      "learning_rate": 9.206365946442545e-06,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8187379658222198,
      "num_tokens": 29230787.0,
      "step": 24360
    },
    {
      "entropy": 1.9009444043040276,
      "epoch": 0.07554486774610787,
      "grad_norm": 10.091814041137695,
      "learning_rate": 9.204476802321853e-06,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8159004956483841,
      "num_tokens": 29243031.0,
      "step": 24370
    },
    {
      "entropy": 1.9343391075730323,
      "epoch": 0.07557586687115757,
      "grad_norm": 11.57731819152832,
      "learning_rate": 9.202588820679873e-06,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8243474781513214,
      "num_tokens": 29254685.0,
      "step": 24380
    },
    {
      "entropy": 1.795560409873724,
      "epoch": 0.07560686599620725,
      "grad_norm": 2.2424561977386475,
      "learning_rate": 9.20070200032488e-06,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8258658021688461,
      "num_tokens": 29268202.0,
      "step": 24390
    },
    {
      "entropy": 1.896498514711857,
      "epoch": 0.07563786512125695,
      "grad_norm": 11.616437911987305,
      "learning_rate": 9.198816340066862e-06,
      "loss": 0.6379,
      "mean_token_accuracy": 0.8165041267871856,
      "num_tokens": 29280463.0,
      "step": 24400
    },
    {
      "entropy": 1.9246143698692322,
      "epoch": 0.07566886424630664,
      "grad_norm": 10.278308868408203,
      "learning_rate": 9.19693183871751e-06,
      "loss": 0.5943,
      "mean_token_accuracy": 0.8151498556137085,
      "num_tokens": 29292665.0,
      "step": 24410
    },
    {
      "entropy": 2.0228449046611785,
      "epoch": 0.07569986337135634,
      "grad_norm": 9.317980766296387,
      "learning_rate": 9.19504849509022e-06,
      "loss": 0.6758,
      "mean_token_accuracy": 0.7997980415821075,
      "num_tokens": 29304324.0,
      "step": 24420
    },
    {
      "entropy": 1.9563984453678132,
      "epoch": 0.07573086249640604,
      "grad_norm": 4.3756422996521,
      "learning_rate": 9.19316630800009e-06,
      "loss": 0.6041,
      "mean_token_accuracy": 0.8073103711009025,
      "num_tokens": 29316485.0,
      "step": 24430
    },
    {
      "entropy": 1.9286874875426292,
      "epoch": 0.07576186162145573,
      "grad_norm": 9.686562538146973,
      "learning_rate": 9.191285276263909e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8244198396801948,
      "num_tokens": 29329320.0,
      "step": 24440
    },
    {
      "entropy": 1.9644469603896142,
      "epoch": 0.07579286074650543,
      "grad_norm": 8.78291130065918,
      "learning_rate": 9.18940539870017e-06,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8178132340312004,
      "num_tokens": 29340967.0,
      "step": 24450
    },
    {
      "entropy": 1.981376151740551,
      "epoch": 0.07582385987155513,
      "grad_norm": 10.671409606933594,
      "learning_rate": 9.187526674129046e-06,
      "loss": 0.6293,
      "mean_token_accuracy": 0.807844452559948,
      "num_tokens": 29353256.0,
      "step": 24460
    },
    {
      "entropy": 1.9730531215667724,
      "epoch": 0.07585485899660482,
      "grad_norm": 9.958746910095215,
      "learning_rate": 9.185649101372406e-06,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8276485189795494,
      "num_tokens": 29363717.0,
      "step": 24470
    },
    {
      "entropy": 1.9136406406760216,
      "epoch": 0.07588585812165452,
      "grad_norm": 5.672607898712158,
      "learning_rate": 9.1837726792538e-06,
      "loss": 0.5799,
      "mean_token_accuracy": 0.8139689520001412,
      "num_tokens": 29375747.0,
      "step": 24480
    },
    {
      "entropy": 1.8405794501304626,
      "epoch": 0.07591685724670422,
      "grad_norm": 9.598546028137207,
      "learning_rate": 9.18189740659846e-06,
      "loss": 0.5324,
      "mean_token_accuracy": 0.825704300403595,
      "num_tokens": 29388350.0,
      "step": 24490
    },
    {
      "entropy": 1.9111091524362565,
      "epoch": 0.07594785637175391,
      "grad_norm": 11.881855964660645,
      "learning_rate": 9.180023282233297e-06,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8082716703414917,
      "num_tokens": 29400465.0,
      "step": 24500
    },
    {
      "entropy": 1.9688442632555962,
      "epoch": 0.0759788554968036,
      "grad_norm": 9.742838859558105,
      "learning_rate": 9.178150304986897e-06,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8214067354798317,
      "num_tokens": 29411568.0,
      "step": 24510
    },
    {
      "entropy": 1.8937147334218025,
      "epoch": 0.07600985462185329,
      "grad_norm": 10.891999244689941,
      "learning_rate": 9.17627847368952e-06,
      "loss": 0.5447,
      "mean_token_accuracy": 0.8176762789487839,
      "num_tokens": 29424053.0,
      "step": 24520
    },
    {
      "entropy": 1.831982211768627,
      "epoch": 0.07604085374690299,
      "grad_norm": 9.699090003967285,
      "learning_rate": 9.174407787173092e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8441127195954323,
      "num_tokens": 29436498.0,
      "step": 24530
    },
    {
      "entropy": 1.829953595995903,
      "epoch": 0.07607185287195269,
      "grad_norm": 10.976648330688477,
      "learning_rate": 9.172538244271205e-06,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8100987046957016,
      "num_tokens": 29449302.0,
      "step": 24540
    },
    {
      "entropy": 1.8781290456652642,
      "epoch": 0.07610285199700238,
      "grad_norm": 10.623594284057617,
      "learning_rate": 9.170669843819118e-06,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8065455496311188,
      "num_tokens": 29461309.0,
      "step": 24550
    },
    {
      "entropy": 1.9357499971985817,
      "epoch": 0.07613385112205208,
      "grad_norm": 5.401109218597412,
      "learning_rate": 9.16880258465375e-06,
      "loss": 0.6238,
      "mean_token_accuracy": 0.809730452299118,
      "num_tokens": 29473381.0,
      "step": 24560
    },
    {
      "entropy": 1.9581341043114662,
      "epoch": 0.07616485024710178,
      "grad_norm": 9.8287935256958,
      "learning_rate": 9.166936465613671e-06,
      "loss": 0.606,
      "mean_token_accuracy": 0.8168662905693054,
      "num_tokens": 29484514.0,
      "step": 24570
    },
    {
      "entropy": 1.8992894351482392,
      "epoch": 0.07619584937215147,
      "grad_norm": 10.469278335571289,
      "learning_rate": 9.165071485539113e-06,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8262490287423134,
      "num_tokens": 29496684.0,
      "step": 24580
    },
    {
      "entropy": 1.856746034324169,
      "epoch": 0.07622684849720117,
      "grad_norm": 7.977566719055176,
      "learning_rate": 9.163207643271953e-06,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8371643677353859,
      "num_tokens": 29508911.0,
      "step": 24590
    },
    {
      "entropy": 1.8318717867136,
      "epoch": 0.07625784762225087,
      "grad_norm": 7.776499271392822,
      "learning_rate": 9.161344937655717e-06,
      "loss": 0.5138,
      "mean_token_accuracy": 0.8247979655861855,
      "num_tokens": 29522385.0,
      "step": 24600
    },
    {
      "entropy": 1.9623229920864105,
      "epoch": 0.07628884674730056,
      "grad_norm": 11.194042205810547,
      "learning_rate": 9.159483367535581e-06,
      "loss": 0.5956,
      "mean_token_accuracy": 0.8133369132876396,
      "num_tokens": 29533515.0,
      "step": 24610
    },
    {
      "entropy": 1.8496048122644424,
      "epoch": 0.07631984587235026,
      "grad_norm": 9.94607162475586,
      "learning_rate": 9.157622931758355e-06,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8247012510895729,
      "num_tokens": 29546981.0,
      "step": 24620
    },
    {
      "entropy": 1.8561659142374993,
      "epoch": 0.07635084499739994,
      "grad_norm": 4.343288421630859,
      "learning_rate": 9.155763629172494e-06,
      "loss": 0.5875,
      "mean_token_accuracy": 0.8228087961673737,
      "num_tokens": 29559906.0,
      "step": 24630
    },
    {
      "entropy": 1.8643113687634467,
      "epoch": 0.07638184412244964,
      "grad_norm": 3.739152193069458,
      "learning_rate": 9.153905458628086e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8250072717666626,
      "num_tokens": 29572877.0,
      "step": 24640
    },
    {
      "entropy": 1.9850792646408082,
      "epoch": 0.07641284324749933,
      "grad_norm": 11.358891487121582,
      "learning_rate": 9.152048418976852e-06,
      "loss": 0.656,
      "mean_token_accuracy": 0.8045062452554703,
      "num_tokens": 29584254.0,
      "step": 24650
    },
    {
      "entropy": 1.930943602323532,
      "epoch": 0.07644384237254903,
      "grad_norm": 9.818647384643555,
      "learning_rate": 9.150192509072147e-06,
      "loss": 0.5624,
      "mean_token_accuracy": 0.811825430393219,
      "num_tokens": 29596694.0,
      "step": 24660
    },
    {
      "entropy": 1.963192854821682,
      "epoch": 0.07647484149759873,
      "grad_norm": 10.888368606567383,
      "learning_rate": 9.148337727768948e-06,
      "loss": 0.6148,
      "mean_token_accuracy": 0.8172325447201729,
      "num_tokens": 29607626.0,
      "step": 24670
    },
    {
      "entropy": 1.8728375658392906,
      "epoch": 0.07650584062264842,
      "grad_norm": 9.656998634338379,
      "learning_rate": 9.146484073923858e-06,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8192928373813629,
      "num_tokens": 29620518.0,
      "step": 24680
    },
    {
      "entropy": 1.9853510811924935,
      "epoch": 0.07653683974769812,
      "grad_norm": 9.287243843078613,
      "learning_rate": 9.144631546395098e-06,
      "loss": 0.6036,
      "mean_token_accuracy": 0.814111416041851,
      "num_tokens": 29632155.0,
      "step": 24690
    },
    {
      "entropy": 1.974936306476593,
      "epoch": 0.07656783887274782,
      "grad_norm": 9.676168441772461,
      "learning_rate": 9.142780144042515e-06,
      "loss": 0.6892,
      "mean_token_accuracy": 0.8018039375543594,
      "num_tokens": 29643855.0,
      "step": 24700
    },
    {
      "entropy": 1.9820726066827774,
      "epoch": 0.07659883799779751,
      "grad_norm": 10.02213191986084,
      "learning_rate": 9.140929865727566e-06,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8112702906131745,
      "num_tokens": 29655657.0,
      "step": 24710
    },
    {
      "entropy": 1.909799675643444,
      "epoch": 0.07662983712284721,
      "grad_norm": 9.104336738586426,
      "learning_rate": 9.139080710313316e-06,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8222946509718895,
      "num_tokens": 29667862.0,
      "step": 24720
    },
    {
      "entropy": 1.992139181494713,
      "epoch": 0.07666083624789691,
      "grad_norm": 4.088809013366699,
      "learning_rate": 9.137232676664449e-06,
      "loss": 0.66,
      "mean_token_accuracy": 0.7951893359422684,
      "num_tokens": 29679790.0,
      "step": 24730
    },
    {
      "entropy": 1.9576111137866974,
      "epoch": 0.0766918353729466,
      "grad_norm": 11.400434494018555,
      "learning_rate": 9.135385763647246e-06,
      "loss": 0.6328,
      "mean_token_accuracy": 0.8115091070532798,
      "num_tokens": 29691260.0,
      "step": 24740
    },
    {
      "entropy": 1.8499963372945785,
      "epoch": 0.0767228344979963,
      "grad_norm": 9.526920318603516,
      "learning_rate": 9.1335399701296e-06,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8311129599809647,
      "num_tokens": 29704733.0,
      "step": 24750
    },
    {
      "entropy": 1.848706914484501,
      "epoch": 0.07675383362304598,
      "grad_norm": 9.934120178222656,
      "learning_rate": 9.131695294980995e-06,
      "loss": 0.5474,
      "mean_token_accuracy": 0.831969042122364,
      "num_tokens": 29717478.0,
      "step": 24760
    },
    {
      "entropy": 1.912044520676136,
      "epoch": 0.07678483274809568,
      "grad_norm": 10.17774772644043,
      "learning_rate": 9.129851737072522e-06,
      "loss": 0.63,
      "mean_token_accuracy": 0.8104033455252647,
      "num_tokens": 29728251.0,
      "step": 24770
    },
    {
      "entropy": 2.015045040845871,
      "epoch": 0.07681583187314538,
      "grad_norm": 7.955008506774902,
      "learning_rate": 9.128009295276862e-06,
      "loss": 0.6655,
      "mean_token_accuracy": 0.8075503215193749,
      "num_tokens": 29739613.0,
      "step": 24780
    },
    {
      "entropy": 1.8380137085914612,
      "epoch": 0.07684683099819507,
      "grad_norm": 6.014031887054443,
      "learning_rate": 9.126167968468289e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8174324378371238,
      "num_tokens": 29752547.0,
      "step": 24790
    },
    {
      "entropy": 2.001574045419693,
      "epoch": 0.07687783012324477,
      "grad_norm": 10.860368728637695,
      "learning_rate": 9.124327755522661e-06,
      "loss": 0.6676,
      "mean_token_accuracy": 0.808248932659626,
      "num_tokens": 29762867.0,
      "step": 24800
    },
    {
      "entropy": 2.0134127497673036,
      "epoch": 0.07690882924829447,
      "grad_norm": 9.547008514404297,
      "learning_rate": 9.122488655317434e-06,
      "loss": 0.7027,
      "mean_token_accuracy": 0.804673321545124,
      "num_tokens": 29773812.0,
      "step": 24810
    },
    {
      "entropy": 1.8855068862438202,
      "epoch": 0.07693982837334416,
      "grad_norm": 11.452570915222168,
      "learning_rate": 9.120650666731632e-06,
      "loss": 0.56,
      "mean_token_accuracy": 0.8221283480525017,
      "num_tokens": 29785964.0,
      "step": 24820
    },
    {
      "entropy": 1.8937954485416413,
      "epoch": 0.07697082749839386,
      "grad_norm": 9.164377212524414,
      "learning_rate": 9.118813788645872e-06,
      "loss": 0.6028,
      "mean_token_accuracy": 0.8206618830561638,
      "num_tokens": 29797873.0,
      "step": 24830
    },
    {
      "entropy": 1.8926649376749993,
      "epoch": 0.07700182662344356,
      "grad_norm": 10.76089096069336,
      "learning_rate": 9.116978019942341e-06,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8190899342298508,
      "num_tokens": 29809898.0,
      "step": 24840
    },
    {
      "entropy": 1.739300973713398,
      "epoch": 0.07703282574849325,
      "grad_norm": 4.010213851928711,
      "learning_rate": 9.115143359504806e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8370007425546646,
      "num_tokens": 29824298.0,
      "step": 24850
    },
    {
      "entropy": 1.8673224583268166,
      "epoch": 0.07706382487354295,
      "grad_norm": 7.60786247253418,
      "learning_rate": 9.113309806218598e-06,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8228920385241508,
      "num_tokens": 29836384.0,
      "step": 24860
    },
    {
      "entropy": 1.8416576564311982,
      "epoch": 0.07709482399859265,
      "grad_norm": 10.18614387512207,
      "learning_rate": 9.111477358970625e-06,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8315700188279151,
      "num_tokens": 29848465.0,
      "step": 24870
    },
    {
      "entropy": 1.833722086250782,
      "epoch": 0.07712582312364233,
      "grad_norm": 9.657119750976562,
      "learning_rate": 9.109646016649358e-06,
      "loss": 0.5351,
      "mean_token_accuracy": 0.822961401939392,
      "num_tokens": 29860315.0,
      "step": 24880
    },
    {
      "entropy": 1.9155391678214073,
      "epoch": 0.07715682224869203,
      "grad_norm": 9.408291816711426,
      "learning_rate": 9.107815778144829e-06,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8233883902430534,
      "num_tokens": 29871950.0,
      "step": 24890
    },
    {
      "entropy": 1.8746279999613762,
      "epoch": 0.07718782137374172,
      "grad_norm": 9.93213176727295,
      "learning_rate": 9.105986642348637e-06,
      "loss": 0.6101,
      "mean_token_accuracy": 0.8215365827083587,
      "num_tokens": 29883850.0,
      "step": 24900
    },
    {
      "entropy": 1.997341087460518,
      "epoch": 0.07721882049879142,
      "grad_norm": 10.22244930267334,
      "learning_rate": 9.104158608153925e-06,
      "loss": 0.6552,
      "mean_token_accuracy": 0.8083247482776642,
      "num_tokens": 29894752.0,
      "step": 24910
    },
    {
      "entropy": 1.9021727859973907,
      "epoch": 0.07724981962384112,
      "grad_norm": 10.58688735961914,
      "learning_rate": 9.10233167445541e-06,
      "loss": 0.5976,
      "mean_token_accuracy": 0.8210710018873215,
      "num_tokens": 29906184.0,
      "step": 24920
    },
    {
      "entropy": 1.9085007548332213,
      "epoch": 0.07728081874889081,
      "grad_norm": 10.130495071411133,
      "learning_rate": 9.100505840149343e-06,
      "loss": 0.5777,
      "mean_token_accuracy": 0.8177463442087174,
      "num_tokens": 29918911.0,
      "step": 24930
    },
    {
      "entropy": 1.949122653901577,
      "epoch": 0.07731181787394051,
      "grad_norm": 10.247212409973145,
      "learning_rate": 9.09868110413354e-06,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8257071673870087,
      "num_tokens": 29930418.0,
      "step": 24940
    },
    {
      "entropy": 1.8943613111972808,
      "epoch": 0.0773428169989902,
      "grad_norm": 8.908804893493652,
      "learning_rate": 9.096857465307348e-06,
      "loss": 0.5782,
      "mean_token_accuracy": 0.8173254355788231,
      "num_tokens": 29942796.0,
      "step": 24950
    },
    {
      "entropy": 1.8852560982108115,
      "epoch": 0.0773738161240399,
      "grad_norm": 9.335540771484375,
      "learning_rate": 9.095034922571667e-06,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8281083196401596,
      "num_tokens": 29954762.0,
      "step": 24960
    },
    {
      "entropy": 1.9339411437511445,
      "epoch": 0.0774048152490896,
      "grad_norm": 8.73486042022705,
      "learning_rate": 9.093213474828937e-06,
      "loss": 0.6839,
      "mean_token_accuracy": 0.8056422159075737,
      "num_tokens": 29966209.0,
      "step": 24970
    },
    {
      "entropy": 1.8483040496706962,
      "epoch": 0.0774358143741393,
      "grad_norm": 16.429473876953125,
      "learning_rate": 9.091393120983134e-06,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8213301405310631,
      "num_tokens": 29979277.0,
      "step": 24980
    },
    {
      "entropy": 1.9246578827500342,
      "epoch": 0.07746681349918899,
      "grad_norm": 11.819490432739258,
      "learning_rate": 9.089573859939769e-06,
      "loss": 0.6234,
      "mean_token_accuracy": 0.8150523856282235,
      "num_tokens": 29991214.0,
      "step": 24990
    },
    {
      "entropy": 1.8754414036870002,
      "epoch": 0.07749781262423869,
      "grad_norm": 3.053318500518799,
      "learning_rate": 9.087755690605889e-06,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8337288603186608,
      "num_tokens": 30004088.0,
      "step": 25000
    },
    {
      "entropy": 1.958650803565979,
      "epoch": 0.07752881174928837,
      "grad_norm": 9.075653076171875,
      "learning_rate": 9.085938611890065e-06,
      "loss": 0.6506,
      "mean_token_accuracy": 0.8138946458697319,
      "num_tokens": 30015387.0,
      "step": 25010
    },
    {
      "entropy": 1.831257238984108,
      "epoch": 0.07755981087433807,
      "grad_norm": 7.205061912536621,
      "learning_rate": 9.084122622702402e-06,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8284579709172248,
      "num_tokens": 30028069.0,
      "step": 25020
    },
    {
      "entropy": 1.921176479756832,
      "epoch": 0.07759080999938776,
      "grad_norm": 9.383119583129883,
      "learning_rate": 9.082307721954523e-06,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8308942928910256,
      "num_tokens": 30039720.0,
      "step": 25030
    },
    {
      "entropy": 1.881229367852211,
      "epoch": 0.07762180912443746,
      "grad_norm": 9.154582023620605,
      "learning_rate": 9.080493908559574e-06,
      "loss": 0.5715,
      "mean_token_accuracy": 0.8158265233039856,
      "num_tokens": 30052310.0,
      "step": 25040
    },
    {
      "entropy": 1.9333179131150247,
      "epoch": 0.07765280824948716,
      "grad_norm": 10.319096565246582,
      "learning_rate": 9.078681181432226e-06,
      "loss": 0.656,
      "mean_token_accuracy": 0.8071909308433532,
      "num_tokens": 30063377.0,
      "step": 25050
    },
    {
      "entropy": 1.9245691820979118,
      "epoch": 0.07768380737453685,
      "grad_norm": 9.286566734313965,
      "learning_rate": 9.076869539488652e-06,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8230229288339614,
      "num_tokens": 30074509.0,
      "step": 25060
    },
    {
      "entropy": 1.9185208335518837,
      "epoch": 0.07771480649958655,
      "grad_norm": 10.67673110961914,
      "learning_rate": 9.075058981646555e-06,
      "loss": 0.6388,
      "mean_token_accuracy": 0.810713978111744,
      "num_tokens": 30086905.0,
      "step": 25070
    },
    {
      "entropy": 1.9412970080971719,
      "epoch": 0.07774580562463625,
      "grad_norm": 8.668137550354004,
      "learning_rate": 9.073249506825138e-06,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8281246423721313,
      "num_tokens": 30098285.0,
      "step": 25080
    },
    {
      "entropy": 1.9020920038223266,
      "epoch": 0.07777680474968594,
      "grad_norm": 4.553186416625977,
      "learning_rate": 9.071441113945115e-06,
      "loss": 0.5749,
      "mean_token_accuracy": 0.8125508442521095,
      "num_tokens": 30109763.0,
      "step": 25090
    },
    {
      "entropy": 1.9090612187981606,
      "epoch": 0.07780780387473564,
      "grad_norm": 9.096549987792969,
      "learning_rate": 9.069633801928704e-06,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8202937185764313,
      "num_tokens": 30121903.0,
      "step": 25100
    },
    {
      "entropy": 1.9104878604412079,
      "epoch": 0.07783880299978534,
      "grad_norm": 10.058632850646973,
      "learning_rate": 9.067827569699626e-06,
      "loss": 0.6086,
      "mean_token_accuracy": 0.8223781570792198,
      "num_tokens": 30134136.0,
      "step": 25110
    },
    {
      "entropy": 1.9066944912075996,
      "epoch": 0.07786980212483503,
      "grad_norm": 7.3937835693359375,
      "learning_rate": 9.066022416183104e-06,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8184296682476997,
      "num_tokens": 30145673.0,
      "step": 25120
    },
    {
      "entropy": 1.977784439921379,
      "epoch": 0.07790080124988472,
      "grad_norm": 8.76796817779541,
      "learning_rate": 9.064218340305854e-06,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8132604837417603,
      "num_tokens": 30156471.0,
      "step": 25130
    },
    {
      "entropy": 1.886358195543289,
      "epoch": 0.07793180037493441,
      "grad_norm": 8.512279510498047,
      "learning_rate": 9.06241534099609e-06,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8291906744241715,
      "num_tokens": 30168512.0,
      "step": 25140
    },
    {
      "entropy": 1.9338295564055443,
      "epoch": 0.07796279949998411,
      "grad_norm": 8.981732368469238,
      "learning_rate": 9.060613417183516e-06,
      "loss": 0.6373,
      "mean_token_accuracy": 0.8155599415302277,
      "num_tokens": 30180559.0,
      "step": 25150
    },
    {
      "entropy": 1.851930246502161,
      "epoch": 0.0779937986250338,
      "grad_norm": 11.297672271728516,
      "learning_rate": 9.058812567799327e-06,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8301180645823478,
      "num_tokens": 30194350.0,
      "step": 25160
    },
    {
      "entropy": 1.932929702103138,
      "epoch": 0.0780247977500835,
      "grad_norm": 4.530979156494141,
      "learning_rate": 9.0570127917762e-06,
      "loss": 0.5372,
      "mean_token_accuracy": 0.823866055905819,
      "num_tokens": 30206582.0,
      "step": 25170
    },
    {
      "entropy": 1.9074819549918174,
      "epoch": 0.0780557968751332,
      "grad_norm": 4.458383560180664,
      "learning_rate": 9.055214088048302e-06,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8161739498376847,
      "num_tokens": 30218782.0,
      "step": 25180
    },
    {
      "entropy": 1.9132570594549179,
      "epoch": 0.0780867960001829,
      "grad_norm": Infinity,
      "learning_rate": 9.053416455551274e-06,
      "loss": 0.5894,
      "mean_token_accuracy": 0.8196236163377761,
      "num_tokens": 30230586.0,
      "step": 25190
    },
    {
      "entropy": 1.8898057445883751,
      "epoch": 0.07811779512523259,
      "grad_norm": 10.808612823486328,
      "learning_rate": 9.051619893222242e-06,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8312712505459785,
      "num_tokens": 30242839.0,
      "step": 25200
    },
    {
      "entropy": 1.8855190083384514,
      "epoch": 0.07814879425028229,
      "grad_norm": 10.371384620666504,
      "learning_rate": 9.049824399999807e-06,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8088703706860543,
      "num_tokens": 30255525.0,
      "step": 25210
    },
    {
      "entropy": 1.8693858250975608,
      "epoch": 0.07817979337533199,
      "grad_norm": 11.21320629119873,
      "learning_rate": 9.048029974824037e-06,
      "loss": 0.5543,
      "mean_token_accuracy": 0.826404581964016,
      "num_tokens": 30268078.0,
      "step": 25220
    },
    {
      "entropy": 1.9001759216189384,
      "epoch": 0.07821079250038168,
      "grad_norm": 8.04379940032959,
      "learning_rate": 9.046236616636477e-06,
      "loss": 0.6188,
      "mean_token_accuracy": 0.8241765573620796,
      "num_tokens": 30280040.0,
      "step": 25230
    },
    {
      "entropy": 1.9273234009742737,
      "epoch": 0.07824179162543138,
      "grad_norm": 11.019646644592285,
      "learning_rate": 9.044444324380139e-06,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8227365925908089,
      "num_tokens": 30290832.0,
      "step": 25240
    },
    {
      "entropy": 1.9361531496047975,
      "epoch": 0.07827279075048106,
      "grad_norm": 3.8891756534576416,
      "learning_rate": 9.042653096999496e-06,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8081121280789375,
      "num_tokens": 30302299.0,
      "step": 25250
    },
    {
      "entropy": 1.96910317838192,
      "epoch": 0.07830378987553076,
      "grad_norm": 8.430257797241211,
      "learning_rate": 9.04086293344049e-06,
      "loss": 0.5955,
      "mean_token_accuracy": 0.8174209460616112,
      "num_tokens": 30313745.0,
      "step": 25260
    },
    {
      "entropy": 1.8507313832640648,
      "epoch": 0.07833478900058045,
      "grad_norm": 10.676153182983398,
      "learning_rate": 9.039073832650518e-06,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8205841019749641,
      "num_tokens": 30326883.0,
      "step": 25270
    },
    {
      "entropy": 1.963607743382454,
      "epoch": 0.07836578812563015,
      "grad_norm": 10.588327407836914,
      "learning_rate": 9.037285793578439e-06,
      "loss": 0.6089,
      "mean_token_accuracy": 0.8108539909124375,
      "num_tokens": 30338102.0,
      "step": 25280
    },
    {
      "entropy": 1.9715974554419518,
      "epoch": 0.07839678725067985,
      "grad_norm": 9.336774826049805,
      "learning_rate": 9.035498815174564e-06,
      "loss": 0.6256,
      "mean_token_accuracy": 0.8076756104826928,
      "num_tokens": 30349489.0,
      "step": 25290
    },
    {
      "entropy": 1.9550383672118188,
      "epoch": 0.07842778637572954,
      "grad_norm": 9.92357063293457,
      "learning_rate": 9.033712896390654e-06,
      "loss": 0.6597,
      "mean_token_accuracy": 0.803634025156498,
      "num_tokens": 30361182.0,
      "step": 25300
    },
    {
      "entropy": 1.9395698383450508,
      "epoch": 0.07845878550077924,
      "grad_norm": 8.71617603302002,
      "learning_rate": 9.031928036179925e-06,
      "loss": 0.631,
      "mean_token_accuracy": 0.8065466418862343,
      "num_tokens": 30372996.0,
      "step": 25310
    },
    {
      "entropy": 1.847324576973915,
      "epoch": 0.07848978462582894,
      "grad_norm": 5.6480584144592285,
      "learning_rate": 9.030144233497038e-06,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8144370660185813,
      "num_tokens": 30386024.0,
      "step": 25320
    },
    {
      "entropy": 1.8323638439178467,
      "epoch": 0.07852078375087863,
      "grad_norm": 10.176698684692383,
      "learning_rate": 9.028361487298097e-06,
      "loss": 0.5583,
      "mean_token_accuracy": 0.8260697081685067,
      "num_tokens": 30399072.0,
      "step": 25330
    },
    {
      "entropy": 1.9654901623725891,
      "epoch": 0.07855178287592833,
      "grad_norm": 10.919507026672363,
      "learning_rate": 9.026579796540651e-06,
      "loss": 0.6562,
      "mean_token_accuracy": 0.8032170712947846,
      "num_tokens": 30410445.0,
      "step": 25340
    },
    {
      "entropy": 1.8728495821356774,
      "epoch": 0.07858278200097803,
      "grad_norm": 9.558503150939941,
      "learning_rate": 9.024799160183686e-06,
      "loss": 0.5595,
      "mean_token_accuracy": 0.82089733928442,
      "num_tokens": 30422438.0,
      "step": 25350
    },
    {
      "entropy": 1.8767936065793038,
      "epoch": 0.07861378112602772,
      "grad_norm": 6.74028205871582,
      "learning_rate": 9.023019577187625e-06,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8307792738080024,
      "num_tokens": 30435068.0,
      "step": 25360
    },
    {
      "entropy": 1.9264275386929512,
      "epoch": 0.07864478025107742,
      "grad_norm": 9.8104829788208,
      "learning_rate": 9.021241046514326e-06,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8222647801041603,
      "num_tokens": 30446563.0,
      "step": 25370
    },
    {
      "entropy": 1.93300940990448,
      "epoch": 0.0786757793761271,
      "grad_norm": 9.438268661499023,
      "learning_rate": 9.019463567127084e-06,
      "loss": 0.6245,
      "mean_token_accuracy": 0.813123632967472,
      "num_tokens": 30457702.0,
      "step": 25380
    },
    {
      "entropy": 1.9117926597595214,
      "epoch": 0.0787067785011768,
      "grad_norm": 10.080045700073242,
      "learning_rate": 9.017687137990611e-06,
      "loss": 0.6594,
      "mean_token_accuracy": 0.8094756618142128,
      "num_tokens": 30469833.0,
      "step": 25390
    },
    {
      "entropy": 1.9238570496439933,
      "epoch": 0.0787377776262265,
      "grad_norm": 9.324179649353027,
      "learning_rate": 9.01591175807106e-06,
      "loss": 0.6385,
      "mean_token_accuracy": 0.8032417684793473,
      "num_tokens": 30481722.0,
      "step": 25400
    },
    {
      "entropy": 1.9335221163928509,
      "epoch": 0.07876877675127619,
      "grad_norm": 10.291169166564941,
      "learning_rate": 9.014137426335997e-06,
      "loss": 0.6627,
      "mean_token_accuracy": 0.8041873052716255,
      "num_tokens": 30494711.0,
      "step": 25410
    },
    {
      "entropy": 1.8992304280400276,
      "epoch": 0.07879977587632589,
      "grad_norm": 9.40573787689209,
      "learning_rate": 9.012364141754415e-06,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8209843635559082,
      "num_tokens": 30506481.0,
      "step": 25420
    },
    {
      "entropy": 1.9901362270116807,
      "epoch": 0.07883077500137559,
      "grad_norm": 11.049663543701172,
      "learning_rate": 9.010591903296726e-06,
      "loss": 0.652,
      "mean_token_accuracy": 0.8064600333571434,
      "num_tokens": 30517738.0,
      "step": 25430
    },
    {
      "entropy": 1.9595788344740868,
      "epoch": 0.07886177412642528,
      "grad_norm": 9.368852615356445,
      "learning_rate": 9.008820709934756e-06,
      "loss": 0.6615,
      "mean_token_accuracy": 0.7978012830018997,
      "num_tokens": 30529230.0,
      "step": 25440
    },
    {
      "entropy": 1.8642796009778977,
      "epoch": 0.07889277325147498,
      "grad_norm": 5.745087146759033,
      "learning_rate": 9.00705056064175e-06,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8303910657763481,
      "num_tokens": 30542462.0,
      "step": 25450
    },
    {
      "entropy": 1.890346498787403,
      "epoch": 0.07892377237652468,
      "grad_norm": 12.036408424377441,
      "learning_rate": 9.00528145439236e-06,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8230859890580178,
      "num_tokens": 30554735.0,
      "step": 25460
    },
    {
      "entropy": 1.9690846800804138,
      "epoch": 0.07895477150157437,
      "grad_norm": 10.138855934143066,
      "learning_rate": 9.00351339016265e-06,
      "loss": 0.6098,
      "mean_token_accuracy": 0.8199924737215042,
      "num_tokens": 30566201.0,
      "step": 25470
    },
    {
      "entropy": 2.0111054688692094,
      "epoch": 0.07898577062662407,
      "grad_norm": 8.961263656616211,
      "learning_rate": 9.001746366930088e-06,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8159183651208878,
      "num_tokens": 30577578.0,
      "step": 25480
    },
    {
      "entropy": 1.9971545487642288,
      "epoch": 0.07901676975167377,
      "grad_norm": 8.601164817810059,
      "learning_rate": 8.99998038367355e-06,
      "loss": 0.6461,
      "mean_token_accuracy": 0.8124226480722427,
      "num_tokens": 30589416.0,
      "step": 25490
    },
    {
      "entropy": 2.0272182136774064,
      "epoch": 0.07904776887672345,
      "grad_norm": 9.910305976867676,
      "learning_rate": 8.99821543937331e-06,
      "loss": 0.7675,
      "mean_token_accuracy": 0.8009614288806916,
      "num_tokens": 30601049.0,
      "step": 25500
    },
    {
      "entropy": 1.9942003041505814,
      "epoch": 0.07907876800177314,
      "grad_norm": 9.830111503601074,
      "learning_rate": 8.996451533011044e-06,
      "loss": 0.62,
      "mean_token_accuracy": 0.8120215192437172,
      "num_tokens": 30611769.0,
      "step": 25510
    },
    {
      "entropy": 1.929137869179249,
      "epoch": 0.07910976712682284,
      "grad_norm": 10.363052368164062,
      "learning_rate": 8.994688663569825e-06,
      "loss": 0.6134,
      "mean_token_accuracy": 0.8164379015564919,
      "num_tokens": 30623572.0,
      "step": 25520
    },
    {
      "entropy": 1.9560409665107727,
      "epoch": 0.07914076625187254,
      "grad_norm": 10.759134292602539,
      "learning_rate": 8.992926830034117e-06,
      "loss": 0.609,
      "mean_token_accuracy": 0.8174306198954582,
      "num_tokens": 30634807.0,
      "step": 25530
    },
    {
      "entropy": 2.0076246559619904,
      "epoch": 0.07917176537692223,
      "grad_norm": 10.531962394714355,
      "learning_rate": 8.991166031389779e-06,
      "loss": 0.7254,
      "mean_token_accuracy": 0.7958659037947655,
      "num_tokens": 30645607.0,
      "step": 25540
    },
    {
      "entropy": 1.8035003036260604,
      "epoch": 0.07920276450197193,
      "grad_norm": 8.862020492553711,
      "learning_rate": 8.989406266624054e-06,
      "loss": 0.4378,
      "mean_token_accuracy": 0.8288759797811508,
      "num_tokens": 30659160.0,
      "step": 25550
    },
    {
      "entropy": 1.9678311094641685,
      "epoch": 0.07923376362702163,
      "grad_norm": 9.012432098388672,
      "learning_rate": 8.987647534725585e-06,
      "loss": 0.6402,
      "mean_token_accuracy": 0.8036282330751419,
      "num_tokens": 30670411.0,
      "step": 25560
    },
    {
      "entropy": 1.9711192145943641,
      "epoch": 0.07926476275207132,
      "grad_norm": 10.310850143432617,
      "learning_rate": 8.985889834684384e-06,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8189939826726913,
      "num_tokens": 30682267.0,
      "step": 25570
    },
    {
      "entropy": 1.792180709540844,
      "epoch": 0.07929576187712102,
      "grad_norm": 12.734272956848145,
      "learning_rate": 8.984133165491855e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8344605669379235,
      "num_tokens": 30695155.0,
      "step": 25580
    },
    {
      "entropy": 1.9018129274249076,
      "epoch": 0.07932676100217072,
      "grad_norm": 9.23828125,
      "learning_rate": 8.982377526140776e-06,
      "loss": 0.6423,
      "mean_token_accuracy": 0.8115654811263084,
      "num_tokens": 30708151.0,
      "step": 25590
    },
    {
      "entropy": 1.9214498043060302,
      "epoch": 0.07935776012722041,
      "grad_norm": 4.616427421569824,
      "learning_rate": 8.98062291562531e-06,
      "loss": 0.5874,
      "mean_token_accuracy": 0.817217455804348,
      "num_tokens": 30719730.0,
      "step": 25600
    },
    {
      "entropy": 1.8673983976244926,
      "epoch": 0.07938875925227011,
      "grad_norm": 9.74424934387207,
      "learning_rate": 8.978869332940982e-06,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8221541255712509,
      "num_tokens": 30732691.0,
      "step": 25610
    },
    {
      "entropy": 1.8662855371832847,
      "epoch": 0.0794197583773198,
      "grad_norm": 10.79581069946289,
      "learning_rate": 8.977116777084705e-06,
      "loss": 0.5544,
      "mean_token_accuracy": 0.8190583303570748,
      "num_tokens": 30745114.0,
      "step": 25620
    },
    {
      "entropy": 2.0325272887945176,
      "epoch": 0.07945075750236949,
      "grad_norm": 10.473299026489258,
      "learning_rate": 8.97536524705475e-06,
      "loss": 0.677,
      "mean_token_accuracy": 0.8067957848310471,
      "num_tokens": 30756265.0,
      "step": 25630
    },
    {
      "entropy": 1.974963715672493,
      "epoch": 0.07948175662741919,
      "grad_norm": 8.578821182250977,
      "learning_rate": 8.97361474185076e-06,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8203337907791137,
      "num_tokens": 30767729.0,
      "step": 25640
    },
    {
      "entropy": 1.9518730476498605,
      "epoch": 0.07951275575246888,
      "grad_norm": 9.881929397583008,
      "learning_rate": 8.971865260473745e-06,
      "loss": 0.6164,
      "mean_token_accuracy": 0.8127556294202805,
      "num_tokens": 30779093.0,
      "step": 25650
    },
    {
      "entropy": 1.9101403415203095,
      "epoch": 0.07954375487751858,
      "grad_norm": 9.09304428100586,
      "learning_rate": 8.970116801926072e-06,
      "loss": 0.5419,
      "mean_token_accuracy": 0.833724494278431,
      "num_tokens": 30790837.0,
      "step": 25660
    },
    {
      "entropy": 1.8745130628347397,
      "epoch": 0.07957475400256828,
      "grad_norm": 9.2150239944458,
      "learning_rate": 8.968369365211478e-06,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8339875742793084,
      "num_tokens": 30802641.0,
      "step": 25670
    },
    {
      "entropy": 1.94149309694767,
      "epoch": 0.07960575312761797,
      "grad_norm": 9.093975067138672,
      "learning_rate": 8.966622949335044e-06,
      "loss": 0.6163,
      "mean_token_accuracy": 0.8239960089325905,
      "num_tokens": 30813720.0,
      "step": 25680
    },
    {
      "entropy": 1.8855563715100288,
      "epoch": 0.07963675225266767,
      "grad_norm": 4.550987243652344,
      "learning_rate": 8.964877553303222e-06,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8233138158917427,
      "num_tokens": 30825831.0,
      "step": 25690
    },
    {
      "entropy": 1.9300659999251366,
      "epoch": 0.07966775137771737,
      "grad_norm": 10.610151290893555,
      "learning_rate": 8.963133176123809e-06,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8244445592164993,
      "num_tokens": 30837291.0,
      "step": 25700
    },
    {
      "entropy": 1.9518328681588173,
      "epoch": 0.07969875050276706,
      "grad_norm": 10.140637397766113,
      "learning_rate": 8.96138981680595e-06,
      "loss": 0.6133,
      "mean_token_accuracy": 0.8174855709075928,
      "num_tokens": 30848725.0,
      "step": 25710
    },
    {
      "entropy": 1.9481216147542,
      "epoch": 0.07972974962781676,
      "grad_norm": 5.342643737792969,
      "learning_rate": 8.959647474360146e-06,
      "loss": 0.6286,
      "mean_token_accuracy": 0.8104046359658241,
      "num_tokens": 30860275.0,
      "step": 25720
    },
    {
      "entropy": 1.9668898478150367,
      "epoch": 0.07976074875286646,
      "grad_norm": 9.690641403198242,
      "learning_rate": 8.95790614779824e-06,
      "loss": 0.644,
      "mean_token_accuracy": 0.8048096477985383,
      "num_tokens": 30872020.0,
      "step": 25730
    },
    {
      "entropy": 1.895005388557911,
      "epoch": 0.07979174787791615,
      "grad_norm": 13.030941009521484,
      "learning_rate": 8.956165836133419e-06,
      "loss": 0.6495,
      "mean_token_accuracy": 0.8069834470748901,
      "num_tokens": 30884853.0,
      "step": 25740
    },
    {
      "entropy": 1.9691629260778427,
      "epoch": 0.07982274700296584,
      "grad_norm": 11.10064697265625,
      "learning_rate": 8.954426538380212e-06,
      "loss": 0.6367,
      "mean_token_accuracy": 0.8148859232664108,
      "num_tokens": 30895623.0,
      "step": 25750
    },
    {
      "entropy": 1.9379816070199012,
      "epoch": 0.07985374612801553,
      "grad_norm": 7.397519588470459,
      "learning_rate": 8.952688253554488e-06,
      "loss": 0.6139,
      "mean_token_accuracy": 0.820967635512352,
      "num_tokens": 30907044.0,
      "step": 25760
    },
    {
      "entropy": 1.8707928493618966,
      "epoch": 0.07988474525306523,
      "grad_norm": 8.89204216003418,
      "learning_rate": 8.950950980673451e-06,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8097473427653312,
      "num_tokens": 30919270.0,
      "step": 25770
    },
    {
      "entropy": 1.8771313697099685,
      "epoch": 0.07991574437811493,
      "grad_norm": 9.017126083374023,
      "learning_rate": 8.949214718755644e-06,
      "loss": 0.6379,
      "mean_token_accuracy": 0.8157517641782761,
      "num_tokens": 30931930.0,
      "step": 25780
    },
    {
      "entropy": 1.9514517337083817,
      "epoch": 0.07994674350316462,
      "grad_norm": 9.412887573242188,
      "learning_rate": 8.947479466820933e-06,
      "loss": 0.6404,
      "mean_token_accuracy": 0.807883444428444,
      "num_tokens": 30943436.0,
      "step": 25790
    },
    {
      "entropy": 1.9551150798797607,
      "epoch": 0.07997774262821432,
      "grad_norm": 10.580997467041016,
      "learning_rate": 8.945745223890525e-06,
      "loss": 0.635,
      "mean_token_accuracy": 0.8118805930018425,
      "num_tokens": 30955093.0,
      "step": 25800
    },
    {
      "entropy": 1.9215492516756059,
      "epoch": 0.08000874175326401,
      "grad_norm": 4.973093509674072,
      "learning_rate": 8.944011988986943e-06,
      "loss": 0.6226,
      "mean_token_accuracy": 0.8189863070845604,
      "num_tokens": 30966468.0,
      "step": 25810
    },
    {
      "entropy": 1.9604627519845963,
      "epoch": 0.08003974087831371,
      "grad_norm": 11.397422790527344,
      "learning_rate": 8.942279761134045e-06,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8068236276507378,
      "num_tokens": 30978392.0,
      "step": 25820
    },
    {
      "entropy": 1.8480182617902756,
      "epoch": 0.08007074000336341,
      "grad_norm": 4.0751261711120605,
      "learning_rate": 8.940548539357008e-06,
      "loss": 0.5905,
      "mean_token_accuracy": 0.81248699426651,
      "num_tokens": 30992369.0,
      "step": 25830
    },
    {
      "entropy": 1.8566480681300164,
      "epoch": 0.0801017391284131,
      "grad_norm": 6.9857916831970215,
      "learning_rate": 8.938818322682328e-06,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8251223266124725,
      "num_tokens": 31005133.0,
      "step": 25840
    },
    {
      "entropy": 1.951548993587494,
      "epoch": 0.0801327382534628,
      "grad_norm": 9.003771781921387,
      "learning_rate": 8.937089110137822e-06,
      "loss": 0.6049,
      "mean_token_accuracy": 0.812137958407402,
      "num_tokens": 31017000.0,
      "step": 25850
    },
    {
      "entropy": 1.9318700328469276,
      "epoch": 0.0801637373785125,
      "grad_norm": 10.689835548400879,
      "learning_rate": 8.935360900752618e-06,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8234897822141647,
      "num_tokens": 31028427.0,
      "step": 25860
    },
    {
      "entropy": 1.9848022490739823,
      "epoch": 0.08019473650356218,
      "grad_norm": 9.83168888092041,
      "learning_rate": 8.933633693557168e-06,
      "loss": 0.6557,
      "mean_token_accuracy": 0.8102818682789803,
      "num_tokens": 31039440.0,
      "step": 25870
    },
    {
      "entropy": 1.8783807411789895,
      "epoch": 0.08022573562861188,
      "grad_norm": 10.153984069824219,
      "learning_rate": 8.931907487583224e-06,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8286775290966034,
      "num_tokens": 31051086.0,
      "step": 25880
    },
    {
      "entropy": 1.9235938981175422,
      "epoch": 0.08025673475366157,
      "grad_norm": 10.353632926940918,
      "learning_rate": 8.930182281863854e-06,
      "loss": 0.6119,
      "mean_token_accuracy": 0.8167414903640747,
      "num_tokens": 31062372.0,
      "step": 25890
    },
    {
      "entropy": 1.9412924006581307,
      "epoch": 0.08028773387871127,
      "grad_norm": 9.35558032989502,
      "learning_rate": 8.928458075433428e-06,
      "loss": 0.6887,
      "mean_token_accuracy": 0.7980827406048775,
      "num_tokens": 31074103.0,
      "step": 25900
    },
    {
      "entropy": 1.9117850810289383,
      "epoch": 0.08031873300376097,
      "grad_norm": 8.50671100616455,
      "learning_rate": 8.926734867327626e-06,
      "loss": 0.5881,
      "mean_token_accuracy": 0.80945935100317,
      "num_tokens": 31086274.0,
      "step": 25910
    },
    {
      "entropy": 1.9611097499728203,
      "epoch": 0.08034973212881066,
      "grad_norm": 8.947006225585938,
      "learning_rate": 8.925012656583428e-06,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8106518238782883,
      "num_tokens": 31097609.0,
      "step": 25920
    },
    {
      "entropy": 1.941481387615204,
      "epoch": 0.08038073125386036,
      "grad_norm": 4.701364040374756,
      "learning_rate": 8.923291442239114e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8272410377860069,
      "num_tokens": 31109738.0,
      "step": 25930
    },
    {
      "entropy": 1.9850254267454148,
      "epoch": 0.08041173037891006,
      "grad_norm": 10.166258811950684,
      "learning_rate": 8.921571223334262e-06,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8136166602373123,
      "num_tokens": 31120962.0,
      "step": 25940
    },
    {
      "entropy": 1.8603035241365433,
      "epoch": 0.08044272950395975,
      "grad_norm": 9.798973083496094,
      "learning_rate": 8.919851998909738e-06,
      "loss": 0.5282,
      "mean_token_accuracy": 0.833336153626442,
      "num_tokens": 31133382.0,
      "step": 25950
    },
    {
      "entropy": 1.9528677210211753,
      "epoch": 0.08047372862900945,
      "grad_norm": 10.009119033813477,
      "learning_rate": 8.918133768007718e-06,
      "loss": 0.6604,
      "mean_token_accuracy": 0.806072898209095,
      "num_tokens": 31144362.0,
      "step": 25960
    },
    {
      "entropy": 1.8188411056995393,
      "epoch": 0.08050472775405915,
      "grad_norm": 8.478353500366211,
      "learning_rate": 8.916416529671652e-06,
      "loss": 0.5355,
      "mean_token_accuracy": 0.8259258419275284,
      "num_tokens": 31157130.0,
      "step": 25970
    },
    {
      "entropy": 1.855710941553116,
      "epoch": 0.08053572687910884,
      "grad_norm": 4.531567096710205,
      "learning_rate": 8.914700282946292e-06,
      "loss": 0.5377,
      "mean_token_accuracy": 0.826113897562027,
      "num_tokens": 31169756.0,
      "step": 25980
    },
    {
      "entropy": 1.8293490082025528,
      "epoch": 0.08056672600415853,
      "grad_norm": 4.700442790985107,
      "learning_rate": 8.912985026877668e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8252740427851677,
      "num_tokens": 31182565.0,
      "step": 25990
    },
    {
      "entropy": 1.8819095849990846,
      "epoch": 0.08059772512920822,
      "grad_norm": 10.227608680725098,
      "learning_rate": 8.911270760513097e-06,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8177149787545204,
      "num_tokens": 31194457.0,
      "step": 26000
    },
    {
      "entropy": 1.9495192840695381,
      "epoch": 0.08062872425425792,
      "grad_norm": 10.328841209411621,
      "learning_rate": 8.909557482901179e-06,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8159580215811729,
      "num_tokens": 31205886.0,
      "step": 26010
    },
    {
      "entropy": 1.9724682122468948,
      "epoch": 0.08065972337930762,
      "grad_norm": 10.351836204528809,
      "learning_rate": 8.907845193091793e-06,
      "loss": 0.6892,
      "mean_token_accuracy": 0.7933256924152374,
      "num_tokens": 31216798.0,
      "step": 26020
    },
    {
      "entropy": 1.8941216632723807,
      "epoch": 0.08069072250435731,
      "grad_norm": 11.160825729370117,
      "learning_rate": 8.906133890136095e-06,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8280467942357064,
      "num_tokens": 31229398.0,
      "step": 26030
    },
    {
      "entropy": 1.9749949797987938,
      "epoch": 0.08072172162940701,
      "grad_norm": 7.923883438110352,
      "learning_rate": 8.90442357308652e-06,
      "loss": 0.6518,
      "mean_token_accuracy": 0.8020041942596435,
      "num_tokens": 31240647.0,
      "step": 26040
    },
    {
      "entropy": 1.9157247439026832,
      "epoch": 0.0807527207544567,
      "grad_norm": 8.247703552246094,
      "learning_rate": 8.902714240996773e-06,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8275446742773056,
      "num_tokens": 31253058.0,
      "step": 26050
    },
    {
      "entropy": 1.9141668871045112,
      "epoch": 0.0807837198795064,
      "grad_norm": 4.465513706207275,
      "learning_rate": 8.901005892921827e-06,
      "loss": 0.5543,
      "mean_token_accuracy": 0.81632010191679,
      "num_tokens": 31265597.0,
      "step": 26060
    },
    {
      "entropy": 1.998997524380684,
      "epoch": 0.0808147190045561,
      "grad_norm": 9.465398788452148,
      "learning_rate": 8.899298527917932e-06,
      "loss": 0.625,
      "mean_token_accuracy": 0.8183861076831818,
      "num_tokens": 31276394.0,
      "step": 26070
    },
    {
      "entropy": 1.9701352685689926,
      "epoch": 0.0808457181296058,
      "grad_norm": 9.879446029663086,
      "learning_rate": 8.897592145042599e-06,
      "loss": 0.6342,
      "mean_token_accuracy": 0.8082750916481019,
      "num_tokens": 31287446.0,
      "step": 26080
    },
    {
      "entropy": 1.8816271275281906,
      "epoch": 0.08087671725465549,
      "grad_norm": 10.568394660949707,
      "learning_rate": 8.895886743354607e-06,
      "loss": 0.538,
      "mean_token_accuracy": 0.8256219759583473,
      "num_tokens": 31299916.0,
      "step": 26090
    },
    {
      "entropy": 1.9000601902604104,
      "epoch": 0.08090771637970519,
      "grad_norm": 9.578839302062988,
      "learning_rate": 8.89418232191399e-06,
      "loss": 0.57,
      "mean_token_accuracy": 0.8156951382756233,
      "num_tokens": 31313478.0,
      "step": 26100
    },
    {
      "entropy": 1.9269771069288253,
      "epoch": 0.08093871550475489,
      "grad_norm": 5.241694927215576,
      "learning_rate": 8.892478879782055e-06,
      "loss": 0.5895,
      "mean_token_accuracy": 0.822375500202179,
      "num_tokens": 31325006.0,
      "step": 26110
    },
    {
      "entropy": 1.915390558540821,
      "epoch": 0.08096971462980457,
      "grad_norm": 10.392666816711426,
      "learning_rate": 8.89077641602135e-06,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8208850100636482,
      "num_tokens": 31336358.0,
      "step": 26120
    },
    {
      "entropy": 1.8701285541057586,
      "epoch": 0.08100071375485426,
      "grad_norm": 8.39531135559082,
      "learning_rate": 8.889074929695697e-06,
      "loss": 0.5223,
      "mean_token_accuracy": 0.8237580001354218,
      "num_tokens": 31349262.0,
      "step": 26130
    },
    {
      "entropy": 1.9644879460334779,
      "epoch": 0.08103171287990396,
      "grad_norm": 8.174306869506836,
      "learning_rate": 8.887374419870162e-06,
      "loss": 0.6391,
      "mean_token_accuracy": 0.8160557880997658,
      "num_tokens": 31360978.0,
      "step": 26140
    },
    {
      "entropy": 1.870239832997322,
      "epoch": 0.08106271200495366,
      "grad_norm": 4.718651294708252,
      "learning_rate": 8.885674885611059e-06,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8182685613632202,
      "num_tokens": 31374020.0,
      "step": 26150
    },
    {
      "entropy": 1.8591389670968055,
      "epoch": 0.08109371113000335,
      "grad_norm": 8.56635570526123,
      "learning_rate": 8.883976325985959e-06,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8392164379358291,
      "num_tokens": 31386295.0,
      "step": 26160
    },
    {
      "entropy": 2.010419914126396,
      "epoch": 0.08112471025505305,
      "grad_norm": 10.94079303741455,
      "learning_rate": 8.882278740063671e-06,
      "loss": 0.7017,
      "mean_token_accuracy": 0.804059025645256,
      "num_tokens": 31397492.0,
      "step": 26170
    },
    {
      "entropy": 1.9494641929864884,
      "epoch": 0.08115570938010275,
      "grad_norm": 3.204777717590332,
      "learning_rate": 8.880582126914265e-06,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8109059870243073,
      "num_tokens": 31408976.0,
      "step": 26180
    },
    {
      "entropy": 1.9177653357386588,
      "epoch": 0.08118670850515244,
      "grad_norm": 9.458409309387207,
      "learning_rate": 8.878886485609038e-06,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8205662295222282,
      "num_tokens": 31421691.0,
      "step": 26190
    },
    {
      "entropy": 1.941348561644554,
      "epoch": 0.08121770763020214,
      "grad_norm": 10.314962387084961,
      "learning_rate": 8.877191815220537e-06,
      "loss": 0.609,
      "mean_token_accuracy": 0.8177227556705475,
      "num_tokens": 31433019.0,
      "step": 26200
    },
    {
      "entropy": 1.8470569260418415,
      "epoch": 0.08124870675525184,
      "grad_norm": 10.691038131713867,
      "learning_rate": 8.87549811482254e-06,
      "loss": 0.4794,
      "mean_token_accuracy": 0.8288433074951171,
      "num_tokens": 31446595.0,
      "step": 26210
    },
    {
      "entropy": 1.9533916339278221,
      "epoch": 0.08127970588030153,
      "grad_norm": 4.70416784286499,
      "learning_rate": 8.873805383490072e-06,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8235921949148178,
      "num_tokens": 31458289.0,
      "step": 26220
    },
    {
      "entropy": 1.9918930500745773,
      "epoch": 0.08131070500535123,
      "grad_norm": 10.034157752990723,
      "learning_rate": 8.872113620299381e-06,
      "loss": 0.6885,
      "mean_token_accuracy": 0.8030604913830757,
      "num_tokens": 31469163.0,
      "step": 26230
    },
    {
      "entropy": 1.8951246917247773,
      "epoch": 0.08134170413040091,
      "grad_norm": 8.459050178527832,
      "learning_rate": 8.870422824327956e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8330509856343269,
      "num_tokens": 31482597.0,
      "step": 26240
    },
    {
      "entropy": 1.901189935207367,
      "epoch": 0.08137270325545061,
      "grad_norm": 4.526102542877197,
      "learning_rate": 8.868732994654511e-06,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8234093725681305,
      "num_tokens": 31495707.0,
      "step": 26250
    },
    {
      "entropy": 1.8694349378347397,
      "epoch": 0.0814037023805003,
      "grad_norm": 8.740363121032715,
      "learning_rate": 8.86704413035899e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8234608098864555,
      "num_tokens": 31507761.0,
      "step": 26260
    },
    {
      "entropy": 1.90776207447052,
      "epoch": 0.08143470150555,
      "grad_norm": 8.299092292785645,
      "learning_rate": 8.865356230522566e-06,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8166111215949059,
      "num_tokens": 31519733.0,
      "step": 26270
    },
    {
      "entropy": 1.988849925994873,
      "epoch": 0.0814657006305997,
      "grad_norm": 10.71021842956543,
      "learning_rate": 8.86366929422763e-06,
      "loss": 0.6454,
      "mean_token_accuracy": 0.8144510120153428,
      "num_tokens": 31530442.0,
      "step": 26280
    },
    {
      "entropy": 1.9403239041566849,
      "epoch": 0.0814966997556494,
      "grad_norm": 10.783156394958496,
      "learning_rate": 8.861983320557797e-06,
      "loss": 0.619,
      "mean_token_accuracy": 0.8142760202288628,
      "num_tokens": 31541855.0,
      "step": 26290
    },
    {
      "entropy": 1.9194020926952362,
      "epoch": 0.08152769888069909,
      "grad_norm": 5.271920204162598,
      "learning_rate": 8.860298308597903e-06,
      "loss": 0.6006,
      "mean_token_accuracy": 0.8099767789244652,
      "num_tokens": 31553683.0,
      "step": 26300
    },
    {
      "entropy": 1.9054380610585213,
      "epoch": 0.08155869800574879,
      "grad_norm": 12.495100975036621,
      "learning_rate": 8.858614257434004e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8234122693538666,
      "num_tokens": 31566209.0,
      "step": 26310
    },
    {
      "entropy": 1.9283027410507203,
      "epoch": 0.08158969713079849,
      "grad_norm": 10.337760925292969,
      "learning_rate": 8.856931166153366e-06,
      "loss": 0.557,
      "mean_token_accuracy": 0.827317263185978,
      "num_tokens": 31578676.0,
      "step": 26320
    },
    {
      "entropy": 1.9200524538755417,
      "epoch": 0.08162069625584818,
      "grad_norm": 9.985169410705566,
      "learning_rate": 8.855249033844469e-06,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8291199162602425,
      "num_tokens": 31591513.0,
      "step": 26330
    },
    {
      "entropy": 1.8813577458262443,
      "epoch": 0.08165169538089788,
      "grad_norm": 10.70927906036377,
      "learning_rate": 8.85356785959701e-06,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8120563924312592,
      "num_tokens": 31604610.0,
      "step": 26340
    },
    {
      "entropy": 1.9429839253425598,
      "epoch": 0.08168269450594758,
      "grad_norm": 7.395925998687744,
      "learning_rate": 8.851887642501889e-06,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8196773245930672,
      "num_tokens": 31616846.0,
      "step": 26350
    },
    {
      "entropy": 1.8979680463671684,
      "epoch": 0.08171369363099726,
      "grad_norm": 10.052593231201172,
      "learning_rate": 8.850208381651215e-06,
      "loss": 0.5895,
      "mean_token_accuracy": 0.8154531285166741,
      "num_tokens": 31629253.0,
      "step": 26360
    },
    {
      "entropy": 1.894950045645237,
      "epoch": 0.08174469275604695,
      "grad_norm": 10.194618225097656,
      "learning_rate": 8.848530076138306e-06,
      "loss": 0.5183,
      "mean_token_accuracy": 0.827742937207222,
      "num_tokens": 31642045.0,
      "step": 26370
    },
    {
      "entropy": 1.9756985664367677,
      "epoch": 0.08177569188109665,
      "grad_norm": 9.307488441467285,
      "learning_rate": 8.846852725057679e-06,
      "loss": 0.6164,
      "mean_token_accuracy": 0.8118034496903419,
      "num_tokens": 31653833.0,
      "step": 26380
    },
    {
      "entropy": 1.977025419473648,
      "epoch": 0.08180669100614635,
      "grad_norm": 8.45008659362793,
      "learning_rate": 8.845176327505053e-06,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8323731452226639,
      "num_tokens": 31664531.0,
      "step": 26390
    },
    {
      "entropy": 1.8568702943623066,
      "epoch": 0.08183769013119604,
      "grad_norm": 9.432966232299805,
      "learning_rate": 8.843500882577342e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.83240677267313,
      "num_tokens": 31677924.0,
      "step": 26400
    },
    {
      "entropy": 1.9465472564101218,
      "epoch": 0.08186868925624574,
      "grad_norm": 2.6528499126434326,
      "learning_rate": 8.841826389372667e-06,
      "loss": 0.57,
      "mean_token_accuracy": 0.819884067773819,
      "num_tokens": 31689433.0,
      "step": 26410
    },
    {
      "entropy": 1.9157054662704467,
      "epoch": 0.08189968838129544,
      "grad_norm": 4.077538013458252,
      "learning_rate": 8.840152846990336e-06,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8310910239815712,
      "num_tokens": 31701899.0,
      "step": 26420
    },
    {
      "entropy": 1.9014649242162704,
      "epoch": 0.08193068750634513,
      "grad_norm": 9.67369556427002,
      "learning_rate": 8.838480254530852e-06,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8307635292410851,
      "num_tokens": 31713964.0,
      "step": 26430
    },
    {
      "entropy": 1.944985429942608,
      "epoch": 0.08196168663139483,
      "grad_norm": 10.318603515625,
      "learning_rate": 8.836808611095908e-06,
      "loss": 0.589,
      "mean_token_accuracy": 0.8197949156165123,
      "num_tokens": 31725749.0,
      "step": 26440
    },
    {
      "entropy": 1.9587625324726106,
      "epoch": 0.08199268575644453,
      "grad_norm": 9.986980438232422,
      "learning_rate": 8.835137915788388e-06,
      "loss": 0.6011,
      "mean_token_accuracy": 0.816116102039814,
      "num_tokens": 31738281.0,
      "step": 26450
    },
    {
      "entropy": 1.9987529665231705,
      "epoch": 0.08202368488149422,
      "grad_norm": 9.245450973510742,
      "learning_rate": 8.83346816771236e-06,
      "loss": 0.637,
      "mean_token_accuracy": 0.8139676377177238,
      "num_tokens": 31749103.0,
      "step": 26460
    },
    {
      "entropy": 1.9783283829689027,
      "epoch": 0.08205468400654392,
      "grad_norm": 13.349959373474121,
      "learning_rate": 8.831799365973078e-06,
      "loss": 0.6254,
      "mean_token_accuracy": 0.8167249724268913,
      "num_tokens": 31760327.0,
      "step": 26470
    },
    {
      "entropy": 1.9021371573209762,
      "epoch": 0.08208568313159362,
      "grad_norm": 4.939640998840332,
      "learning_rate": 8.83013150967698e-06,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8197105556726456,
      "num_tokens": 31772545.0,
      "step": 26480
    },
    {
      "entropy": 1.8698833830654622,
      "epoch": 0.0821166822566433,
      "grad_norm": 9.22522258758545,
      "learning_rate": 8.828464597931686e-06,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8225023493170738,
      "num_tokens": 31785337.0,
      "step": 26490
    },
    {
      "entropy": 1.8712293177843093,
      "epoch": 0.082147681381693,
      "grad_norm": 5.361429214477539,
      "learning_rate": 8.82679862984599e-06,
      "loss": 0.5483,
      "mean_token_accuracy": 0.8150455951690674,
      "num_tokens": 31797319.0,
      "step": 26500
    },
    {
      "entropy": 1.9659972831606864,
      "epoch": 0.0821786805067427,
      "grad_norm": 9.854397773742676,
      "learning_rate": 8.825133604529864e-06,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8131707355380058,
      "num_tokens": 31808978.0,
      "step": 26510
    },
    {
      "entropy": 1.9501794785261155,
      "epoch": 0.08220967963179239,
      "grad_norm": 8.959293365478516,
      "learning_rate": 8.823469521094459e-06,
      "loss": 0.6272,
      "mean_token_accuracy": 0.8212861150503159,
      "num_tokens": 31821515.0,
      "step": 26520
    },
    {
      "entropy": 1.9314236760139465,
      "epoch": 0.08224067875684209,
      "grad_norm": 10.00307846069336,
      "learning_rate": 8.821806378652095e-06,
      "loss": 0.5826,
      "mean_token_accuracy": 0.8202818527817726,
      "num_tokens": 31832945.0,
      "step": 26530
    },
    {
      "entropy": 1.9181413248181343,
      "epoch": 0.08227167788189178,
      "grad_norm": 12.737542152404785,
      "learning_rate": 8.820144176316263e-06,
      "loss": 0.6233,
      "mean_token_accuracy": 0.808757683634758,
      "num_tokens": 31844982.0,
      "step": 26540
    },
    {
      "entropy": 1.9316234186291694,
      "epoch": 0.08230267700694148,
      "grad_norm": 4.668799877166748,
      "learning_rate": 8.818482913201624e-06,
      "loss": 0.5973,
      "mean_token_accuracy": 0.81618000715971,
      "num_tokens": 31857158.0,
      "step": 26550
    },
    {
      "entropy": 1.9605771273374557,
      "epoch": 0.08233367613199118,
      "grad_norm": 10.899618148803711,
      "learning_rate": 8.816822588424007e-06,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8206842705607414,
      "num_tokens": 31868956.0,
      "step": 26560
    },
    {
      "entropy": 1.9107506170868873,
      "epoch": 0.08236467525704087,
      "grad_norm": 10.741750717163086,
      "learning_rate": 8.815163201100404e-06,
      "loss": 0.564,
      "mean_token_accuracy": 0.8113860443234444,
      "num_tokens": 31881384.0,
      "step": 26570
    },
    {
      "entropy": 1.940708489716053,
      "epoch": 0.08239567438209057,
      "grad_norm": 10.441813468933105,
      "learning_rate": 8.813504750348967e-06,
      "loss": 0.584,
      "mean_token_accuracy": 0.8169033840298653,
      "num_tokens": 31892982.0,
      "step": 26580
    },
    {
      "entropy": 1.9362832650542259,
      "epoch": 0.08242667350714027,
      "grad_norm": 9.666556358337402,
      "learning_rate": 8.811847235289013e-06,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8026597559452057,
      "num_tokens": 31905412.0,
      "step": 26590
    },
    {
      "entropy": 1.9204092472791672,
      "epoch": 0.08245767263218996,
      "grad_norm": 8.493167877197266,
      "learning_rate": 8.81019065504102e-06,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8150411576032639,
      "num_tokens": 31917332.0,
      "step": 26600
    },
    {
      "entropy": 1.9749287247657776,
      "epoch": 0.08248867175723965,
      "grad_norm": 9.949286460876465,
      "learning_rate": 8.808535008726616e-06,
      "loss": 0.6233,
      "mean_token_accuracy": 0.8155123621225357,
      "num_tokens": 31928762.0,
      "step": 26610
    },
    {
      "entropy": 2.0281356394290926,
      "epoch": 0.08251967088228934,
      "grad_norm": 8.458736419677734,
      "learning_rate": 8.806880295468594e-06,
      "loss": 0.6129,
      "mean_token_accuracy": 0.817957803606987,
      "num_tokens": 31939691.0,
      "step": 26620
    },
    {
      "entropy": 1.9947254791855813,
      "epoch": 0.08255067000733904,
      "grad_norm": 10.101909637451172,
      "learning_rate": 8.805226514390884e-06,
      "loss": 0.5895,
      "mean_token_accuracy": 0.8237784549593925,
      "num_tokens": 31950789.0,
      "step": 26630
    },
    {
      "entropy": 1.9266445934772491,
      "epoch": 0.08258166913238874,
      "grad_norm": 9.334656715393066,
      "learning_rate": 8.803573664618587e-06,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8326213672757149,
      "num_tokens": 31962662.0,
      "step": 26640
    },
    {
      "entropy": 1.9811443358659744,
      "epoch": 0.08261266825743843,
      "grad_norm": 9.728004455566406,
      "learning_rate": 8.801921745277938e-06,
      "loss": 0.6228,
      "mean_token_accuracy": 0.8104926988482475,
      "num_tokens": 31973878.0,
      "step": 26650
    },
    {
      "entropy": 1.9071193888783455,
      "epoch": 0.08264366738248813,
      "grad_norm": 9.647929191589355,
      "learning_rate": 8.800270755496327e-06,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8240789666771888,
      "num_tokens": 31985601.0,
      "step": 26660
    },
    {
      "entropy": 1.8988082259893417,
      "epoch": 0.08267466650753783,
      "grad_norm": 5.0942888259887695,
      "learning_rate": 8.798620694402286e-06,
      "loss": 0.5738,
      "mean_token_accuracy": 0.819773106276989,
      "num_tokens": 31999208.0,
      "step": 26670
    },
    {
      "entropy": 1.9712405875325203,
      "epoch": 0.08270566563258752,
      "grad_norm": 8.369913101196289,
      "learning_rate": 8.796971561125492e-06,
      "loss": 0.6086,
      "mean_token_accuracy": 0.8169108390808105,
      "num_tokens": 32010782.0,
      "step": 26680
    },
    {
      "entropy": 1.868271279335022,
      "epoch": 0.08273666475763722,
      "grad_norm": 8.779671669006348,
      "learning_rate": 8.795323354796762e-06,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8314287766814232,
      "num_tokens": 32023787.0,
      "step": 26690
    },
    {
      "entropy": 1.9078305020928383,
      "epoch": 0.08276766388268691,
      "grad_norm": 9.038331985473633,
      "learning_rate": 8.79367607454805e-06,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8266749441623688,
      "num_tokens": 32036579.0,
      "step": 26700
    },
    {
      "entropy": 1.9659512534737587,
      "epoch": 0.08279866300773661,
      "grad_norm": 10.663753509521484,
      "learning_rate": 8.792029719512458e-06,
      "loss": 0.5953,
      "mean_token_accuracy": 0.8178058177232742,
      "num_tokens": 32048734.0,
      "step": 26710
    },
    {
      "entropy": 1.8505152672529221,
      "epoch": 0.08282966213278631,
      "grad_norm": 3.9875552654266357,
      "learning_rate": 8.79038428882421e-06,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8262492626905441,
      "num_tokens": 32061882.0,
      "step": 26720
    },
    {
      "entropy": 1.913353630900383,
      "epoch": 0.08286066125783599,
      "grad_norm": 10.27352523803711,
      "learning_rate": 8.788739781618678e-06,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8218828424811363,
      "num_tokens": 32074037.0,
      "step": 26730
    },
    {
      "entropy": 1.9119721353054047,
      "epoch": 0.08289166038288569,
      "grad_norm": 9.0624361038208,
      "learning_rate": 8.78709619703235e-06,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8198665246367455,
      "num_tokens": 32086164.0,
      "step": 26740
    },
    {
      "entropy": 1.8503568902611733,
      "epoch": 0.08292265950793538,
      "grad_norm": 5.034000396728516,
      "learning_rate": 8.785453534202857e-06,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8217618718743325,
      "num_tokens": 32099236.0,
      "step": 26750
    },
    {
      "entropy": 1.9170492082834243,
      "epoch": 0.08295365863298508,
      "grad_norm": 11.386276245117188,
      "learning_rate": 8.78381179226895e-06,
      "loss": 0.5962,
      "mean_token_accuracy": 0.8131098374724388,
      "num_tokens": 32111036.0,
      "step": 26760
    },
    {
      "entropy": 2.001404981315136,
      "epoch": 0.08298465775803478,
      "grad_norm": 9.401607513427734,
      "learning_rate": 8.782170970370514e-06,
      "loss": 0.6513,
      "mean_token_accuracy": 0.8121735200285911,
      "num_tokens": 32122745.0,
      "step": 26770
    },
    {
      "entropy": 1.9728690326213836,
      "epoch": 0.08301565688308447,
      "grad_norm": 9.115731239318848,
      "learning_rate": 8.78053106764855e-06,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8155710816383361,
      "num_tokens": 32133822.0,
      "step": 26780
    },
    {
      "entropy": 1.952053852379322,
      "epoch": 0.08304665600813417,
      "grad_norm": 9.419805526733398,
      "learning_rate": 8.778892083245187e-06,
      "loss": 0.6176,
      "mean_token_accuracy": 0.814921198785305,
      "num_tokens": 32144770.0,
      "step": 26790
    },
    {
      "entropy": 1.9340526655316352,
      "epoch": 0.08307765513318387,
      "grad_norm": 4.676142692565918,
      "learning_rate": 8.77725401630367e-06,
      "loss": 0.6304,
      "mean_token_accuracy": 0.7994276747107506,
      "num_tokens": 32156515.0,
      "step": 26800
    },
    {
      "entropy": 2.002475252747536,
      "epoch": 0.08310865425823356,
      "grad_norm": 5.354742527008057,
      "learning_rate": 8.775616865968369e-06,
      "loss": 0.7259,
      "mean_token_accuracy": 0.7925671577453614,
      "num_tokens": 32167752.0,
      "step": 26810
    },
    {
      "entropy": 1.8802199259400367,
      "epoch": 0.08313965338328326,
      "grad_norm": 8.681201934814453,
      "learning_rate": 8.773980631384764e-06,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8213910296559334,
      "num_tokens": 32181313.0,
      "step": 26820
    },
    {
      "entropy": 1.873046538233757,
      "epoch": 0.08317065250833296,
      "grad_norm": 8.961077690124512,
      "learning_rate": 8.772345311699455e-06,
      "loss": 0.558,
      "mean_token_accuracy": 0.8188267648220062,
      "num_tokens": 32194510.0,
      "step": 26830
    },
    {
      "entropy": 1.8043782696127892,
      "epoch": 0.08320165163338265,
      "grad_norm": 10.20888614654541,
      "learning_rate": 8.770710906060152e-06,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8352167114615441,
      "num_tokens": 32207477.0,
      "step": 26840
    },
    {
      "entropy": 1.962397077679634,
      "epoch": 0.08323265075843235,
      "grad_norm": 8.981500625610352,
      "learning_rate": 8.769077413615676e-06,
      "loss": 0.6166,
      "mean_token_accuracy": 0.8194408491253853,
      "num_tokens": 32218668.0,
      "step": 26850
    },
    {
      "entropy": 1.9099099516868592,
      "epoch": 0.08326364988348203,
      "grad_norm": 9.33209228515625,
      "learning_rate": 8.76744483351596e-06,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8329472854733467,
      "num_tokens": 32230093.0,
      "step": 26860
    },
    {
      "entropy": 1.8509590789675712,
      "epoch": 0.08329464900853173,
      "grad_norm": 10.720943450927734,
      "learning_rate": 8.76581316491204e-06,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8222440704703331,
      "num_tokens": 32242430.0,
      "step": 26870
    },
    {
      "entropy": 1.7675331860780716,
      "epoch": 0.08332564813358143,
      "grad_norm": 5.470025539398193,
      "learning_rate": 8.764182406956064e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8318305298686027,
      "num_tokens": 32255794.0,
      "step": 26880
    },
    {
      "entropy": 1.8293626666069032,
      "epoch": 0.08335664725863112,
      "grad_norm": 6.280379772186279,
      "learning_rate": 8.762552558801276e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8191826656460762,
      "num_tokens": 32269026.0,
      "step": 26890
    },
    {
      "entropy": 1.8529397904872895,
      "epoch": 0.08338764638368082,
      "grad_norm": 9.650927543640137,
      "learning_rate": 8.760923619602028e-06,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8250053748488426,
      "num_tokens": 32282150.0,
      "step": 26900
    },
    {
      "entropy": 1.8897050678730012,
      "epoch": 0.08341864550873052,
      "grad_norm": 4.701704978942871,
      "learning_rate": 8.75929558851377e-06,
      "loss": 0.564,
      "mean_token_accuracy": 0.8231826663017273,
      "num_tokens": 32293837.0,
      "step": 26910
    },
    {
      "entropy": 1.8501434102654457,
      "epoch": 0.08344964463378021,
      "grad_norm": 10.15555191040039,
      "learning_rate": 8.757668464693049e-06,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8285914018750191,
      "num_tokens": 32306371.0,
      "step": 26920
    },
    {
      "entropy": 1.9136781513690948,
      "epoch": 0.08348064375882991,
      "grad_norm": 9.14686107635498,
      "learning_rate": 8.756042247297512e-06,
      "loss": 0.5924,
      "mean_token_accuracy": 0.8250029042363167,
      "num_tokens": 32318441.0,
      "step": 26930
    },
    {
      "entropy": 1.9553735256195068,
      "epoch": 0.0835116428838796,
      "grad_norm": 4.363174915313721,
      "learning_rate": 8.754416935485893e-06,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8168588444590569,
      "num_tokens": 32329443.0,
      "step": 26940
    },
    {
      "entropy": 1.839969304203987,
      "epoch": 0.0835426420089293,
      "grad_norm": 9.89714241027832,
      "learning_rate": 8.75279252841803e-06,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8351193264126777,
      "num_tokens": 32341869.0,
      "step": 26950
    },
    {
      "entropy": 1.9383182168006896,
      "epoch": 0.083573641133979,
      "grad_norm": 10.119979858398438,
      "learning_rate": 8.751169025254838e-06,
      "loss": 0.593,
      "mean_token_accuracy": 0.8233784183859825,
      "num_tokens": 32352867.0,
      "step": 26960
    },
    {
      "entropy": 1.8974655985832214,
      "epoch": 0.0836046402590287,
      "grad_norm": 9.094542503356934,
      "learning_rate": 8.749546425158334e-06,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8213450491428376,
      "num_tokens": 32365021.0,
      "step": 26970
    },
    {
      "entropy": 1.8861443296074867,
      "epoch": 0.08363563938407838,
      "grad_norm": 9.384647369384766,
      "learning_rate": 8.747924727291615e-06,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8222412168979645,
      "num_tokens": 32376467.0,
      "step": 26980
    },
    {
      "entropy": 1.9021775022149086,
      "epoch": 0.08366663850912807,
      "grad_norm": 8.179243087768555,
      "learning_rate": 8.746303930818864e-06,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8159692242741585,
      "num_tokens": 32388816.0,
      "step": 26990
    },
    {
      "entropy": 1.9401640132069589,
      "epoch": 0.08369763763417777,
      "grad_norm": 4.109061241149902,
      "learning_rate": 8.744684034905353e-06,
      "loss": 0.5723,
      "mean_token_accuracy": 0.8248335152864457,
      "num_tokens": 32400534.0,
      "step": 27000
    },
    {
      "entropy": 1.8398920968174934,
      "epoch": 0.08372863675922747,
      "grad_norm": 10.484107971191406,
      "learning_rate": 8.743065038717426e-06,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8244616940617562,
      "num_tokens": 32413551.0,
      "step": 27010
    },
    {
      "entropy": 1.9067912593483924,
      "epoch": 0.08375963588427716,
      "grad_norm": 4.866901397705078,
      "learning_rate": 8.741446941422514e-06,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8243859261274338,
      "num_tokens": 32425381.0,
      "step": 27020
    },
    {
      "entropy": 1.839262606203556,
      "epoch": 0.08379063500932686,
      "grad_norm": 11.031305313110352,
      "learning_rate": 8.739829742189128e-06,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8227956235408783,
      "num_tokens": 32438069.0,
      "step": 27030
    },
    {
      "entropy": 1.8895979836583137,
      "epoch": 0.08382163413437656,
      "grad_norm": 8.922094345092773,
      "learning_rate": 8.738213440186849e-06,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8271268501877784,
      "num_tokens": 32449380.0,
      "step": 27040
    },
    {
      "entropy": 1.945242629945278,
      "epoch": 0.08385263325942625,
      "grad_norm": 4.774547576904297,
      "learning_rate": 8.736598034586335e-06,
      "loss": 0.609,
      "mean_token_accuracy": 0.8129522934556007,
      "num_tokens": 32460665.0,
      "step": 27050
    },
    {
      "entropy": 1.9358567342162132,
      "epoch": 0.08388363238447595,
      "grad_norm": 5.142744064331055,
      "learning_rate": 8.734983524559322e-06,
      "loss": 0.6,
      "mean_token_accuracy": 0.8112214103341102,
      "num_tokens": 32472553.0,
      "step": 27060
    },
    {
      "entropy": 1.909812480211258,
      "epoch": 0.08391463150952565,
      "grad_norm": 8.57243824005127,
      "learning_rate": 8.733369909278609e-06,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8224716767668724,
      "num_tokens": 32485563.0,
      "step": 27070
    },
    {
      "entropy": 2.0142262816429137,
      "epoch": 0.08394563063457534,
      "grad_norm": 10.417340278625488,
      "learning_rate": 8.731757187918067e-06,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8098721027374267,
      "num_tokens": 32496322.0,
      "step": 27080
    },
    {
      "entropy": 1.8523387983441353,
      "epoch": 0.08397662975962504,
      "grad_norm": 4.585434436798096,
      "learning_rate": 8.730145359652638e-06,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8351857841014863,
      "num_tokens": 32509323.0,
      "step": 27090
    },
    {
      "entropy": 1.8732295900583267,
      "epoch": 0.08400762888467472,
      "grad_norm": 4.835305690765381,
      "learning_rate": 8.728534423658325e-06,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8177412390708924,
      "num_tokens": 32521503.0,
      "step": 27100
    },
    {
      "entropy": 2.0055402636528017,
      "epoch": 0.08403862800972442,
      "grad_norm": 13.625639915466309,
      "learning_rate": 8.726924379112201e-06,
      "loss": 0.6815,
      "mean_token_accuracy": 0.8050232946872711,
      "num_tokens": 32532199.0,
      "step": 27110
    },
    {
      "entropy": 1.8963544055819512,
      "epoch": 0.08406962713477412,
      "grad_norm": 9.827751159667969,
      "learning_rate": 8.725315225192391e-06,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8158974751830101,
      "num_tokens": 32544370.0,
      "step": 27120
    },
    {
      "entropy": 1.957877866923809,
      "epoch": 0.08410062625982381,
      "grad_norm": 4.588069915771484,
      "learning_rate": 8.723706961078094e-06,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8149573966860771,
      "num_tokens": 32555644.0,
      "step": 27130
    },
    {
      "entropy": 1.9230869844555856,
      "epoch": 0.08413162538487351,
      "grad_norm": 8.736885070800781,
      "learning_rate": 8.722099585949552e-06,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8175855800509453,
      "num_tokens": 32568234.0,
      "step": 27140
    },
    {
      "entropy": 1.9423416420817374,
      "epoch": 0.0841626245099232,
      "grad_norm": 10.53701400756836,
      "learning_rate": 8.720493098988078e-06,
      "loss": 0.6455,
      "mean_token_accuracy": 0.8082656994462013,
      "num_tokens": 32580063.0,
      "step": 27150
    },
    {
      "entropy": 2.0061039671301844,
      "epoch": 0.0841936236349729,
      "grad_norm": 9.278496742248535,
      "learning_rate": 8.718887499376033e-06,
      "loss": 0.6392,
      "mean_token_accuracy": 0.8064600452780724,
      "num_tokens": 32591372.0,
      "step": 27160
    },
    {
      "entropy": 1.9022212833166123,
      "epoch": 0.0842246227600226,
      "grad_norm": 4.704682350158691,
      "learning_rate": 8.717282786296834e-06,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8284071788191796,
      "num_tokens": 32604362.0,
      "step": 27170
    },
    {
      "entropy": 1.858892096579075,
      "epoch": 0.0842556218850723,
      "grad_norm": 10.126176834106445,
      "learning_rate": 8.715678958934944e-06,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8314953356981277,
      "num_tokens": 32616980.0,
      "step": 27180
    },
    {
      "entropy": 1.8593590274453162,
      "epoch": 0.08428662101012199,
      "grad_norm": 10.302115440368652,
      "learning_rate": 8.714076016475885e-06,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8241918250918389,
      "num_tokens": 32629349.0,
      "step": 27190
    },
    {
      "entropy": 1.8986007198691368,
      "epoch": 0.08431762013517169,
      "grad_norm": 3.36916446685791,
      "learning_rate": 8.712473958106222e-06,
      "loss": 0.5957,
      "mean_token_accuracy": 0.8168155118823052,
      "num_tokens": 32641755.0,
      "step": 27200
    },
    {
      "entropy": 1.8978888988494873,
      "epoch": 0.08434861926022139,
      "grad_norm": 9.608877182006836,
      "learning_rate": 8.710872783013563e-06,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8153728187084198,
      "num_tokens": 32653282.0,
      "step": 27210
    },
    {
      "entropy": 1.9542479366064072,
      "epoch": 0.08437961838527108,
      "grad_norm": 4.4257307052612305,
      "learning_rate": 8.709272490386569e-06,
      "loss": 0.6091,
      "mean_token_accuracy": 0.806995865702629,
      "num_tokens": 32664764.0,
      "step": 27220
    },
    {
      "entropy": 1.9367131859064102,
      "epoch": 0.08441061751032077,
      "grad_norm": 8.415755271911621,
      "learning_rate": 8.707673079414937e-06,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8194612860679626,
      "num_tokens": 32675941.0,
      "step": 27230
    },
    {
      "entropy": 1.9739065006375314,
      "epoch": 0.08444161663537046,
      "grad_norm": 9.644390106201172,
      "learning_rate": 8.706074549289411e-06,
      "loss": 0.6278,
      "mean_token_accuracy": 0.8121867910027504,
      "num_tokens": 32687272.0,
      "step": 27240
    },
    {
      "entropy": 1.9072382494807243,
      "epoch": 0.08447261576042016,
      "grad_norm": 8.06834888458252,
      "learning_rate": 8.704476899201766e-06,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8260136678814888,
      "num_tokens": 32698817.0,
      "step": 27250
    },
    {
      "entropy": 1.9281651645898819,
      "epoch": 0.08450361488546985,
      "grad_norm": 10.1871976852417,
      "learning_rate": 8.702880128344827e-06,
      "loss": 0.5762,
      "mean_token_accuracy": 0.82305498868227,
      "num_tokens": 32710416.0,
      "step": 27260
    },
    {
      "entropy": 1.9203084349632262,
      "epoch": 0.08453461401051955,
      "grad_norm": 11.28444766998291,
      "learning_rate": 8.701284235912444e-06,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8235008075833321,
      "num_tokens": 32722557.0,
      "step": 27270
    },
    {
      "entropy": 1.8537453413009644,
      "epoch": 0.08456561313556925,
      "grad_norm": 10.914976119995117,
      "learning_rate": 8.699689221099508e-06,
      "loss": 0.6454,
      "mean_token_accuracy": 0.8067585110664368,
      "num_tokens": 32735577.0,
      "step": 27280
    },
    {
      "entropy": 1.9035761684179306,
      "epoch": 0.08459661226061894,
      "grad_norm": 11.605010986328125,
      "learning_rate": 8.698095083101939e-06,
      "loss": 0.5898,
      "mean_token_accuracy": 0.8235433489084244,
      "num_tokens": 32747673.0,
      "step": 27290
    },
    {
      "entropy": 1.8419725641608238,
      "epoch": 0.08462761138566864,
      "grad_norm": 7.603212833404541,
      "learning_rate": 8.69650182111669e-06,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8381339445710182,
      "num_tokens": 32760397.0,
      "step": 27300
    },
    {
      "entropy": 1.9057568103075027,
      "epoch": 0.08465861051071834,
      "grad_norm": 7.779855251312256,
      "learning_rate": 8.694909434341743e-06,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8320439457893372,
      "num_tokens": 32772961.0,
      "step": 27310
    },
    {
      "entropy": 1.9533362567424775,
      "epoch": 0.08468960963576803,
      "grad_norm": 9.11958122253418,
      "learning_rate": 8.693317921976107e-06,
      "loss": 0.6301,
      "mean_token_accuracy": 0.8056516513228417,
      "num_tokens": 32785213.0,
      "step": 27320
    },
    {
      "entropy": 1.9149364829063416,
      "epoch": 0.08472060876081773,
      "grad_norm": 8.80299186706543,
      "learning_rate": 8.69172728321982e-06,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8212979912757874,
      "num_tokens": 32798167.0,
      "step": 27330
    },
    {
      "entropy": 1.8609929516911508,
      "epoch": 0.08475160788586743,
      "grad_norm": 4.60377311706543,
      "learning_rate": 8.690137517273937e-06,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8336883813142777,
      "num_tokens": 32810453.0,
      "step": 27340
    },
    {
      "entropy": 1.9059410750865937,
      "epoch": 0.08478260701091711,
      "grad_norm": 10.732145309448242,
      "learning_rate": 8.688548623340543e-06,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8180224969983101,
      "num_tokens": 32823220.0,
      "step": 27350
    },
    {
      "entropy": 1.9536007195711136,
      "epoch": 0.08481360613596681,
      "grad_norm": 8.611103057861328,
      "learning_rate": 8.68696060062274e-06,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8148555681109428,
      "num_tokens": 32834092.0,
      "step": 27360
    },
    {
      "entropy": 1.8823547706007957,
      "epoch": 0.0848446052610165,
      "grad_norm": 8.003419876098633,
      "learning_rate": 8.685373448324655e-06,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8131691709160804,
      "num_tokens": 32846184.0,
      "step": 27370
    },
    {
      "entropy": 1.8589939393103123,
      "epoch": 0.0848756043860662,
      "grad_norm": 9.513988494873047,
      "learning_rate": 8.683787165651419e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8292522057890892,
      "num_tokens": 32859312.0,
      "step": 27380
    },
    {
      "entropy": 1.9123133316636085,
      "epoch": 0.0849066035111159,
      "grad_norm": 9.80207633972168,
      "learning_rate": 8.682201751809196e-06,
      "loss": 0.6113,
      "mean_token_accuracy": 0.8205790624022484,
      "num_tokens": 32870081.0,
      "step": 27390
    },
    {
      "entropy": 1.8975965559482575,
      "epoch": 0.0849376026361656,
      "grad_norm": 9.128314971923828,
      "learning_rate": 8.680617206005148e-06,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8289038851857186,
      "num_tokens": 32881096.0,
      "step": 27400
    },
    {
      "entropy": 1.9172597080469131,
      "epoch": 0.08496860176121529,
      "grad_norm": 10.668315887451172,
      "learning_rate": 8.679033527447462e-06,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8238224640488625,
      "num_tokens": 32893038.0,
      "step": 27410
    },
    {
      "entropy": 1.8307318970561028,
      "epoch": 0.08499960088626499,
      "grad_norm": 4.935861587524414,
      "learning_rate": 8.67745071534533e-06,
      "loss": 0.531,
      "mean_token_accuracy": 0.8222571074962616,
      "num_tokens": 32906388.0,
      "step": 27420
    },
    {
      "entropy": 1.9008280768990518,
      "epoch": 0.08503060001131468,
      "grad_norm": 4.503681182861328,
      "learning_rate": 8.675868768908956e-06,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8253033936023713,
      "num_tokens": 32918054.0,
      "step": 27430
    },
    {
      "entropy": 1.8841553077101707,
      "epoch": 0.08506159913636438,
      "grad_norm": 8.340595245361328,
      "learning_rate": 8.674287687349546e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8274075031280518,
      "num_tokens": 32929483.0,
      "step": 27440
    },
    {
      "entropy": 1.8089099466800689,
      "epoch": 0.08509259826141408,
      "grad_norm": 4.920588493347168,
      "learning_rate": 8.672707469879315e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8309017539024353,
      "num_tokens": 32942288.0,
      "step": 27450
    },
    {
      "entropy": 1.8523821212351321,
      "epoch": 0.08512359738646377,
      "grad_norm": 10.191000938415527,
      "learning_rate": 8.67112811571149e-06,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8139420121908187,
      "num_tokens": 32954595.0,
      "step": 27460
    },
    {
      "entropy": 1.878308503329754,
      "epoch": 0.08515459651151346,
      "grad_norm": 5.282871246337891,
      "learning_rate": 8.669549624060287e-06,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8135267734527588,
      "num_tokens": 32966387.0,
      "step": 27470
    },
    {
      "entropy": 1.8660302713513375,
      "epoch": 0.08518559563656315,
      "grad_norm": 7.17582893371582,
      "learning_rate": 8.66797199414093e-06,
      "loss": 0.6105,
      "mean_token_accuracy": 0.8228353872895241,
      "num_tokens": 32978739.0,
      "step": 27480
    },
    {
      "entropy": 1.855283497273922,
      "epoch": 0.08521659476161285,
      "grad_norm": 9.164436340332031,
      "learning_rate": 8.666395225169643e-06,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8205430552363395,
      "num_tokens": 32991000.0,
      "step": 27490
    },
    {
      "entropy": 1.916937005519867,
      "epoch": 0.08524759388666255,
      "grad_norm": 5.297219276428223,
      "learning_rate": 8.664819316363645e-06,
      "loss": 0.6027,
      "mean_token_accuracy": 0.8212128937244415,
      "num_tokens": 33002576.0,
      "step": 27500
    },
    {
      "entropy": 1.9633919060230256,
      "epoch": 0.08527859301171224,
      "grad_norm": 10.939813613891602,
      "learning_rate": 8.663244266941157e-06,
      "loss": 0.6718,
      "mean_token_accuracy": 0.8004844337701797,
      "num_tokens": 33013283.0,
      "step": 27510
    },
    {
      "entropy": 1.8730249777436256,
      "epoch": 0.08530959213676194,
      "grad_norm": 5.108644962310791,
      "learning_rate": 8.661670076121382e-06,
      "loss": 0.6221,
      "mean_token_accuracy": 0.810503962635994,
      "num_tokens": 33025788.0,
      "step": 27520
    },
    {
      "entropy": 1.9476985320448876,
      "epoch": 0.08534059126181164,
      "grad_norm": 9.54196548461914,
      "learning_rate": 8.66009674312453e-06,
      "loss": 0.6355,
      "mean_token_accuracy": 0.800178411602974,
      "num_tokens": 33036893.0,
      "step": 27530
    },
    {
      "entropy": 1.9610710114240646,
      "epoch": 0.08537159038686133,
      "grad_norm": 8.600343704223633,
      "learning_rate": 8.658524267171792e-06,
      "loss": 0.6988,
      "mean_token_accuracy": 0.8083438903093338,
      "num_tokens": 33048666.0,
      "step": 27540
    },
    {
      "entropy": 1.8692133530974389,
      "epoch": 0.08540258951191103,
      "grad_norm": 10.251091957092285,
      "learning_rate": 8.656952647485353e-06,
      "loss": 0.561,
      "mean_token_accuracy": 0.8302003368735313,
      "num_tokens": 33061005.0,
      "step": 27550
    },
    {
      "entropy": 1.827210921049118,
      "epoch": 0.08543358863696073,
      "grad_norm": 9.993083000183105,
      "learning_rate": 8.655381883288387e-06,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8159284323453904,
      "num_tokens": 33074475.0,
      "step": 27560
    },
    {
      "entropy": 1.8685495540499688,
      "epoch": 0.08546458776201042,
      "grad_norm": 10.750574111938477,
      "learning_rate": 8.653811973805048e-06,
      "loss": 0.5993,
      "mean_token_accuracy": 0.8224136650562286,
      "num_tokens": 33086089.0,
      "step": 27570
    },
    {
      "entropy": 1.7942854017019272,
      "epoch": 0.08549558688706012,
      "grad_norm": 4.533649921417236,
      "learning_rate": 8.652242918260485e-06,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8248818814754486,
      "num_tokens": 33100203.0,
      "step": 27580
    },
    {
      "entropy": 1.8375580973923207,
      "epoch": 0.08552658601210981,
      "grad_norm": 4.471881866455078,
      "learning_rate": 8.650674715880821e-06,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8313728928565979,
      "num_tokens": 33113514.0,
      "step": 27590
    },
    {
      "entropy": 1.9055291563272476,
      "epoch": 0.0855575851371595,
      "grad_norm": 8.167454719543457,
      "learning_rate": 8.649107365893162e-06,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8212564826011658,
      "num_tokens": 33125377.0,
      "step": 27600
    },
    {
      "entropy": 1.882322046160698,
      "epoch": 0.0855885842622092,
      "grad_norm": 9.761778831481934,
      "learning_rate": 8.647540867525599e-06,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8162475794553756,
      "num_tokens": 33137354.0,
      "step": 27610
    },
    {
      "entropy": 1.8892990604043007,
      "epoch": 0.08561958338725889,
      "grad_norm": 9.397979736328125,
      "learning_rate": 8.645975220007197e-06,
      "loss": 0.6297,
      "mean_token_accuracy": 0.8177220180630684,
      "num_tokens": 33149211.0,
      "step": 27620
    },
    {
      "entropy": 1.9387479245662689,
      "epoch": 0.08565058251230859,
      "grad_norm": 10.300222396850586,
      "learning_rate": 8.644410422567995e-06,
      "loss": 0.6018,
      "mean_token_accuracy": 0.8117999002337456,
      "num_tokens": 33160821.0,
      "step": 27630
    },
    {
      "entropy": 1.942291297018528,
      "epoch": 0.08568158163735828,
      "grad_norm": 11.509066581726074,
      "learning_rate": 8.642846474439016e-06,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8200451254844665,
      "num_tokens": 33172491.0,
      "step": 27640
    },
    {
      "entropy": 1.954495507478714,
      "epoch": 0.08571258076240798,
      "grad_norm": 8.897340774536133,
      "learning_rate": 8.641283374852245e-06,
      "loss": 0.5856,
      "mean_token_accuracy": 0.8219591468572617,
      "num_tokens": 33184078.0,
      "step": 27650
    },
    {
      "entropy": 1.8060628071427345,
      "epoch": 0.08574357988745768,
      "grad_norm": 8.12181282043457,
      "learning_rate": 8.639721123040653e-06,
      "loss": 0.49,
      "mean_token_accuracy": 0.8268592938780784,
      "num_tokens": 33198036.0,
      "step": 27660
    },
    {
      "entropy": 1.7999069422483445,
      "epoch": 0.08577457901250737,
      "grad_norm": 4.471950054168701,
      "learning_rate": 8.638159718238167e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8397438317537308,
      "num_tokens": 33212045.0,
      "step": 27670
    },
    {
      "entropy": 1.9364047437906264,
      "epoch": 0.08580557813755707,
      "grad_norm": 13.346952438354492,
      "learning_rate": 8.636599159679694e-06,
      "loss": 0.6085,
      "mean_token_accuracy": 0.8085330918431282,
      "num_tokens": 33223752.0,
      "step": 27680
    },
    {
      "entropy": 1.8672024756669998,
      "epoch": 0.08583657726260677,
      "grad_norm": 5.810976028442383,
      "learning_rate": 8.635039446601096e-06,
      "loss": 0.5647,
      "mean_token_accuracy": 0.8230971753597259,
      "num_tokens": 33236114.0,
      "step": 27690
    },
    {
      "entropy": 1.8852705493569375,
      "epoch": 0.08586757638765646,
      "grad_norm": 5.571743011474609,
      "learning_rate": 8.633480578239217e-06,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8273131296038627,
      "num_tokens": 33248369.0,
      "step": 27700
    },
    {
      "entropy": 1.9336923122406007,
      "epoch": 0.08589857551270616,
      "grad_norm": 11.164706230163574,
      "learning_rate": 8.63192255383185e-06,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8112567231059075,
      "num_tokens": 33260442.0,
      "step": 27710
    },
    {
      "entropy": 1.9060131937265397,
      "epoch": 0.08592957463775584,
      "grad_norm": 10.689841270446777,
      "learning_rate": 8.630365372617761e-06,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8298930734395981,
      "num_tokens": 33272270.0,
      "step": 27720
    },
    {
      "entropy": 1.9362266033887863,
      "epoch": 0.08596057376280554,
      "grad_norm": 9.383460998535156,
      "learning_rate": 8.62880903383667e-06,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8200540199875832,
      "num_tokens": 33283981.0,
      "step": 27730
    },
    {
      "entropy": 1.9314474761486053,
      "epoch": 0.08599157288785524,
      "grad_norm": 10.130805969238281,
      "learning_rate": 8.627253536729257e-06,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8344287112355232,
      "num_tokens": 33296450.0,
      "step": 27740
    },
    {
      "entropy": 1.944339656829834,
      "epoch": 0.08602257201290493,
      "grad_norm": 8.83008861541748,
      "learning_rate": 8.625698880537165e-06,
      "loss": 0.6563,
      "mean_token_accuracy": 0.8038821011781693,
      "num_tokens": 33307571.0,
      "step": 27750
    },
    {
      "entropy": 1.9081848710775375,
      "epoch": 0.08605357113795463,
      "grad_norm": 10.480369567871094,
      "learning_rate": 8.62414506450299e-06,
      "loss": 0.5913,
      "mean_token_accuracy": 0.817870119214058,
      "num_tokens": 33319700.0,
      "step": 27760
    },
    {
      "entropy": 1.84724163711071,
      "epoch": 0.08608457026300433,
      "grad_norm": 3.8803551197052,
      "learning_rate": 8.622592087870282e-06,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8268263578414917,
      "num_tokens": 33331971.0,
      "step": 27770
    },
    {
      "entropy": 1.934218955039978,
      "epoch": 0.08611556938805402,
      "grad_norm": 9.977314949035645,
      "learning_rate": 8.621039949883543e-06,
      "loss": 0.6326,
      "mean_token_accuracy": 0.8152559250593185,
      "num_tokens": 33342967.0,
      "step": 27780
    },
    {
      "entropy": 1.8323631912469864,
      "epoch": 0.08614656851310372,
      "grad_norm": 12.091215133666992,
      "learning_rate": 8.619488649788232e-06,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8163023605942726,
      "num_tokens": 33356235.0,
      "step": 27790
    },
    {
      "entropy": 1.8991770133376122,
      "epoch": 0.08617756763815342,
      "grad_norm": 7.5659074783325195,
      "learning_rate": 8.617938186830752e-06,
      "loss": 0.5869,
      "mean_token_accuracy": 0.828110148012638,
      "num_tokens": 33368988.0,
      "step": 27800
    },
    {
      "entropy": 1.8398756310343742,
      "epoch": 0.08620856676320311,
      "grad_norm": 8.638795852661133,
      "learning_rate": 8.616388560258459e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8359197080135345,
      "num_tokens": 33381666.0,
      "step": 27810
    },
    {
      "entropy": 1.9264942169189454,
      "epoch": 0.08623956588825281,
      "grad_norm": 9.415760040283203,
      "learning_rate": 8.61483976931965e-06,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8242600187659264,
      "num_tokens": 33393125.0,
      "step": 27820
    },
    {
      "entropy": 1.841173852980137,
      "epoch": 0.0862705650133025,
      "grad_norm": 6.242795944213867,
      "learning_rate": 8.613291813263578e-06,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8402754053473472,
      "num_tokens": 33405715.0,
      "step": 27830
    },
    {
      "entropy": 1.9486567616462707,
      "epoch": 0.08630156413835219,
      "grad_norm": 8.530654907226562,
      "learning_rate": 8.61174469134043e-06,
      "loss": 0.621,
      "mean_token_accuracy": 0.8173292249441146,
      "num_tokens": 33417314.0,
      "step": 27840
    },
    {
      "entropy": 1.8448371395468712,
      "epoch": 0.08633256326340188,
      "grad_norm": 5.01255464553833,
      "learning_rate": 8.610198402801338e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8294866099953652,
      "num_tokens": 33431319.0,
      "step": 27850
    },
    {
      "entropy": 1.8705769084393977,
      "epoch": 0.08636356238845158,
      "grad_norm": 12.166780471801758,
      "learning_rate": 8.60865294689838e-06,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8117830008268356,
      "num_tokens": 33444973.0,
      "step": 27860
    },
    {
      "entropy": 1.9853614136576652,
      "epoch": 0.08639456151350128,
      "grad_norm": 9.717790603637695,
      "learning_rate": 8.607108322884566e-06,
      "loss": 0.61,
      "mean_token_accuracy": 0.8108811169862747,
      "num_tokens": 33456479.0,
      "step": 27870
    },
    {
      "entropy": 1.918572799861431,
      "epoch": 0.08642556063855097,
      "grad_norm": 4.174164295196533,
      "learning_rate": 8.605564530013847e-06,
      "loss": 0.5333,
      "mean_token_accuracy": 0.829804676771164,
      "num_tokens": 33468303.0,
      "step": 27880
    },
    {
      "entropy": 1.9406263917684554,
      "epoch": 0.08645655976360067,
      "grad_norm": 8.830784797668457,
      "learning_rate": 8.604021567541113e-06,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8110644370317459,
      "num_tokens": 33480915.0,
      "step": 27890
    },
    {
      "entropy": 1.9681098356842994,
      "epoch": 0.08648755888865037,
      "grad_norm": 9.740206718444824,
      "learning_rate": 8.602479434722184e-06,
      "loss": 0.623,
      "mean_token_accuracy": 0.8138311117887497,
      "num_tokens": 33491964.0,
      "step": 27900
    },
    {
      "entropy": 1.9589456513524055,
      "epoch": 0.08651855801370006,
      "grad_norm": 9.058309555053711,
      "learning_rate": 8.600938130813817e-06,
      "loss": 0.6245,
      "mean_token_accuracy": 0.814877088367939,
      "num_tokens": 33503440.0,
      "step": 27910
    },
    {
      "entropy": 1.787692840397358,
      "epoch": 0.08654955713874976,
      "grad_norm": 9.325605392456055,
      "learning_rate": 8.5993976550737e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8466595977544784,
      "num_tokens": 33517709.0,
      "step": 27920
    },
    {
      "entropy": 1.9136761993169784,
      "epoch": 0.08658055626379946,
      "grad_norm": 11.386701583862305,
      "learning_rate": 8.59785800676045e-06,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8248209357261658,
      "num_tokens": 33529539.0,
      "step": 27930
    },
    {
      "entropy": 1.8507066249847413,
      "epoch": 0.08661155538884915,
      "grad_norm": 9.556991577148438,
      "learning_rate": 8.596319185133614e-06,
      "loss": 0.521,
      "mean_token_accuracy": 0.8304374098777771,
      "num_tokens": 33542002.0,
      "step": 27940
    },
    {
      "entropy": 1.9005508705973626,
      "epoch": 0.08664255451389885,
      "grad_norm": 4.965198516845703,
      "learning_rate": 8.594781189453666e-06,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8208310827612877,
      "num_tokens": 33553517.0,
      "step": 27950
    },
    {
      "entropy": 1.9146973922848702,
      "epoch": 0.08667355363894855,
      "grad_norm": 4.9398932456970215,
      "learning_rate": 8.593244018982006e-06,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8191220477223397,
      "num_tokens": 33565833.0,
      "step": 27960
    },
    {
      "entropy": 1.8798900321125984,
      "epoch": 0.08670455276399823,
      "grad_norm": 8.898138046264648,
      "learning_rate": 8.59170767298096e-06,
      "loss": 0.6295,
      "mean_token_accuracy": 0.8155390247702599,
      "num_tokens": 33578565.0,
      "step": 27970
    },
    {
      "entropy": 1.8256605304777622,
      "epoch": 0.08673555188904793,
      "grad_norm": 9.652932167053223,
      "learning_rate": 8.590172150713773e-06,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8326058983802795,
      "num_tokens": 33591102.0,
      "step": 27980
    },
    {
      "entropy": 1.8813355445861817,
      "epoch": 0.08676655101409762,
      "grad_norm": 9.704791069030762,
      "learning_rate": 8.588637451444612e-06,
      "loss": 0.5626,
      "mean_token_accuracy": 0.8161522090435028,
      "num_tokens": 33603187.0,
      "step": 27990
    },
    {
      "entropy": 1.926676908135414,
      "epoch": 0.08679755013914732,
      "grad_norm": 10.068709373474121,
      "learning_rate": 8.587103574438569e-06,
      "loss": 0.644,
      "mean_token_accuracy": 0.8118028625845909,
      "num_tokens": 33614612.0,
      "step": 28000
    },
    {
      "entropy": 1.880669642984867,
      "epoch": 0.08682854926419702,
      "grad_norm": 8.847538948059082,
      "learning_rate": 8.585570518961651e-06,
      "loss": 0.5402,
      "mean_token_accuracy": 0.827701772749424,
      "num_tokens": 33626936.0,
      "step": 28010
    },
    {
      "entropy": 1.9508138015866279,
      "epoch": 0.08685954838924671,
      "grad_norm": 11.043745994567871,
      "learning_rate": 8.58403828428078e-06,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8156369537115097,
      "num_tokens": 33638555.0,
      "step": 28020
    },
    {
      "entropy": 1.9032750859856606,
      "epoch": 0.08689054751429641,
      "grad_norm": 10.193146705627441,
      "learning_rate": 8.582506869663793e-06,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8273995086550713,
      "num_tokens": 33650029.0,
      "step": 28030
    },
    {
      "entropy": 1.8772203177213669,
      "epoch": 0.0869215466393461,
      "grad_norm": 4.957283020019531,
      "learning_rate": 8.580976274379448e-06,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8220570385456085,
      "num_tokens": 33662687.0,
      "step": 28040
    },
    {
      "entropy": 2.005272647738457,
      "epoch": 0.0869525457643958,
      "grad_norm": 8.876986503601074,
      "learning_rate": 8.579446497697407e-06,
      "loss": 0.6347,
      "mean_token_accuracy": 0.809257960319519,
      "num_tokens": 33673133.0,
      "step": 28050
    },
    {
      "entropy": 1.897057008743286,
      "epoch": 0.0869835448894455,
      "grad_norm": 4.687375068664551,
      "learning_rate": 8.57791753888825e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8260800316929817,
      "num_tokens": 33685382.0,
      "step": 28060
    },
    {
      "entropy": 1.890799196064472,
      "epoch": 0.0870145440144952,
      "grad_norm": 11.013267517089844,
      "learning_rate": 8.576389397223463e-06,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8162667453289032,
      "num_tokens": 33698473.0,
      "step": 28070
    },
    {
      "entropy": 1.8947896853089332,
      "epoch": 0.08704554313954489,
      "grad_norm": 9.438980102539062,
      "learning_rate": 8.574862071975438e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.8390587836503982,
      "num_tokens": 33710540.0,
      "step": 28080
    },
    {
      "entropy": 1.833718155324459,
      "epoch": 0.08707654226459458,
      "grad_norm": 9.037769317626953,
      "learning_rate": 8.57333556241748e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8308623015880585,
      "num_tokens": 33723318.0,
      "step": 28090
    },
    {
      "entropy": 1.9327234029769897,
      "epoch": 0.08710754138964427,
      "grad_norm": 8.257326126098633,
      "learning_rate": 8.571809867823794e-06,
      "loss": 0.6263,
      "mean_token_accuracy": 0.8134735986590386,
      "num_tokens": 33734540.0,
      "step": 28100
    },
    {
      "entropy": 1.9399518385529517,
      "epoch": 0.08713854051469397,
      "grad_norm": 8.866934776306152,
      "learning_rate": 8.57028498746949e-06,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8291549518704414,
      "num_tokens": 33745641.0,
      "step": 28110
    },
    {
      "entropy": 1.8807589188218117,
      "epoch": 0.08716953963974367,
      "grad_norm": 9.743132591247559,
      "learning_rate": 8.568760920630582e-06,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8171397164463997,
      "num_tokens": 33757675.0,
      "step": 28120
    },
    {
      "entropy": 1.8112970426678658,
      "epoch": 0.08720053876479336,
      "grad_norm": 12.117599487304688,
      "learning_rate": 8.567237666583983e-06,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8401974871754646,
      "num_tokens": 33769971.0,
      "step": 28130
    },
    {
      "entropy": 1.8710318520665168,
      "epoch": 0.08723153788984306,
      "grad_norm": 10.758975982666016,
      "learning_rate": 8.565715224607507e-06,
      "loss": 0.5972,
      "mean_token_accuracy": 0.8087448701262474,
      "num_tokens": 33782332.0,
      "step": 28140
    },
    {
      "entropy": 1.9059083193540574,
      "epoch": 0.08726253701489275,
      "grad_norm": 9.423166275024414,
      "learning_rate": 8.564193593979863e-06,
      "loss": 0.566,
      "mean_token_accuracy": 0.8211286991834641,
      "num_tokens": 33794157.0,
      "step": 28150
    },
    {
      "entropy": 1.8057139664888382,
      "epoch": 0.08729353613994245,
      "grad_norm": 4.398457050323486,
      "learning_rate": 8.562672773980662e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8441807180643082,
      "num_tokens": 33806432.0,
      "step": 28160
    },
    {
      "entropy": 1.8262229442596436,
      "epoch": 0.08732453526499215,
      "grad_norm": 8.604774475097656,
      "learning_rate": 8.561152763890406e-06,
      "loss": 0.519,
      "mean_token_accuracy": 0.8298747554421425,
      "num_tokens": 33818953.0,
      "step": 28170
    },
    {
      "entropy": 1.856579264998436,
      "epoch": 0.08735553439004184,
      "grad_norm": 9.146932601928711,
      "learning_rate": 8.559633562990491e-06,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8260318920016289,
      "num_tokens": 33831302.0,
      "step": 28180
    },
    {
      "entropy": 1.9090891823172569,
      "epoch": 0.08738653351509154,
      "grad_norm": 10.917224884033203,
      "learning_rate": 8.558115170563206e-06,
      "loss": 0.6262,
      "mean_token_accuracy": 0.7991869211196899,
      "num_tokens": 33843087.0,
      "step": 28190
    },
    {
      "entropy": 1.8630158439278603,
      "epoch": 0.08741753264014124,
      "grad_norm": 8.141376495361328,
      "learning_rate": 8.556597585891731e-06,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8259263783693314,
      "num_tokens": 33855443.0,
      "step": 28200
    },
    {
      "entropy": 1.9804678469896317,
      "epoch": 0.08744853176519093,
      "grad_norm": 9.60744857788086,
      "learning_rate": 8.555080808260135e-06,
      "loss": 0.6753,
      "mean_token_accuracy": 0.8023278743028641,
      "num_tokens": 33866719.0,
      "step": 28210
    },
    {
      "entropy": 1.933399637043476,
      "epoch": 0.08747953089024062,
      "grad_norm": 7.662554740905762,
      "learning_rate": 8.55356483695338e-06,
      "loss": 0.5787,
      "mean_token_accuracy": 0.8204610332846641,
      "num_tokens": 33877680.0,
      "step": 28220
    },
    {
      "entropy": 1.936586219072342,
      "epoch": 0.08751053001529031,
      "grad_norm": 9.148261070251465,
      "learning_rate": 8.552049671257301e-06,
      "loss": 0.6002,
      "mean_token_accuracy": 0.810947285592556,
      "num_tokens": 33890103.0,
      "step": 28230
    },
    {
      "entropy": 1.9918580889701842,
      "epoch": 0.08754152914034001,
      "grad_norm": 9.620087623596191,
      "learning_rate": 8.550535310458639e-06,
      "loss": 0.6424,
      "mean_token_accuracy": 0.8145117297768593,
      "num_tokens": 33901085.0,
      "step": 28240
    },
    {
      "entropy": 1.9204299062490464,
      "epoch": 0.0875725282653897,
      "grad_norm": 9.934686660766602,
      "learning_rate": 8.549021753844996e-06,
      "loss": 0.5643,
      "mean_token_accuracy": 0.82352195084095,
      "num_tokens": 33913338.0,
      "step": 28250
    },
    {
      "entropy": 1.8327893808484077,
      "epoch": 0.0876035273904394,
      "grad_norm": 5.132232666015625,
      "learning_rate": 8.547509000704874e-06,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8444669589400291,
      "num_tokens": 33926686.0,
      "step": 28260
    },
    {
      "entropy": 1.8552611097693443,
      "epoch": 0.0876345265154891,
      "grad_norm": 10.373604774475098,
      "learning_rate": 8.545997050327648e-06,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8225721046328545,
      "num_tokens": 33939038.0,
      "step": 28270
    },
    {
      "entropy": 1.9178115367889403,
      "epoch": 0.0876655256405388,
      "grad_norm": 11.540125846862793,
      "learning_rate": 8.544485902003573e-06,
      "loss": 0.6322,
      "mean_token_accuracy": 0.8020992532372475,
      "num_tokens": 33950992.0,
      "step": 28280
    },
    {
      "entropy": 1.852571301162243,
      "epoch": 0.0876965247655885,
      "grad_norm": 10.074957847595215,
      "learning_rate": 8.542975555023786e-06,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8297949224710465,
      "num_tokens": 33963043.0,
      "step": 28290
    },
    {
      "entropy": 1.815762387216091,
      "epoch": 0.08772752389063819,
      "grad_norm": 5.080463886260986,
      "learning_rate": 8.541466008680297e-06,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8305084735155106,
      "num_tokens": 33976042.0,
      "step": 28300
    },
    {
      "entropy": 1.7797249928116798,
      "epoch": 0.08775852301568789,
      "grad_norm": 11.101983070373535,
      "learning_rate": 8.539957262265988e-06,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8253014713525773,
      "num_tokens": 33989173.0,
      "step": 28310
    },
    {
      "entropy": 1.9091511026024819,
      "epoch": 0.08778952214073758,
      "grad_norm": 6.8226141929626465,
      "learning_rate": 8.538449315074628e-06,
      "loss": 0.6012,
      "mean_token_accuracy": 0.8207783192396164,
      "num_tokens": 34000588.0,
      "step": 28320
    },
    {
      "entropy": 1.8104485064744948,
      "epoch": 0.08782052126578728,
      "grad_norm": 8.398052215576172,
      "learning_rate": 8.536942166400845e-06,
      "loss": 0.456,
      "mean_token_accuracy": 0.8447774976491929,
      "num_tokens": 34013462.0,
      "step": 28330
    },
    {
      "entropy": 1.8076317757368088,
      "epoch": 0.08785152039083696,
      "grad_norm": 8.194868087768555,
      "learning_rate": 8.535435815540142e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8241015315055847,
      "num_tokens": 34025542.0,
      "step": 28340
    },
    {
      "entropy": 1.8683214783668518,
      "epoch": 0.08788251951588666,
      "grad_norm": 9.149380683898926,
      "learning_rate": 8.533930261788897e-06,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8132390677928925,
      "num_tokens": 34037377.0,
      "step": 28350
    },
    {
      "entropy": 1.8220268458127975,
      "epoch": 0.08791351864093636,
      "grad_norm": 4.427048683166504,
      "learning_rate": 8.532425504444351e-06,
      "loss": 0.5305,
      "mean_token_accuracy": 0.82462307959795,
      "num_tokens": 34050269.0,
      "step": 28360
    },
    {
      "entropy": 1.804908536374569,
      "epoch": 0.08794451776598605,
      "grad_norm": 4.890988349914551,
      "learning_rate": 8.530921542804612e-06,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8353535622358322,
      "num_tokens": 34062946.0,
      "step": 28370
    },
    {
      "entropy": 1.832678309082985,
      "epoch": 0.08797551689103575,
      "grad_norm": 9.909698486328125,
      "learning_rate": 8.52941837616866e-06,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8312997654080391,
      "num_tokens": 34075013.0,
      "step": 28380
    },
    {
      "entropy": 1.7965776398777962,
      "epoch": 0.08800651601608545,
      "grad_norm": 2.5267937183380127,
      "learning_rate": 8.527916003836331e-06,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8231059461832047,
      "num_tokens": 34087851.0,
      "step": 28390
    },
    {
      "entropy": 1.8382470428943634,
      "epoch": 0.08803751514113514,
      "grad_norm": 9.315332412719727,
      "learning_rate": 8.52641442510833e-06,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8272110924124718,
      "num_tokens": 34099777.0,
      "step": 28400
    },
    {
      "entropy": 1.7765842095017432,
      "epoch": 0.08806851426618484,
      "grad_norm": 9.057583808898926,
      "learning_rate": 8.524913639286219e-06,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8396439641714096,
      "num_tokens": 34111989.0,
      "step": 28410
    },
    {
      "entropy": 1.9059796720743178,
      "epoch": 0.08809951339123454,
      "grad_norm": 8.058761596679688,
      "learning_rate": 8.523413645672424e-06,
      "loss": 0.6247,
      "mean_token_accuracy": 0.811768627166748,
      "num_tokens": 34123482.0,
      "step": 28420
    },
    {
      "entropy": 1.8845535546541214,
      "epoch": 0.08813051251628423,
      "grad_norm": 7.467669486999512,
      "learning_rate": 8.52191444357023e-06,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8204743906855583,
      "num_tokens": 34134946.0,
      "step": 28430
    },
    {
      "entropy": 1.789374603331089,
      "epoch": 0.08816151164133393,
      "grad_norm": 8.87814712524414,
      "learning_rate": 8.520416032283778e-06,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8291935101151466,
      "num_tokens": 34147310.0,
      "step": 28440
    },
    {
      "entropy": 1.9232223629951477,
      "epoch": 0.08819251076638362,
      "grad_norm": 10.156554222106934,
      "learning_rate": 8.518918411118063e-06,
      "loss": 0.6747,
      "mean_token_accuracy": 0.8002053126692772,
      "num_tokens": 34157822.0,
      "step": 28450
    },
    {
      "entropy": 1.8766271471977234,
      "epoch": 0.08822350989143331,
      "grad_norm": 9.71645450592041,
      "learning_rate": 8.51742157937894e-06,
      "loss": 0.6254,
      "mean_token_accuracy": 0.8169643610715867,
      "num_tokens": 34168930.0,
      "step": 28460
    },
    {
      "entropy": 1.9275714561343193,
      "epoch": 0.088254509016483,
      "grad_norm": 10.583243370056152,
      "learning_rate": 8.515925536373112e-06,
      "loss": 0.6441,
      "mean_token_accuracy": 0.8056027069687843,
      "num_tokens": 34181366.0,
      "step": 28470
    },
    {
      "entropy": 1.8285674542188644,
      "epoch": 0.0882855081415327,
      "grad_norm": 12.124537467956543,
      "learning_rate": 8.51443028140814e-06,
      "loss": 0.5686,
      "mean_token_accuracy": 0.81962631046772,
      "num_tokens": 34193201.0,
      "step": 28480
    },
    {
      "entropy": 1.8924355298280715,
      "epoch": 0.0883165072665824,
      "grad_norm": 8.544037818908691,
      "learning_rate": 8.512935813792427e-06,
      "loss": 0.5901,
      "mean_token_accuracy": 0.8254889070987701,
      "num_tokens": 34203925.0,
      "step": 28490
    },
    {
      "entropy": 1.8146496564149857,
      "epoch": 0.0883475063916321,
      "grad_norm": 9.721524238586426,
      "learning_rate": 8.511442132835237e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.8225750029087067,
      "num_tokens": 34216079.0,
      "step": 28500
    },
    {
      "entropy": 1.8287780940532685,
      "epoch": 0.08837850551668179,
      "grad_norm": 10.420062065124512,
      "learning_rate": 8.509949237846672e-06,
      "loss": 0.6115,
      "mean_token_accuracy": 0.8089086845517158,
      "num_tokens": 34228708.0,
      "step": 28510
    },
    {
      "entropy": 1.8081783071160316,
      "epoch": 0.08840950464173149,
      "grad_norm": 5.157535076141357,
      "learning_rate": 8.508457128137686e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8270225182175637,
      "num_tokens": 34241771.0,
      "step": 28520
    },
    {
      "entropy": 1.8582287296652793,
      "epoch": 0.08844050376678118,
      "grad_norm": 4.54521369934082,
      "learning_rate": 8.506965803020078e-06,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8173837080597878,
      "num_tokens": 34253510.0,
      "step": 28530
    },
    {
      "entropy": 1.769499270617962,
      "epoch": 0.08847150289183088,
      "grad_norm": 11.29238224029541,
      "learning_rate": 8.50547526180649e-06,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8185042932629585,
      "num_tokens": 34266457.0,
      "step": 28540
    },
    {
      "entropy": 1.8402513667941094,
      "epoch": 0.08850250201688058,
      "grad_norm": 10.66517162322998,
      "learning_rate": 8.503985503810404e-06,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8161203816533089,
      "num_tokens": 34278814.0,
      "step": 28550
    },
    {
      "entropy": 1.9259721651673316,
      "epoch": 0.08853350114193027,
      "grad_norm": 11.027581214904785,
      "learning_rate": 8.502496528346151e-06,
      "loss": 0.6318,
      "mean_token_accuracy": 0.8031986460089684,
      "num_tokens": 34290769.0,
      "step": 28560
    },
    {
      "entropy": 1.8921418815851212,
      "epoch": 0.08856450026697997,
      "grad_norm": 8.231386184692383,
      "learning_rate": 8.501008334728893e-06,
      "loss": 0.573,
      "mean_token_accuracy": 0.8195232212543487,
      "num_tokens": 34301680.0,
      "step": 28570
    },
    {
      "entropy": 1.9299334168434144,
      "epoch": 0.08859549939202967,
      "grad_norm": 12.117446899414062,
      "learning_rate": 8.49952092227464e-06,
      "loss": 0.6296,
      "mean_token_accuracy": 0.8005195692181587,
      "num_tokens": 34313378.0,
      "step": 28580
    },
    {
      "entropy": 1.8172454446554185,
      "epoch": 0.08862649851707935,
      "grad_norm": 8.979788780212402,
      "learning_rate": 8.498034290300233e-06,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8196427300572395,
      "num_tokens": 34326259.0,
      "step": 28590
    },
    {
      "entropy": 1.8738806433975697,
      "epoch": 0.08865749764212905,
      "grad_norm": 9.108292579650879,
      "learning_rate": 8.496548438123347e-06,
      "loss": 0.5925,
      "mean_token_accuracy": 0.8214808851480484,
      "num_tokens": 34337849.0,
      "step": 28600
    },
    {
      "entropy": 1.8880690574645995,
      "epoch": 0.08868849676717874,
      "grad_norm": 10.18707275390625,
      "learning_rate": 8.495063365062501e-06,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8180713891983032,
      "num_tokens": 34349609.0,
      "step": 28610
    },
    {
      "entropy": 1.7710354149341583,
      "epoch": 0.08871949589222844,
      "grad_norm": 2.5999815464019775,
      "learning_rate": 8.493579070437038e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8418709754943847,
      "num_tokens": 34363340.0,
      "step": 28620
    },
    {
      "entropy": 1.7987074330449104,
      "epoch": 0.08875049501727814,
      "grad_norm": 4.221140384674072,
      "learning_rate": 8.492095553567142e-06,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8264068886637688,
      "num_tokens": 34376348.0,
      "step": 28630
    },
    {
      "entropy": 1.8191845625638963,
      "epoch": 0.08878149414232783,
      "grad_norm": 8.881943702697754,
      "learning_rate": 8.49061281377382e-06,
      "loss": 0.615,
      "mean_token_accuracy": 0.8144074931740761,
      "num_tokens": 34389472.0,
      "step": 28640
    },
    {
      "entropy": 1.910823555290699,
      "epoch": 0.08881249326737753,
      "grad_norm": 10.219226837158203,
      "learning_rate": 8.489130850378912e-06,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8214432463049889,
      "num_tokens": 34400353.0,
      "step": 28650
    },
    {
      "entropy": 1.817768232524395,
      "epoch": 0.08884349239242723,
      "grad_norm": 9.067606925964355,
      "learning_rate": 8.487649662705087e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8307969868183136,
      "num_tokens": 34412601.0,
      "step": 28660
    },
    {
      "entropy": 1.7433785900473595,
      "epoch": 0.08887449151747692,
      "grad_norm": 9.326837539672852,
      "learning_rate": 8.48616925007584e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8333866909146309,
      "num_tokens": 34425813.0,
      "step": 28670
    },
    {
      "entropy": 1.8864496439695357,
      "epoch": 0.08890549064252662,
      "grad_norm": 8.599132537841797,
      "learning_rate": 8.484689611815491e-06,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8141407266259193,
      "num_tokens": 34437369.0,
      "step": 28680
    },
    {
      "entropy": 1.8374801024794578,
      "epoch": 0.08893648976757632,
      "grad_norm": 9.512642860412598,
      "learning_rate": 8.483210747249186e-06,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8316513374447823,
      "num_tokens": 34449662.0,
      "step": 28690
    },
    {
      "entropy": 1.7934413641691207,
      "epoch": 0.08896748889262601,
      "grad_norm": 5.296740531921387,
      "learning_rate": 8.481732655702892e-06,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8272898524999619,
      "num_tokens": 34462445.0,
      "step": 28700
    },
    {
      "entropy": 1.8716158524155617,
      "epoch": 0.0889984880176757,
      "grad_norm": 9.011677742004395,
      "learning_rate": 8.4802553365034e-06,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8340117856860161,
      "num_tokens": 34474062.0,
      "step": 28710
    },
    {
      "entropy": 1.801230075955391,
      "epoch": 0.08902948714272539,
      "grad_norm": 8.55329418182373,
      "learning_rate": 8.478778788978323e-06,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8217922821640968,
      "num_tokens": 34487372.0,
      "step": 28720
    },
    {
      "entropy": 1.8363505557179451,
      "epoch": 0.08906048626777509,
      "grad_norm": 9.565589904785156,
      "learning_rate": 8.477303012456088e-06,
      "loss": 0.5462,
      "mean_token_accuracy": 0.825613497197628,
      "num_tokens": 34499424.0,
      "step": 28730
    },
    {
      "entropy": 1.8904439568519593,
      "epoch": 0.08909148539282478,
      "grad_norm": 9.289710998535156,
      "learning_rate": 8.47582800626594e-06,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8117494031786918,
      "num_tokens": 34510549.0,
      "step": 28740
    },
    {
      "entropy": 1.876707810163498,
      "epoch": 0.08912248451787448,
      "grad_norm": 10.488570213317871,
      "learning_rate": 8.474353769737951e-06,
      "loss": 0.584,
      "mean_token_accuracy": 0.8244671374559402,
      "num_tokens": 34522326.0,
      "step": 28750
    },
    {
      "entropy": 1.8384212747216224,
      "epoch": 0.08915348364292418,
      "grad_norm": 4.727441787719727,
      "learning_rate": 8.472880302202995e-06,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8281288787722587,
      "num_tokens": 34533723.0,
      "step": 28760
    },
    {
      "entropy": 1.8969373285770417,
      "epoch": 0.08918448276797387,
      "grad_norm": 10.660612106323242,
      "learning_rate": 8.471407602992768e-06,
      "loss": 0.6065,
      "mean_token_accuracy": 0.8140498995780945,
      "num_tokens": 34545247.0,
      "step": 28770
    },
    {
      "entropy": 1.8668588057160378,
      "epoch": 0.08921548189302357,
      "grad_norm": 5.187861919403076,
      "learning_rate": 8.469935671439776e-06,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8087451472878456,
      "num_tokens": 34557397.0,
      "step": 28780
    },
    {
      "entropy": 1.9032288029789926,
      "epoch": 0.08924648101807327,
      "grad_norm": 5.01014518737793,
      "learning_rate": 8.468464506877338e-06,
      "loss": 0.5857,
      "mean_token_accuracy": 0.8231427073478699,
      "num_tokens": 34568813.0,
      "step": 28790
    },
    {
      "entropy": 1.83003838211298,
      "epoch": 0.08927748014312296,
      "grad_norm": 7.490168571472168,
      "learning_rate": 8.46699410863958e-06,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8283394366502762,
      "num_tokens": 34581102.0,
      "step": 28800
    },
    {
      "entropy": 1.8458102241158485,
      "epoch": 0.08930847926817266,
      "grad_norm": 11.137908935546875,
      "learning_rate": 8.465524476061445e-06,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8272756159305572,
      "num_tokens": 34593458.0,
      "step": 28810
    },
    {
      "entropy": 1.7978875115513802,
      "epoch": 0.08933947839322236,
      "grad_norm": 13.252009391784668,
      "learning_rate": 8.464055608478673e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8363189071416854,
      "num_tokens": 34607214.0,
      "step": 28820
    },
    {
      "entropy": 1.8788623362779617,
      "epoch": 0.08937047751827204,
      "grad_norm": 8.798412322998047,
      "learning_rate": 8.46258750522782e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8359053075313568,
      "num_tokens": 34618731.0,
      "step": 28830
    },
    {
      "entropy": 1.8706136047840118,
      "epoch": 0.08940147664332174,
      "grad_norm": 10.804203033447266,
      "learning_rate": 8.46112016564624e-06,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8170384958386421,
      "num_tokens": 34629767.0,
      "step": 28840
    },
    {
      "entropy": 1.9193976387381553,
      "epoch": 0.08943247576837143,
      "grad_norm": 9.608490943908691,
      "learning_rate": 8.459653589072098e-06,
      "loss": 0.5879,
      "mean_token_accuracy": 0.823236158490181,
      "num_tokens": 34641783.0,
      "step": 28850
    },
    {
      "entropy": 1.953177237510681,
      "epoch": 0.08946347489342113,
      "grad_norm": 9.958290100097656,
      "learning_rate": 8.458187774844355e-06,
      "loss": 0.6343,
      "mean_token_accuracy": 0.8059092834591866,
      "num_tokens": 34652879.0,
      "step": 28860
    },
    {
      "entropy": 1.9278223618865014,
      "epoch": 0.08949447401847083,
      "grad_norm": 9.370678901672363,
      "learning_rate": 8.456722722302779e-06,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8173175677657127,
      "num_tokens": 34664517.0,
      "step": 28870
    },
    {
      "entropy": 1.743288952112198,
      "epoch": 0.08952547314352052,
      "grad_norm": 4.402005672454834,
      "learning_rate": 8.45525843078793e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8388036683201789,
      "num_tokens": 34678318.0,
      "step": 28880
    },
    {
      "entropy": 1.82712120115757,
      "epoch": 0.08955647226857022,
      "grad_norm": 5.668634414672852,
      "learning_rate": 8.453794899641178e-06,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8243148773908615,
      "num_tokens": 34690574.0,
      "step": 28890
    },
    {
      "entropy": 1.794793924689293,
      "epoch": 0.08958747139361992,
      "grad_norm": 4.694520950317383,
      "learning_rate": 8.452332128204687e-06,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8249142169952393,
      "num_tokens": 34702705.0,
      "step": 28900
    },
    {
      "entropy": 1.9069462090730667,
      "epoch": 0.08961847051866961,
      "grad_norm": 9.739960670471191,
      "learning_rate": 8.450870115821412e-06,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8125949770212173,
      "num_tokens": 34713359.0,
      "step": 28910
    },
    {
      "entropy": 1.843550091981888,
      "epoch": 0.08964946964371931,
      "grad_norm": 9.711644172668457,
      "learning_rate": 8.449408861835107e-06,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8326569676399231,
      "num_tokens": 34724846.0,
      "step": 28920
    },
    {
      "entropy": 1.8499679207801818,
      "epoch": 0.089680468768769,
      "grad_norm": 9.517023086547852,
      "learning_rate": 8.447948365590324e-06,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8209212452173233,
      "num_tokens": 34736090.0,
      "step": 28930
    },
    {
      "entropy": 1.8778686568140983,
      "epoch": 0.0897114678938187,
      "grad_norm": 7.4227423667907715,
      "learning_rate": 8.446488626432398e-06,
      "loss": 0.5703,
      "mean_token_accuracy": 0.8157725527882576,
      "num_tokens": 34748597.0,
      "step": 28940
    },
    {
      "entropy": 1.901959116756916,
      "epoch": 0.0897424670188684,
      "grad_norm": 8.503968238830566,
      "learning_rate": 8.445029643707466e-06,
      "loss": 0.6077,
      "mean_token_accuracy": 0.8245082676410675,
      "num_tokens": 34760085.0,
      "step": 28950
    },
    {
      "entropy": 1.9186773270368576,
      "epoch": 0.08977346614391808,
      "grad_norm": 9.638235092163086,
      "learning_rate": 8.443571416762454e-06,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8147833585739136,
      "num_tokens": 34771714.0,
      "step": 28960
    },
    {
      "entropy": 1.9039877220988273,
      "epoch": 0.08980446526896778,
      "grad_norm": 10.257258415222168,
      "learning_rate": 8.442113944945066e-06,
      "loss": 0.5929,
      "mean_token_accuracy": 0.8272214874625206,
      "num_tokens": 34783065.0,
      "step": 28970
    },
    {
      "entropy": 1.8442776799201965,
      "epoch": 0.08983546439401748,
      "grad_norm": 8.371413230895996,
      "learning_rate": 8.440657227603809e-06,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8354337051510811,
      "num_tokens": 34796313.0,
      "step": 28980
    },
    {
      "entropy": 1.8650359570980073,
      "epoch": 0.08986646351906717,
      "grad_norm": 8.93613052368164,
      "learning_rate": 8.439201264087966e-06,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8383944362401963,
      "num_tokens": 34809266.0,
      "step": 28990
    },
    {
      "entropy": 1.898645070195198,
      "epoch": 0.08989746264411687,
      "grad_norm": 4.809848785400391,
      "learning_rate": 8.437746053747611e-06,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8344621181488037,
      "num_tokens": 34821442.0,
      "step": 29000
    },
    {
      "entropy": 1.8691908940672874,
      "epoch": 0.08992846176916656,
      "grad_norm": 4.506728649139404,
      "learning_rate": 8.436291595933597e-06,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8261950254440308,
      "num_tokens": 34834092.0,
      "step": 29010
    },
    {
      "entropy": 1.989156812429428,
      "epoch": 0.08995946089421626,
      "grad_norm": 10.102865219116211,
      "learning_rate": 8.434837889997567e-06,
      "loss": 0.6544,
      "mean_token_accuracy": 0.8023478895425796,
      "num_tokens": 34844589.0,
      "step": 29020
    },
    {
      "entropy": 1.8639222919940948,
      "epoch": 0.08999046001926596,
      "grad_norm": 8.607808113098145,
      "learning_rate": 8.433384935291941e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8342802464962006,
      "num_tokens": 34856930.0,
      "step": 29030
    },
    {
      "entropy": 1.888324649631977,
      "epoch": 0.09002145914431565,
      "grad_norm": 15.61713695526123,
      "learning_rate": 8.43193273116992e-06,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8299177408218383,
      "num_tokens": 34868535.0,
      "step": 29040
    },
    {
      "entropy": 1.9326095387339592,
      "epoch": 0.09005245826936535,
      "grad_norm": 13.876715660095215,
      "learning_rate": 8.430481276985486e-06,
      "loss": 0.606,
      "mean_token_accuracy": 0.8128352046012879,
      "num_tokens": 34879434.0,
      "step": 29050
    },
    {
      "entropy": 1.8903497770428657,
      "epoch": 0.09008345739441505,
      "grad_norm": 10.24538803100586,
      "learning_rate": 8.429030572093397e-06,
      "loss": 0.556,
      "mean_token_accuracy": 0.8233076304197311,
      "num_tokens": 34891052.0,
      "step": 29060
    },
    {
      "entropy": 1.8975743114948274,
      "epoch": 0.09011445651946474,
      "grad_norm": 11.500408172607422,
      "learning_rate": 8.427580615849188e-06,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8229522377252578,
      "num_tokens": 34903040.0,
      "step": 29070
    },
    {
      "entropy": 1.8994533449411393,
      "epoch": 0.09014545564451443,
      "grad_norm": 11.263349533081055,
      "learning_rate": 8.426131407609173e-06,
      "loss": 0.5816,
      "mean_token_accuracy": 0.820560471713543,
      "num_tokens": 34914402.0,
      "step": 29080
    },
    {
      "entropy": 1.8676754549145698,
      "epoch": 0.09017645476956412,
      "grad_norm": 9.663436889648438,
      "learning_rate": 8.42468294673044e-06,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8182600021362305,
      "num_tokens": 34926419.0,
      "step": 29090
    },
    {
      "entropy": 1.8573566570878028,
      "epoch": 0.09020745389461382,
      "grad_norm": 11.30569076538086,
      "learning_rate": 8.423235232570846e-06,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8217246904969215,
      "num_tokens": 34938924.0,
      "step": 29100
    },
    {
      "entropy": 1.9108622312545775,
      "epoch": 0.09023845301966352,
      "grad_norm": 5.698962211608887,
      "learning_rate": 8.421788264489021e-06,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8229934841394424,
      "num_tokens": 34950788.0,
      "step": 29110
    },
    {
      "entropy": 1.777287058532238,
      "epoch": 0.09026945214471321,
      "grad_norm": 9.155726432800293,
      "learning_rate": 8.420342041844372e-06,
      "loss": 0.518,
      "mean_token_accuracy": 0.834735095500946,
      "num_tokens": 34963772.0,
      "step": 29120
    },
    {
      "entropy": 1.8479400515556335,
      "epoch": 0.09030045126976291,
      "grad_norm": 4.175968170166016,
      "learning_rate": 8.418896563997072e-06,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8213765308260917,
      "num_tokens": 34976155.0,
      "step": 29130
    },
    {
      "entropy": 1.8892390161752701,
      "epoch": 0.0903314503948126,
      "grad_norm": 10.586323738098145,
      "learning_rate": 8.41745183030806e-06,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8211356267333031,
      "num_tokens": 34988223.0,
      "step": 29140
    },
    {
      "entropy": 1.8357836604118347,
      "epoch": 0.0903624495198623,
      "grad_norm": 10.024797439575195,
      "learning_rate": 8.416007840139042e-06,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8338177174329757,
      "num_tokens": 35001045.0,
      "step": 29150
    },
    {
      "entropy": 1.887669287621975,
      "epoch": 0.090393448644912,
      "grad_norm": 9.929082870483398,
      "learning_rate": 8.4145645928525e-06,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8130573540925979,
      "num_tokens": 35012268.0,
      "step": 29160
    },
    {
      "entropy": 1.8210702255368232,
      "epoch": 0.0904244477699617,
      "grad_norm": 4.600307464599609,
      "learning_rate": 8.413122087811668e-06,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8326424300670624,
      "num_tokens": 35025165.0,
      "step": 29170
    },
    {
      "entropy": 1.9049372583627702,
      "epoch": 0.0904554468950114,
      "grad_norm": 9.302005767822266,
      "learning_rate": 8.411680324380554e-06,
      "loss": 0.5807,
      "mean_token_accuracy": 0.8166706204414368,
      "num_tokens": 35036999.0,
      "step": 29180
    },
    {
      "entropy": 1.8811736673116684,
      "epoch": 0.09048644602006109,
      "grad_norm": 11.642059326171875,
      "learning_rate": 8.410239301923921e-06,
      "loss": 0.5831,
      "mean_token_accuracy": 0.8235431507229805,
      "num_tokens": 35048718.0,
      "step": 29190
    },
    {
      "entropy": 1.7818408221006394,
      "epoch": 0.09051744514511077,
      "grad_norm": 3.9471521377563477,
      "learning_rate": 8.408799019807298e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8274667426943779,
      "num_tokens": 35062426.0,
      "step": 29200
    },
    {
      "entropy": 1.9234388038516044,
      "epoch": 0.09054844427016047,
      "grad_norm": 9.527081489562988,
      "learning_rate": 8.407359477396974e-06,
      "loss": 0.5952,
      "mean_token_accuracy": 0.824344064295292,
      "num_tokens": 35073342.0,
      "step": 29210
    },
    {
      "entropy": 1.9060495615005493,
      "epoch": 0.09057944339521017,
      "grad_norm": 4.219698905944824,
      "learning_rate": 8.405920674059997e-06,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8162388294935227,
      "num_tokens": 35085769.0,
      "step": 29220
    },
    {
      "entropy": 1.9729074537754059,
      "epoch": 0.09061044252025986,
      "grad_norm": 12.57748794555664,
      "learning_rate": 8.404482609164172e-06,
      "loss": 0.678,
      "mean_token_accuracy": 0.8029696837067604,
      "num_tokens": 35096711.0,
      "step": 29230
    },
    {
      "entropy": 1.9721817374229431,
      "epoch": 0.09064144164530956,
      "grad_norm": 9.833843231201172,
      "learning_rate": 8.40304528207806e-06,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8235157504677773,
      "num_tokens": 35107468.0,
      "step": 29240
    },
    {
      "entropy": 1.8956667900085449,
      "epoch": 0.09067244077035926,
      "grad_norm": 8.549565315246582,
      "learning_rate": 8.40160869217098e-06,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8300123199820518,
      "num_tokens": 35119143.0,
      "step": 29250
    },
    {
      "entropy": 1.856950616836548,
      "epoch": 0.09070343989540895,
      "grad_norm": 8.162071228027344,
      "learning_rate": 8.400172838813004e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8311971753835679,
      "num_tokens": 35131680.0,
      "step": 29260
    },
    {
      "entropy": 1.92002714574337,
      "epoch": 0.09073443902045865,
      "grad_norm": 9.52449893951416,
      "learning_rate": 8.398737721374958e-06,
      "loss": 0.5886,
      "mean_token_accuracy": 0.8099859327077865,
      "num_tokens": 35143630.0,
      "step": 29270
    },
    {
      "entropy": 1.8672298848628999,
      "epoch": 0.09076543814550835,
      "grad_norm": 10.40442180633545,
      "learning_rate": 8.39730333922842e-06,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8219729751348496,
      "num_tokens": 35155669.0,
      "step": 29280
    },
    {
      "entropy": 1.8741114243865014,
      "epoch": 0.09079643727055804,
      "grad_norm": 9.638392448425293,
      "learning_rate": 8.39586969174572e-06,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8221320033073425,
      "num_tokens": 35168182.0,
      "step": 29290
    },
    {
      "entropy": 1.8563538879156112,
      "epoch": 0.09082743639560774,
      "grad_norm": 11.128931999206543,
      "learning_rate": 8.394436778299934e-06,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8322840392589569,
      "num_tokens": 35180920.0,
      "step": 29300
    },
    {
      "entropy": 1.940355482697487,
      "epoch": 0.09085843552065744,
      "grad_norm": 12.628369331359863,
      "learning_rate": 8.393004598264892e-06,
      "loss": 0.6027,
      "mean_token_accuracy": 0.8121664762496948,
      "num_tokens": 35191979.0,
      "step": 29310
    },
    {
      "entropy": 1.881583635509014,
      "epoch": 0.09088943464570713,
      "grad_norm": 10.32827091217041,
      "learning_rate": 8.391573151015169e-06,
      "loss": 0.5926,
      "mean_token_accuracy": 0.8216746702790261,
      "num_tokens": 35204154.0,
      "step": 29320
    },
    {
      "entropy": 1.9963495016098023,
      "epoch": 0.09092043377075681,
      "grad_norm": 10.914651870727539,
      "learning_rate": 8.390142435926085e-06,
      "loss": 0.6708,
      "mean_token_accuracy": 0.802430622279644,
      "num_tokens": 35215370.0,
      "step": 29330
    },
    {
      "entropy": 1.807043470442295,
      "epoch": 0.09095143289580651,
      "grad_norm": 6.392316818237305,
      "learning_rate": 8.38871245237371e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.83642118871212,
      "num_tokens": 35228636.0,
      "step": 29340
    },
    {
      "entropy": 1.8242978394031524,
      "epoch": 0.09098243202085621,
      "grad_norm": 8.627922058105469,
      "learning_rate": 8.387283199734848e-06,
      "loss": 0.4963,
      "mean_token_accuracy": 0.833967824280262,
      "num_tokens": 35241156.0,
      "step": 29350
    },
    {
      "entropy": 1.937102773785591,
      "epoch": 0.0910134311459059,
      "grad_norm": 7.6476640701293945,
      "learning_rate": 8.38585467738706e-06,
      "loss": 0.6429,
      "mean_token_accuracy": 0.810026279091835,
      "num_tokens": 35252420.0,
      "step": 29360
    },
    {
      "entropy": 1.8407186821103096,
      "epoch": 0.0910444302709556,
      "grad_norm": 9.836170196533203,
      "learning_rate": 8.38442688470864e-06,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8297666862607003,
      "num_tokens": 35264790.0,
      "step": 29370
    },
    {
      "entropy": 1.8077852353453636,
      "epoch": 0.0910754293960053,
      "grad_norm": 9.396247863769531,
      "learning_rate": 8.382999821078624e-06,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8240919351577759,
      "num_tokens": 35277676.0,
      "step": 29380
    },
    {
      "entropy": 1.8391135558485985,
      "epoch": 0.091106428521055,
      "grad_norm": 4.910935401916504,
      "learning_rate": 8.381573485876786e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8325628146529198,
      "num_tokens": 35290859.0,
      "step": 29390
    },
    {
      "entropy": 1.8129886105656623,
      "epoch": 0.09113742764610469,
      "grad_norm": 9.58067798614502,
      "learning_rate": 8.380147878483645e-06,
      "loss": 0.5201,
      "mean_token_accuracy": 0.826962910592556,
      "num_tokens": 35303398.0,
      "step": 29400
    },
    {
      "entropy": 1.85167246311903,
      "epoch": 0.09116842677115439,
      "grad_norm": 8.074714660644531,
      "learning_rate": 8.378722998280448e-06,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8280769392848015,
      "num_tokens": 35315344.0,
      "step": 29410
    },
    {
      "entropy": 1.992096221446991,
      "epoch": 0.09119942589620408,
      "grad_norm": 9.145973205566406,
      "learning_rate": 8.377298844649186e-06,
      "loss": 0.7051,
      "mean_token_accuracy": 0.7987665429711341,
      "num_tokens": 35326107.0,
      "step": 29420
    },
    {
      "entropy": 1.8843266785144805,
      "epoch": 0.09123042502125378,
      "grad_norm": 9.272721290588379,
      "learning_rate": 8.375875416972584e-06,
      "loss": 0.5242,
      "mean_token_accuracy": 0.83641587048769,
      "num_tokens": 35338238.0,
      "step": 29430
    },
    {
      "entropy": 1.8167478024959565,
      "epoch": 0.09126142414630348,
      "grad_norm": 3.9986329078674316,
      "learning_rate": 8.374452714634094e-06,
      "loss": 0.5826,
      "mean_token_accuracy": 0.8207257181406021,
      "num_tokens": 35351268.0,
      "step": 29440
    },
    {
      "entropy": 1.9153858974575997,
      "epoch": 0.09129242327135316,
      "grad_norm": 10.753203392028809,
      "learning_rate": 8.373030737017907e-06,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8246575996279717,
      "num_tokens": 35362594.0,
      "step": 29450
    },
    {
      "entropy": 1.9501359462738037,
      "epoch": 0.09132342239640286,
      "grad_norm": 9.741477966308594,
      "learning_rate": 8.371609483508947e-06,
      "loss": 0.6487,
      "mean_token_accuracy": 0.8103729501366616,
      "num_tokens": 35373596.0,
      "step": 29460
    },
    {
      "entropy": 1.932808554172516,
      "epoch": 0.09135442152145255,
      "grad_norm": 11.1822509765625,
      "learning_rate": 8.370188953492866e-06,
      "loss": 0.6056,
      "mean_token_accuracy": 0.821463891863823,
      "num_tokens": 35385127.0,
      "step": 29470
    },
    {
      "entropy": 1.8480829164385795,
      "epoch": 0.09138542064650225,
      "grad_norm": 8.782395362854004,
      "learning_rate": 8.368769146356043e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8359311327338219,
      "num_tokens": 35397793.0,
      "step": 29480
    },
    {
      "entropy": 1.9216597527265549,
      "epoch": 0.09141641977155195,
      "grad_norm": 10.21785831451416,
      "learning_rate": 8.36735006148559e-06,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8218403205275535,
      "num_tokens": 35408887.0,
      "step": 29490
    },
    {
      "entropy": 1.854940627515316,
      "epoch": 0.09144741889660164,
      "grad_norm": 9.738880157470703,
      "learning_rate": 8.365931698269346e-06,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8242891728878021,
      "num_tokens": 35421648.0,
      "step": 29500
    },
    {
      "entropy": 1.8743022330105306,
      "epoch": 0.09147841802165134,
      "grad_norm": 10.232392311096191,
      "learning_rate": 8.36451405609587e-06,
      "loss": 0.5753,
      "mean_token_accuracy": 0.8169502720236779,
      "num_tokens": 35434473.0,
      "step": 29510
    },
    {
      "entropy": 1.8739581793546676,
      "epoch": 0.09150941714670104,
      "grad_norm": 9.921628952026367,
      "learning_rate": 8.363097134354453e-06,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8249258384108543,
      "num_tokens": 35446211.0,
      "step": 29520
    },
    {
      "entropy": 1.9153602778911591,
      "epoch": 0.09154041627175073,
      "grad_norm": 9.796987533569336,
      "learning_rate": 8.361680932435107e-06,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8256885960698128,
      "num_tokens": 35457049.0,
      "step": 29530
    },
    {
      "entropy": 1.8282124564051627,
      "epoch": 0.09157141539680043,
      "grad_norm": 8.886029243469238,
      "learning_rate": 8.360265449728567e-06,
      "loss": 0.5452,
      "mean_token_accuracy": 0.8246500983834266,
      "num_tokens": 35470118.0,
      "step": 29540
    },
    {
      "entropy": 1.8801322914659977,
      "epoch": 0.09160241452185013,
      "grad_norm": 6.403395652770996,
      "learning_rate": 8.358850685626288e-06,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8278462216258049,
      "num_tokens": 35483031.0,
      "step": 29550
    },
    {
      "entropy": 1.8667038604617119,
      "epoch": 0.09163341364689982,
      "grad_norm": 8.779369354248047,
      "learning_rate": 8.357436639520454e-06,
      "loss": 0.6143,
      "mean_token_accuracy": 0.8238617271184921,
      "num_tokens": 35494985.0,
      "step": 29560
    },
    {
      "entropy": 1.8358441174030304,
      "epoch": 0.0916644127719495,
      "grad_norm": 9.281476020812988,
      "learning_rate": 8.356023310803953e-06,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8221923336386681,
      "num_tokens": 35506647.0,
      "step": 29570
    },
    {
      "entropy": 1.8791137009859085,
      "epoch": 0.0916954118969992,
      "grad_norm": 11.18396282196045,
      "learning_rate": 8.354610698870407e-06,
      "loss": 0.6371,
      "mean_token_accuracy": 0.8211662247776985,
      "num_tokens": 35518289.0,
      "step": 29580
    },
    {
      "entropy": 1.8068061396479607,
      "epoch": 0.0917264110220489,
      "grad_norm": 9.689407348632812,
      "learning_rate": 8.353198803114144e-06,
      "loss": 0.5216,
      "mean_token_accuracy": 0.819252048432827,
      "num_tokens": 35531769.0,
      "step": 29590
    },
    {
      "entropy": 1.8276824593544005,
      "epoch": 0.0917574101470986,
      "grad_norm": 8.703977584838867,
      "learning_rate": 8.351787622930218e-06,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8255561321973801,
      "num_tokens": 35544219.0,
      "step": 29600
    },
    {
      "entropy": 1.9323107048869133,
      "epoch": 0.09178840927214829,
      "grad_norm": 9.766465187072754,
      "learning_rate": 8.350377157714388e-06,
      "loss": 0.6171,
      "mean_token_accuracy": 0.8179709568619729,
      "num_tokens": 35555186.0,
      "step": 29610
    },
    {
      "entropy": 1.896198531985283,
      "epoch": 0.09181940839719799,
      "grad_norm": 9.552749633789062,
      "learning_rate": 8.348967406863137e-06,
      "loss": 0.6246,
      "mean_token_accuracy": 0.8194371804594993,
      "num_tokens": 35567088.0,
      "step": 29620
    },
    {
      "entropy": 1.9005744218826295,
      "epoch": 0.09185040752224768,
      "grad_norm": 8.323497772216797,
      "learning_rate": 8.347558369773652e-06,
      "loss": 0.5541,
      "mean_token_accuracy": 0.8280677005648613,
      "num_tokens": 35578712.0,
      "step": 29630
    },
    {
      "entropy": 1.8314684435725213,
      "epoch": 0.09188140664729738,
      "grad_norm": 3.150182008743286,
      "learning_rate": 8.346150045843839e-06,
      "loss": 0.5522,
      "mean_token_accuracy": 0.8305447429418564,
      "num_tokens": 35591457.0,
      "step": 29640
    },
    {
      "entropy": 1.8687354385852815,
      "epoch": 0.09191240577234708,
      "grad_norm": 9.428099632263184,
      "learning_rate": 8.344742434472313e-06,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8308294847607612,
      "num_tokens": 35603419.0,
      "step": 29650
    },
    {
      "entropy": 1.8598203748464583,
      "epoch": 0.09194340489739677,
      "grad_norm": 7.902471542358398,
      "learning_rate": 8.343335535058393e-06,
      "loss": 0.508,
      "mean_token_accuracy": 0.8371548101305961,
      "num_tokens": 35615606.0,
      "step": 29660
    },
    {
      "entropy": 1.955969789624214,
      "epoch": 0.09197440402244647,
      "grad_norm": 9.46319580078125,
      "learning_rate": 8.341929347002115e-06,
      "loss": 0.6819,
      "mean_token_accuracy": 0.8057120770215989,
      "num_tokens": 35626376.0,
      "step": 29670
    },
    {
      "entropy": 1.9067956805229187,
      "epoch": 0.09200540314749617,
      "grad_norm": 4.286515235900879,
      "learning_rate": 8.340523869704218e-06,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8169622346758842,
      "num_tokens": 35637722.0,
      "step": 29680
    },
    {
      "entropy": 1.824674019217491,
      "epoch": 0.09203640227254586,
      "grad_norm": 9.380338668823242,
      "learning_rate": 8.33911910256615e-06,
      "loss": 0.5286,
      "mean_token_accuracy": 0.836339183151722,
      "num_tokens": 35650536.0,
      "step": 29690
    },
    {
      "entropy": 1.8453435346484184,
      "epoch": 0.09206740139759555,
      "grad_norm": 4.688096523284912,
      "learning_rate": 8.33771504499006e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8418554350733757,
      "num_tokens": 35663421.0,
      "step": 29700
    },
    {
      "entropy": 1.9342348664999007,
      "epoch": 0.09209840052264524,
      "grad_norm": 8.24661922454834,
      "learning_rate": 8.336311696378805e-06,
      "loss": 0.579,
      "mean_token_accuracy": 0.8277911230921745,
      "num_tokens": 35674421.0,
      "step": 29710
    },
    {
      "entropy": 1.9771510928869247,
      "epoch": 0.09212939964769494,
      "grad_norm": 11.507468223571777,
      "learning_rate": 8.334909056135947e-06,
      "loss": 0.6472,
      "mean_token_accuracy": 0.8096749007701873,
      "num_tokens": 35685226.0,
      "step": 29720
    },
    {
      "entropy": 1.8233107894659042,
      "epoch": 0.09216039877274464,
      "grad_norm": 11.91437816619873,
      "learning_rate": 8.333507123665745e-06,
      "loss": 0.5913,
      "mean_token_accuracy": 0.8291966646909714,
      "num_tokens": 35697511.0,
      "step": 29730
    },
    {
      "entropy": 1.9061884060502052,
      "epoch": 0.09219139789779433,
      "grad_norm": 9.895060539245605,
      "learning_rate": 8.332105898373162e-06,
      "loss": 0.5925,
      "mean_token_accuracy": 0.8209904283285141,
      "num_tokens": 35709184.0,
      "step": 29740
    },
    {
      "entropy": 1.8389362052083016,
      "epoch": 0.09222239702284403,
      "grad_norm": 9.202600479125977,
      "learning_rate": 8.330705379663864e-06,
      "loss": 0.538,
      "mean_token_accuracy": 0.8290344223380088,
      "num_tokens": 35722082.0,
      "step": 29750
    },
    {
      "entropy": 1.836703784763813,
      "epoch": 0.09225339614789373,
      "grad_norm": 9.766961097717285,
      "learning_rate": 8.32930556694421e-06,
      "loss": 0.6326,
      "mean_token_accuracy": 0.8234434753656388,
      "num_tokens": 35735502.0,
      "step": 29760
    },
    {
      "entropy": 1.823054054379463,
      "epoch": 0.09228439527294342,
      "grad_norm": 10.460724830627441,
      "learning_rate": 8.327906459621262e-06,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8367855966091156,
      "num_tokens": 35747655.0,
      "step": 29770
    },
    {
      "entropy": 1.8504664957523347,
      "epoch": 0.09231539439799312,
      "grad_norm": 10.958568572998047,
      "learning_rate": 8.32650805710278e-06,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8216020077466964,
      "num_tokens": 35759453.0,
      "step": 29780
    },
    {
      "entropy": 1.8954664573073388,
      "epoch": 0.09234639352304282,
      "grad_norm": 9.923591613769531,
      "learning_rate": 8.32511035879721e-06,
      "loss": 0.549,
      "mean_token_accuracy": 0.8215544745326042,
      "num_tokens": 35770958.0,
      "step": 29790
    },
    {
      "entropy": 1.8316803082823754,
      "epoch": 0.09237739264809251,
      "grad_norm": 9.00976848602295,
      "learning_rate": 8.323713364113706e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8291282281279564,
      "num_tokens": 35784494.0,
      "step": 29800
    },
    {
      "entropy": 1.8252049744129182,
      "epoch": 0.09240839177314221,
      "grad_norm": 4.551512241363525,
      "learning_rate": 8.322317072462106e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8328192785382271,
      "num_tokens": 35797555.0,
      "step": 29810
    },
    {
      "entropy": 1.86438340395689,
      "epoch": 0.09243939089819189,
      "grad_norm": 12.345170021057129,
      "learning_rate": 8.320921483252948e-06,
      "loss": 0.5755,
      "mean_token_accuracy": 0.8219951093196869,
      "num_tokens": 35810502.0,
      "step": 29820
    },
    {
      "entropy": 1.8994136035442353,
      "epoch": 0.09247039002324159,
      "grad_norm": 11.431099891662598,
      "learning_rate": 8.319526595897457e-06,
      "loss": 0.591,
      "mean_token_accuracy": 0.8122029930353165,
      "num_tokens": 35822303.0,
      "step": 29830
    },
    {
      "entropy": 1.8895697325468064,
      "epoch": 0.09250138914829129,
      "grad_norm": 9.15605354309082,
      "learning_rate": 8.318132409807547e-06,
      "loss": 0.5452,
      "mean_token_accuracy": 0.8253117471933364,
      "num_tokens": 35834691.0,
      "step": 29840
    },
    {
      "entropy": 1.8310220405459403,
      "epoch": 0.09253238827334098,
      "grad_norm": 9.539107322692871,
      "learning_rate": 8.31673892439583e-06,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8205739751458168,
      "num_tokens": 35847122.0,
      "step": 29850
    },
    {
      "entropy": 1.8387802302837373,
      "epoch": 0.09256338739839068,
      "grad_norm": 9.641124725341797,
      "learning_rate": 8.315346139075596e-06,
      "loss": 0.5816,
      "mean_token_accuracy": 0.8287546694278717,
      "num_tokens": 35860015.0,
      "step": 29860
    },
    {
      "entropy": 1.8979364216327668,
      "epoch": 0.09259438652344038,
      "grad_norm": 9.997504234313965,
      "learning_rate": 8.31395405326083e-06,
      "loss": 0.5852,
      "mean_token_accuracy": 0.8195781350135803,
      "num_tokens": 35872240.0,
      "step": 29870
    },
    {
      "entropy": 1.9377481788396835,
      "epoch": 0.09262538564849007,
      "grad_norm": 9.89439868927002,
      "learning_rate": 8.3125626663662e-06,
      "loss": 0.6192,
      "mean_token_accuracy": 0.8212395742535591,
      "num_tokens": 35883875.0,
      "step": 29880
    },
    {
      "entropy": 1.9368197679519654,
      "epoch": 0.09265638477353977,
      "grad_norm": 9.990235328674316,
      "learning_rate": 8.311171977807062e-06,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8287534207105637,
      "num_tokens": 35896237.0,
      "step": 29890
    },
    {
      "entropy": 1.9469063565135003,
      "epoch": 0.09268738389858946,
      "grad_norm": 8.493316650390625,
      "learning_rate": 8.309781986999454e-06,
      "loss": 0.5916,
      "mean_token_accuracy": 0.813089169561863,
      "num_tokens": 35907924.0,
      "step": 29900
    },
    {
      "entropy": 1.9415881425142287,
      "epoch": 0.09271838302363916,
      "grad_norm": 12.5552978515625,
      "learning_rate": 8.3083926933601e-06,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8337080240249634,
      "num_tokens": 35919711.0,
      "step": 29910
    },
    {
      "entropy": 1.9274977430701257,
      "epoch": 0.09274938214868886,
      "grad_norm": 3.963390588760376,
      "learning_rate": 8.307004096306404e-06,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8345837160944939,
      "num_tokens": 35931724.0,
      "step": 29920
    },
    {
      "entropy": 1.9158756092190743,
      "epoch": 0.09278038127373855,
      "grad_norm": 10.906941413879395,
      "learning_rate": 8.30561619525645e-06,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8259297773241997,
      "num_tokens": 35943022.0,
      "step": 29930
    },
    {
      "entropy": 1.8155071794986726,
      "epoch": 0.09281138039878824,
      "grad_norm": 8.987025260925293,
      "learning_rate": 8.304228989629007e-06,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8335763946175575,
      "num_tokens": 35955373.0,
      "step": 29940
    },
    {
      "entropy": 1.8580231979489326,
      "epoch": 0.09284237952383793,
      "grad_norm": 7.32357120513916,
      "learning_rate": 8.302842478843522e-06,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8255671873688698,
      "num_tokens": 35968204.0,
      "step": 29950
    },
    {
      "entropy": 1.9050882533192635,
      "epoch": 0.09287337864888763,
      "grad_norm": 8.925579071044922,
      "learning_rate": 8.301456662320118e-06,
      "loss": 0.6012,
      "mean_token_accuracy": 0.810367950797081,
      "num_tokens": 35979573.0,
      "step": 29960
    },
    {
      "entropy": 1.836830762028694,
      "epoch": 0.09290437777393733,
      "grad_norm": 3.3085451126098633,
      "learning_rate": 8.300071539479595e-06,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8347239702939987,
      "num_tokens": 35992865.0,
      "step": 29970
    },
    {
      "entropy": 1.9335076332092285,
      "epoch": 0.09293537689898702,
      "grad_norm": 4.406698226928711,
      "learning_rate": 8.298687109743434e-06,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8291699022054673,
      "num_tokens": 36004840.0,
      "step": 29980
    },
    {
      "entropy": 1.9243319526314735,
      "epoch": 0.09296637602403672,
      "grad_norm": 9.231583595275879,
      "learning_rate": 8.297303372533783e-06,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8422666415572166,
      "num_tokens": 36016239.0,
      "step": 29990
    },
    {
      "entropy": 1.9396521881222726,
      "epoch": 0.09299737514908642,
      "grad_norm": 10.043663024902344,
      "learning_rate": 8.295920327273474e-06,
      "loss": 0.6083,
      "mean_token_accuracy": 0.8137664362788201,
      "num_tokens": 36027621.0,
      "step": 30000
    },
    {
      "entropy": 1.8998918011784554,
      "epoch": 0.09302837427413611,
      "grad_norm": 3.658256769180298,
      "learning_rate": 8.294537973386005e-06,
      "loss": 0.5306,
      "mean_token_accuracy": 0.8297856241464615,
      "num_tokens": 36039304.0,
      "step": 30010
    },
    {
      "entropy": 1.905530793964863,
      "epoch": 0.09305937339918581,
      "grad_norm": 10.163437843322754,
      "learning_rate": 8.29315631029555e-06,
      "loss": 0.5719,
      "mean_token_accuracy": 0.824944506585598,
      "num_tokens": 36050882.0,
      "step": 30020
    },
    {
      "entropy": 1.7664498060941696,
      "epoch": 0.0930903725242355,
      "grad_norm": 11.35702896118164,
      "learning_rate": 8.291775337426954e-06,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8312182664871216,
      "num_tokens": 36065249.0,
      "step": 30030
    },
    {
      "entropy": 1.8767417997121811,
      "epoch": 0.0931213716492852,
      "grad_norm": 8.986226081848145,
      "learning_rate": 8.290395054205727e-06,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8315106689929962,
      "num_tokens": 36078427.0,
      "step": 30040
    },
    {
      "entropy": 1.9429257303476333,
      "epoch": 0.0931523707743349,
      "grad_norm": 5.139298915863037,
      "learning_rate": 8.289015460058055e-06,
      "loss": 0.6161,
      "mean_token_accuracy": 0.8064095541834831,
      "num_tokens": 36090554.0,
      "step": 30050
    },
    {
      "entropy": 1.8949994623661042,
      "epoch": 0.0931833698993846,
      "grad_norm": 3.8736157417297363,
      "learning_rate": 8.28763655441079e-06,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8159836381673813,
      "num_tokens": 36103427.0,
      "step": 30060
    },
    {
      "entropy": 1.9144406855106353,
      "epoch": 0.09321436902443428,
      "grad_norm": 9.992815971374512,
      "learning_rate": 8.286258336691447e-06,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8263021633028984,
      "num_tokens": 36116559.0,
      "step": 30070
    },
    {
      "entropy": 1.8954633638262748,
      "epoch": 0.09324536814948398,
      "grad_norm": 10.11132526397705,
      "learning_rate": 8.284880806328216e-06,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8312999933958054,
      "num_tokens": 36129181.0,
      "step": 30080
    },
    {
      "entropy": 1.9233060747385025,
      "epoch": 0.09327636727453367,
      "grad_norm": 5.417168617248535,
      "learning_rate": 8.283503962749944e-06,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8156366124749184,
      "num_tokens": 36141429.0,
      "step": 30090
    },
    {
      "entropy": 1.9522355869412422,
      "epoch": 0.09330736639958337,
      "grad_norm": 10.192994117736816,
      "learning_rate": 8.282127805386145e-06,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8214729785919189,
      "num_tokens": 36153159.0,
      "step": 30100
    },
    {
      "entropy": 1.9816941738128662,
      "epoch": 0.09333836552463307,
      "grad_norm": 10.674017906188965,
      "learning_rate": 8.280752333666999e-06,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8102478966116905,
      "num_tokens": 36164765.0,
      "step": 30110
    },
    {
      "entropy": 1.9644594475626946,
      "epoch": 0.09336936464968276,
      "grad_norm": 8.998485565185547,
      "learning_rate": 8.279377547023342e-06,
      "loss": 0.5509,
      "mean_token_accuracy": 0.8299065142869949,
      "num_tokens": 36176082.0,
      "step": 30120
    },
    {
      "entropy": 1.888843522965908,
      "epoch": 0.09340036377473246,
      "grad_norm": 4.065151691436768,
      "learning_rate": 8.278003444886679e-06,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8304722428321838,
      "num_tokens": 36189223.0,
      "step": 30130
    },
    {
      "entropy": 1.8697800204157828,
      "epoch": 0.09343136289978216,
      "grad_norm": 9.428620338439941,
      "learning_rate": 8.276630026689168e-06,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8309311479330063,
      "num_tokens": 36202150.0,
      "step": 30140
    },
    {
      "entropy": 1.933559250831604,
      "epoch": 0.09346236202483185,
      "grad_norm": 9.53989028930664,
      "learning_rate": 8.275257291863631e-06,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8235199645161628,
      "num_tokens": 36213528.0,
      "step": 30150
    },
    {
      "entropy": 1.9191157951951028,
      "epoch": 0.09349336114988155,
      "grad_norm": 11.843420028686523,
      "learning_rate": 8.273885239843545e-06,
      "loss": 0.5874,
      "mean_token_accuracy": 0.8172587484121323,
      "num_tokens": 36225594.0,
      "step": 30160
    },
    {
      "entropy": 1.9522407323122024,
      "epoch": 0.09352436027493125,
      "grad_norm": 10.857767105102539,
      "learning_rate": 8.272513870063048e-06,
      "loss": 0.6238,
      "mean_token_accuracy": 0.8080313906073571,
      "num_tokens": 36236730.0,
      "step": 30170
    },
    {
      "entropy": 1.879507339000702,
      "epoch": 0.09355535939998094,
      "grad_norm": 4.612288951873779,
      "learning_rate": 8.271143181956931e-06,
      "loss": 0.561,
      "mean_token_accuracy": 0.8234292283654213,
      "num_tokens": 36249271.0,
      "step": 30180
    },
    {
      "entropy": 1.904217940568924,
      "epoch": 0.09358635852503062,
      "grad_norm": 9.96085262298584,
      "learning_rate": 8.269773174960643e-06,
      "loss": 0.6201,
      "mean_token_accuracy": 0.816309979557991,
      "num_tokens": 36261216.0,
      "step": 30190
    },
    {
      "entropy": 1.9194680109620095,
      "epoch": 0.09361735765008032,
      "grad_norm": 4.954514503479004,
      "learning_rate": 8.268403848510283e-06,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8259898975491524,
      "num_tokens": 36272893.0,
      "step": 30200
    },
    {
      "entropy": 1.9340100079774856,
      "epoch": 0.09364835677513002,
      "grad_norm": 11.47993278503418,
      "learning_rate": 8.267035202042611e-06,
      "loss": 0.6081,
      "mean_token_accuracy": 0.813405393064022,
      "num_tokens": 36284689.0,
      "step": 30210
    },
    {
      "entropy": 1.904521045088768,
      "epoch": 0.09367935590017971,
      "grad_norm": 9.563363075256348,
      "learning_rate": 8.265667234995031e-06,
      "loss": 0.595,
      "mean_token_accuracy": 0.8256731614470482,
      "num_tokens": 36295838.0,
      "step": 30220
    },
    {
      "entropy": 1.8696922466158867,
      "epoch": 0.09371035502522941,
      "grad_norm": 11.36976432800293,
      "learning_rate": 8.264299946805606e-06,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8290983706712722,
      "num_tokens": 36307929.0,
      "step": 30230
    },
    {
      "entropy": 1.9623483926057816,
      "epoch": 0.09374135415027911,
      "grad_norm": 9.148504257202148,
      "learning_rate": 8.26293333691304e-06,
      "loss": 0.6233,
      "mean_token_accuracy": 0.8086713761091232,
      "num_tokens": 36319244.0,
      "step": 30240
    },
    {
      "entropy": 1.9141168981790542,
      "epoch": 0.0937723532753288,
      "grad_norm": 8.628009796142578,
      "learning_rate": 8.261567404756697e-06,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8221844986081124,
      "num_tokens": 36330692.0,
      "step": 30250
    },
    {
      "entropy": 1.898746033012867,
      "epoch": 0.0938033524003785,
      "grad_norm": 10.499122619628906,
      "learning_rate": 8.260202149776582e-06,
      "loss": 0.5483,
      "mean_token_accuracy": 0.8232525825500489,
      "num_tokens": 36342656.0,
      "step": 30260
    },
    {
      "entropy": 1.8459609940648078,
      "epoch": 0.0938343515254282,
      "grad_norm": 4.076352596282959,
      "learning_rate": 8.258837571413353e-06,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8242526456713677,
      "num_tokens": 36355602.0,
      "step": 30270
    },
    {
      "entropy": 1.870631681382656,
      "epoch": 0.0938653506504779,
      "grad_norm": 9.26065731048584,
      "learning_rate": 8.25747366910831e-06,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8275643572211265,
      "num_tokens": 36368347.0,
      "step": 30280
    },
    {
      "entropy": 1.9731937110424043,
      "epoch": 0.09389634977552759,
      "grad_norm": 10.137398719787598,
      "learning_rate": 8.256110442303401e-06,
      "loss": 0.6493,
      "mean_token_accuracy": 0.8186086341738701,
      "num_tokens": 36378870.0,
      "step": 30290
    },
    {
      "entropy": 1.9244423538446427,
      "epoch": 0.09392734890057729,
      "grad_norm": 8.93706226348877,
      "learning_rate": 8.254747890441217e-06,
      "loss": 0.5874,
      "mean_token_accuracy": 0.8153250128030777,
      "num_tokens": 36389925.0,
      "step": 30300
    },
    {
      "entropy": 1.956382469832897,
      "epoch": 0.09395834802562697,
      "grad_norm": 8.536420822143555,
      "learning_rate": 8.253386012964996e-06,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8204043418169021,
      "num_tokens": 36401021.0,
      "step": 30310
    },
    {
      "entropy": 1.947486911714077,
      "epoch": 0.09398934715067667,
      "grad_norm": 9.407525062561035,
      "learning_rate": 8.252024809318618e-06,
      "loss": 0.6199,
      "mean_token_accuracy": 0.8143399626016616,
      "num_tokens": 36412021.0,
      "step": 30320
    },
    {
      "entropy": 1.9774963051080703,
      "epoch": 0.09402034627572636,
      "grad_norm": 9.66044807434082,
      "learning_rate": 8.250664278946598e-06,
      "loss": 0.6574,
      "mean_token_accuracy": 0.8052627012133599,
      "num_tokens": 36422783.0,
      "step": 30330
    },
    {
      "entropy": 1.9370567843317985,
      "epoch": 0.09405134540077606,
      "grad_norm": 10.304049491882324,
      "learning_rate": 8.249304421294103e-06,
      "loss": 0.617,
      "mean_token_accuracy": 0.8105893135070801,
      "num_tokens": 36434720.0,
      "step": 30340
    },
    {
      "entropy": 1.8703087165951728,
      "epoch": 0.09408234452582576,
      "grad_norm": 9.763945579528809,
      "learning_rate": 8.247945235806933e-06,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8189043670892715,
      "num_tokens": 36447678.0,
      "step": 30350
    },
    {
      "entropy": 1.9350792229175569,
      "epoch": 0.09411334365087545,
      "grad_norm": 8.846423149108887,
      "learning_rate": 8.246586721931527e-06,
      "loss": 0.6098,
      "mean_token_accuracy": 0.8202366888523102,
      "num_tokens": 36459700.0,
      "step": 30360
    },
    {
      "entropy": 1.9018134266138076,
      "epoch": 0.09414434277592515,
      "grad_norm": 8.516942977905273,
      "learning_rate": 8.245228879114964e-06,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8454120337963105,
      "num_tokens": 36471172.0,
      "step": 30370
    },
    {
      "entropy": 1.877271145582199,
      "epoch": 0.09417534190097485,
      "grad_norm": 9.023475646972656,
      "learning_rate": 8.24387170680496e-06,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8345570683479309,
      "num_tokens": 36483107.0,
      "step": 30380
    },
    {
      "entropy": 1.8964007556438447,
      "epoch": 0.09420634102602454,
      "grad_norm": 8.33362102508545,
      "learning_rate": 8.242515204449868e-06,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8253947287797928,
      "num_tokens": 36494970.0,
      "step": 30390
    },
    {
      "entropy": 2.000899037718773,
      "epoch": 0.09423734015107424,
      "grad_norm": 8.705028533935547,
      "learning_rate": 8.241159371498669e-06,
      "loss": 0.669,
      "mean_token_accuracy": 0.8023026213049889,
      "num_tokens": 36505728.0,
      "step": 30400
    },
    {
      "entropy": 1.9049580052495003,
      "epoch": 0.09426833927612394,
      "grad_norm": 9.36442756652832,
      "learning_rate": 8.23980420740099e-06,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8285997763276101,
      "num_tokens": 36517575.0,
      "step": 30410
    },
    {
      "entropy": 1.9479586511850357,
      "epoch": 0.09429933840117363,
      "grad_norm": 4.493463039398193,
      "learning_rate": 8.238449711607085e-06,
      "loss": 0.6312,
      "mean_token_accuracy": 0.8059987321496009,
      "num_tokens": 36529609.0,
      "step": 30420
    },
    {
      "entropy": 1.8796984627842903,
      "epoch": 0.09433033752622333,
      "grad_norm": 5.25566291809082,
      "learning_rate": 8.237095883567837e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8322527810931206,
      "num_tokens": 36541818.0,
      "step": 30430
    },
    {
      "entropy": 1.8748697608709335,
      "epoch": 0.09436133665127301,
      "grad_norm": 7.110718250274658,
      "learning_rate": 8.235742722734768e-06,
      "loss": 0.6065,
      "mean_token_accuracy": 0.820853716135025,
      "num_tokens": 36554237.0,
      "step": 30440
    },
    {
      "entropy": 1.991670235991478,
      "epoch": 0.09439233577632271,
      "grad_norm": 9.52397632598877,
      "learning_rate": 8.234390228560024e-06,
      "loss": 0.6539,
      "mean_token_accuracy": 0.8077353879809379,
      "num_tokens": 36565127.0,
      "step": 30450
    },
    {
      "entropy": 1.8927300944924355,
      "epoch": 0.0944233349013724,
      "grad_norm": 10.027509689331055,
      "learning_rate": 8.233038400496384e-06,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8261799916625023,
      "num_tokens": 36576878.0,
      "step": 30460
    },
    {
      "entropy": 1.9108740240335464,
      "epoch": 0.0944543340264221,
      "grad_norm": 4.891687870025635,
      "learning_rate": 8.231687237997258e-06,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8279662787914276,
      "num_tokens": 36588895.0,
      "step": 30470
    },
    {
      "entropy": 1.8933662503957749,
      "epoch": 0.0944853331514718,
      "grad_norm": 9.058094024658203,
      "learning_rate": 8.230336740516675e-06,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8237976714968681,
      "num_tokens": 36600847.0,
      "step": 30480
    },
    {
      "entropy": 1.9511252835392952,
      "epoch": 0.0945163322765215,
      "grad_norm": 3.954777479171753,
      "learning_rate": 8.228986907509301e-06,
      "loss": 0.598,
      "mean_token_accuracy": 0.8181298822164536,
      "num_tokens": 36612439.0,
      "step": 30490
    },
    {
      "entropy": 1.9695148766040802,
      "epoch": 0.09454733140157119,
      "grad_norm": 9.158353805541992,
      "learning_rate": 8.227637738430418e-06,
      "loss": 0.6427,
      "mean_token_accuracy": 0.812971468269825,
      "num_tokens": 36623202.0,
      "step": 30500
    },
    {
      "entropy": 1.9591504886746407,
      "epoch": 0.09457833052662089,
      "grad_norm": 9.691153526306152,
      "learning_rate": 8.226289232735947e-06,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8101043596863746,
      "num_tokens": 36635035.0,
      "step": 30510
    },
    {
      "entropy": 1.8547694399952888,
      "epoch": 0.09460932965167058,
      "grad_norm": 7.980339527130127,
      "learning_rate": 8.224941389882417e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8369179427623749,
      "num_tokens": 36647914.0,
      "step": 30520
    },
    {
      "entropy": 1.7833900511264802,
      "epoch": 0.09464032877672028,
      "grad_norm": 9.225784301757812,
      "learning_rate": 8.223594209326989e-06,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8371155127882958,
      "num_tokens": 36660801.0,
      "step": 30530
    },
    {
      "entropy": 1.966453444957733,
      "epoch": 0.09467132790176998,
      "grad_norm": 9.450788497924805,
      "learning_rate": 8.222247690527445e-06,
      "loss": 0.6112,
      "mean_token_accuracy": 0.8205273166298866,
      "num_tokens": 36671949.0,
      "step": 30540
    },
    {
      "entropy": 1.9008652031421662,
      "epoch": 0.09470232702681967,
      "grad_norm": 5.126152992248535,
      "learning_rate": 8.220901832942189e-06,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8248556137084961,
      "num_tokens": 36683882.0,
      "step": 30550
    },
    {
      "entropy": 1.8853500545024873,
      "epoch": 0.09473332615186936,
      "grad_norm": 9.585368156433105,
      "learning_rate": 8.219556636030243e-06,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8217202216386795,
      "num_tokens": 36695769.0,
      "step": 30560
    },
    {
      "entropy": 1.9309042051434517,
      "epoch": 0.09476432527691905,
      "grad_norm": 14.714326858520508,
      "learning_rate": 8.21821209925125e-06,
      "loss": 0.603,
      "mean_token_accuracy": 0.8186698570847512,
      "num_tokens": 36707233.0,
      "step": 30570
    },
    {
      "entropy": 1.7473974063992501,
      "epoch": 0.09479532440196875,
      "grad_norm": 10.051163673400879,
      "learning_rate": 8.21686822206547e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8436778649687767,
      "num_tokens": 36721694.0,
      "step": 30580
    },
    {
      "entropy": 1.8862526759505271,
      "epoch": 0.09482632352701845,
      "grad_norm": 11.326528549194336,
      "learning_rate": 8.215525003933785e-06,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8218036189675331,
      "num_tokens": 36734217.0,
      "step": 30590
    },
    {
      "entropy": 1.9055613458156586,
      "epoch": 0.09485732265206814,
      "grad_norm": 9.264266014099121,
      "learning_rate": 8.214182444317686e-06,
      "loss": 0.5895,
      "mean_token_accuracy": 0.821314187347889,
      "num_tokens": 36746342.0,
      "step": 30600
    },
    {
      "entropy": 1.8917794667184353,
      "epoch": 0.09488832177711784,
      "grad_norm": 3.4011175632476807,
      "learning_rate": 8.21284054267929e-06,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8364427119493485,
      "num_tokens": 36758768.0,
      "step": 30610
    },
    {
      "entropy": 1.9201584607362747,
      "epoch": 0.09491932090216754,
      "grad_norm": 9.317156791687012,
      "learning_rate": 8.211499298481317e-06,
      "loss": 0.5849,
      "mean_token_accuracy": 0.8274592280387878,
      "num_tokens": 36770047.0,
      "step": 30620
    },
    {
      "entropy": 1.872726395726204,
      "epoch": 0.09495032002721723,
      "grad_norm": 4.236084938049316,
      "learning_rate": 8.210158711187111e-06,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8388981744647026,
      "num_tokens": 36782594.0,
      "step": 30630
    },
    {
      "entropy": 1.9075644299387933,
      "epoch": 0.09498131915226693,
      "grad_norm": 8.899432182312012,
      "learning_rate": 8.208818780260624e-06,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8260029882192612,
      "num_tokens": 36793757.0,
      "step": 30640
    },
    {
      "entropy": 1.878479714691639,
      "epoch": 0.09501231827731663,
      "grad_norm": 10.720561027526855,
      "learning_rate": 8.207479505166421e-06,
      "loss": 0.5665,
      "mean_token_accuracy": 0.822782176733017,
      "num_tokens": 36806341.0,
      "step": 30650
    },
    {
      "entropy": 1.7969744451344014,
      "epoch": 0.09504331740236632,
      "grad_norm": 1.9851399660110474,
      "learning_rate": 8.206140885369683e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8433524951338768,
      "num_tokens": 36819861.0,
      "step": 30660
    },
    {
      "entropy": 1.973240813612938,
      "epoch": 0.09507431652741602,
      "grad_norm": 10.891983985900879,
      "learning_rate": 8.20480292033619e-06,
      "loss": 0.6501,
      "mean_token_accuracy": 0.8093028724193573,
      "num_tokens": 36830803.0,
      "step": 30670
    },
    {
      "entropy": 1.9500004380941391,
      "epoch": 0.0951053156524657,
      "grad_norm": 8.56579303741455,
      "learning_rate": 8.203465609532345e-06,
      "loss": 0.641,
      "mean_token_accuracy": 0.817388865351677,
      "num_tokens": 36842015.0,
      "step": 30680
    },
    {
      "entropy": 1.945586933195591,
      "epoch": 0.0951363147775154,
      "grad_norm": 10.45765495300293,
      "learning_rate": 8.20212895242515e-06,
      "loss": 0.6136,
      "mean_token_accuracy": 0.804017736017704,
      "num_tokens": 36853267.0,
      "step": 30690
    },
    {
      "entropy": 1.921870057284832,
      "epoch": 0.0951673139025651,
      "grad_norm": 9.248764038085938,
      "learning_rate": 8.20079294848222e-06,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8254543885588645,
      "num_tokens": 36864795.0,
      "step": 30700
    },
    {
      "entropy": 1.945303277671337,
      "epoch": 0.09519831302761479,
      "grad_norm": 8.114592552185059,
      "learning_rate": 8.199457597171774e-06,
      "loss": 0.5779,
      "mean_token_accuracy": 0.825435708463192,
      "num_tokens": 36875864.0,
      "step": 30710
    },
    {
      "entropy": 1.9745984852313996,
      "epoch": 0.09522931215266449,
      "grad_norm": 8.283961296081543,
      "learning_rate": 8.198122897962637e-06,
      "loss": 0.6276,
      "mean_token_accuracy": 0.813007952272892,
      "num_tokens": 36887565.0,
      "step": 30720
    },
    {
      "entropy": 1.971392062306404,
      "epoch": 0.09526031127771419,
      "grad_norm": 9.020630836486816,
      "learning_rate": 8.19678885032424e-06,
      "loss": 0.6255,
      "mean_token_accuracy": 0.8155330777168274,
      "num_tokens": 36899194.0,
      "step": 30730
    },
    {
      "entropy": 1.9447349101305007,
      "epoch": 0.09529131040276388,
      "grad_norm": 9.019769668579102,
      "learning_rate": 8.195455453726619e-06,
      "loss": 0.605,
      "mean_token_accuracy": 0.8189174398779869,
      "num_tokens": 36910579.0,
      "step": 30740
    },
    {
      "entropy": 1.9661620736122132,
      "epoch": 0.09532230952781358,
      "grad_norm": 5.993375778198242,
      "learning_rate": 8.194122707640413e-06,
      "loss": 0.6069,
      "mean_token_accuracy": 0.8191442266106606,
      "num_tokens": 36921589.0,
      "step": 30750
    },
    {
      "entropy": 1.8841305732727052,
      "epoch": 0.09535330865286328,
      "grad_norm": 3.033937692642212,
      "learning_rate": 8.19279061153686e-06,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8168802231550216,
      "num_tokens": 36934205.0,
      "step": 30760
    },
    {
      "entropy": 1.8415063828229905,
      "epoch": 0.09538430777791297,
      "grad_norm": 4.099487781524658,
      "learning_rate": 8.191459164887803e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.830090343952179,
      "num_tokens": 36946800.0,
      "step": 30770
    },
    {
      "entropy": 1.9041469410061835,
      "epoch": 0.09541530690296267,
      "grad_norm": 9.430980682373047,
      "learning_rate": 8.190128367165687e-06,
      "loss": 0.6402,
      "mean_token_accuracy": 0.8095077216625214,
      "num_tokens": 36958615.0,
      "step": 30780
    },
    {
      "entropy": 1.8373140662908554,
      "epoch": 0.09544630602801236,
      "grad_norm": 9.46648120880127,
      "learning_rate": 8.188798217843552e-06,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8348923206329346,
      "num_tokens": 36971891.0,
      "step": 30790
    },
    {
      "entropy": 1.8742023393511773,
      "epoch": 0.09547730515306206,
      "grad_norm": 3.767117738723755,
      "learning_rate": 8.187468716395042e-06,
      "loss": 0.522,
      "mean_token_accuracy": 0.8293885499238968,
      "num_tokens": 36984478.0,
      "step": 30800
    },
    {
      "entropy": 1.8809196785092355,
      "epoch": 0.09550830427811174,
      "grad_norm": 4.8539276123046875,
      "learning_rate": 8.186139862294395e-06,
      "loss": 0.576,
      "mean_token_accuracy": 0.818726259469986,
      "num_tokens": 36996951.0,
      "step": 30810
    },
    {
      "entropy": 1.835391464829445,
      "epoch": 0.09553930340316144,
      "grad_norm": 9.504169464111328,
      "learning_rate": 8.184811655016448e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8270650163292885,
      "num_tokens": 37010328.0,
      "step": 30820
    },
    {
      "entropy": 1.9224170967936516,
      "epoch": 0.09557030252821114,
      "grad_norm": 8.346292495727539,
      "learning_rate": 8.183484094036632e-06,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8288520753383637,
      "num_tokens": 37021973.0,
      "step": 30830
    },
    {
      "entropy": 1.9015969276428222,
      "epoch": 0.09560130165326083,
      "grad_norm": 11.159916877746582,
      "learning_rate": 8.182157178830978e-06,
      "loss": 0.5541,
      "mean_token_accuracy": 0.8173329204320907,
      "num_tokens": 37034374.0,
      "step": 30840
    },
    {
      "entropy": 1.914992319047451,
      "epoch": 0.09563230077831053,
      "grad_norm": 11.362371444702148,
      "learning_rate": 8.180830908876107e-06,
      "loss": 0.598,
      "mean_token_accuracy": 0.8112172558903694,
      "num_tokens": 37046586.0,
      "step": 30850
    },
    {
      "entropy": 1.869801890850067,
      "epoch": 0.09566329990336023,
      "grad_norm": 9.398642539978027,
      "learning_rate": 8.179505283649239e-06,
      "loss": 0.4976,
      "mean_token_accuracy": 0.8326647847890853,
      "num_tokens": 37059363.0,
      "step": 30860
    },
    {
      "entropy": 1.8593883782625198,
      "epoch": 0.09569429902840992,
      "grad_norm": 8.429418563842773,
      "learning_rate": 8.178180302628178e-06,
      "loss": 0.5187,
      "mean_token_accuracy": 0.833593225479126,
      "num_tokens": 37071727.0,
      "step": 30870
    },
    {
      "entropy": 1.8928765952587128,
      "epoch": 0.09572529815345962,
      "grad_norm": 8.581494331359863,
      "learning_rate": 8.176855965291328e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.836057162284851,
      "num_tokens": 37083853.0,
      "step": 30880
    },
    {
      "entropy": 1.854266819357872,
      "epoch": 0.09575629727850932,
      "grad_norm": 4.337390422821045,
      "learning_rate": 8.175532271117681e-06,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8306478515267373,
      "num_tokens": 37096495.0,
      "step": 30890
    },
    {
      "entropy": 1.7374467805027962,
      "epoch": 0.09578729640355901,
      "grad_norm": 8.350436210632324,
      "learning_rate": 8.17420921958682e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.843256051838398,
      "num_tokens": 37110413.0,
      "step": 30900
    },
    {
      "entropy": 1.956564149260521,
      "epoch": 0.09581829552860871,
      "grad_norm": 9.846826553344727,
      "learning_rate": 8.172886810178917e-06,
      "loss": 0.6155,
      "mean_token_accuracy": 0.8104242667555809,
      "num_tokens": 37121734.0,
      "step": 30910
    },
    {
      "entropy": 1.8714107498526573,
      "epoch": 0.0958492946536584,
      "grad_norm": 9.786186218261719,
      "learning_rate": 8.171565042374731e-06,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8231972604990005,
      "num_tokens": 37133847.0,
      "step": 30920
    },
    {
      "entropy": 1.92191222012043,
      "epoch": 0.09588029377870809,
      "grad_norm": 9.510786056518555,
      "learning_rate": 8.17024391565561e-06,
      "loss": 0.5702,
      "mean_token_accuracy": 0.825206808745861,
      "num_tokens": 37145862.0,
      "step": 30930
    },
    {
      "entropy": 1.9616800487041473,
      "epoch": 0.09591129290375779,
      "grad_norm": 10.04150676727295,
      "learning_rate": 8.168923429503489e-06,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8210392817854881,
      "num_tokens": 37156951.0,
      "step": 30940
    },
    {
      "entropy": 1.920756246894598,
      "epoch": 0.09594229202880748,
      "grad_norm": 3.579078435897827,
      "learning_rate": 8.167603583400891e-06,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8260960400104522,
      "num_tokens": 37169346.0,
      "step": 30950
    },
    {
      "entropy": 1.9267018765211106,
      "epoch": 0.09597329115385718,
      "grad_norm": 9.2460355758667,
      "learning_rate": 8.166284376830917e-06,
      "loss": 0.5963,
      "mean_token_accuracy": 0.8267333999276161,
      "num_tokens": 37181057.0,
      "step": 30960
    },
    {
      "entropy": 1.9822622686624527,
      "epoch": 0.09600429027890688,
      "grad_norm": 12.592954635620117,
      "learning_rate": 8.164965809277262e-06,
      "loss": 0.6087,
      "mean_token_accuracy": 0.8141214087605476,
      "num_tokens": 37192122.0,
      "step": 30970
    },
    {
      "entropy": 1.9343804091215133,
      "epoch": 0.09603528940395657,
      "grad_norm": 3.9561359882354736,
      "learning_rate": 8.163647880224195e-06,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8185720920562745,
      "num_tokens": 37203777.0,
      "step": 30980
    },
    {
      "entropy": 1.9694799184799194,
      "epoch": 0.09606628852900627,
      "grad_norm": 9.182045936584473,
      "learning_rate": 8.162330589156574e-06,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8282784789800643,
      "num_tokens": 37215743.0,
      "step": 30990
    },
    {
      "entropy": 1.8368911847472191,
      "epoch": 0.09609728765405597,
      "grad_norm": 4.183026313781738,
      "learning_rate": 8.161013935559836e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8347128361463547,
      "num_tokens": 37229304.0,
      "step": 31000
    },
    {
      "entropy": 1.8615291342139244,
      "epoch": 0.09612828677910566,
      "grad_norm": 10.485788345336914,
      "learning_rate": 8.15969791892e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8415818125009537,
      "num_tokens": 37241842.0,
      "step": 31010
    },
    {
      "entropy": 1.8837064534425736,
      "epoch": 0.09615928590415536,
      "grad_norm": 9.813204765319824,
      "learning_rate": 8.158382538723663e-06,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8298415571451188,
      "num_tokens": 37253894.0,
      "step": 31020
    },
    {
      "entropy": 1.9101071432232857,
      "epoch": 0.09619028502920506,
      "grad_norm": 9.026832580566406,
      "learning_rate": 8.157067794458002e-06,
      "loss": 0.5352,
      "mean_token_accuracy": 0.8250282123684883,
      "num_tokens": 37266281.0,
      "step": 31030
    },
    {
      "entropy": 1.9916753947734833,
      "epoch": 0.09622128415425475,
      "grad_norm": 8.812485694885254,
      "learning_rate": 8.155753685610777e-06,
      "loss": 0.6106,
      "mean_token_accuracy": 0.8189787149429322,
      "num_tokens": 37277277.0,
      "step": 31040
    },
    {
      "entropy": 1.8934721648693085,
      "epoch": 0.09625228327930443,
      "grad_norm": 4.702004909515381,
      "learning_rate": 8.154440211670315e-06,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8383319437503814,
      "num_tokens": 37288790.0,
      "step": 31050
    },
    {
      "entropy": 1.8943874284625053,
      "epoch": 0.09628328240435413,
      "grad_norm": 5.008129119873047,
      "learning_rate": 8.153127372125532e-06,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8305976986885071,
      "num_tokens": 37301541.0,
      "step": 31060
    },
    {
      "entropy": 1.9754459619522096,
      "epoch": 0.09631428152940383,
      "grad_norm": 8.563139915466309,
      "learning_rate": 8.151815166465911e-06,
      "loss": 0.6077,
      "mean_token_accuracy": 0.82650166451931,
      "num_tokens": 37313145.0,
      "step": 31070
    },
    {
      "entropy": 1.9236046463251113,
      "epoch": 0.09634528065445352,
      "grad_norm": 9.933487892150879,
      "learning_rate": 8.150503594181513e-06,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8255937352776528,
      "num_tokens": 37325171.0,
      "step": 31080
    },
    {
      "entropy": 1.8460835695266724,
      "epoch": 0.09637627977950322,
      "grad_norm": 4.579317569732666,
      "learning_rate": 8.149192654762971e-06,
      "loss": 0.5314,
      "mean_token_accuracy": 0.828292778134346,
      "num_tokens": 37338677.0,
      "step": 31090
    },
    {
      "entropy": 1.9498460546135903,
      "epoch": 0.09640727890455292,
      "grad_norm": 10.442451477050781,
      "learning_rate": 8.147882347701493e-06,
      "loss": 0.6204,
      "mean_token_accuracy": 0.818555298447609,
      "num_tokens": 37350287.0,
      "step": 31100
    },
    {
      "entropy": 1.890355022251606,
      "epoch": 0.09643827802960261,
      "grad_norm": 9.938260078430176,
      "learning_rate": 8.146572672488863e-06,
      "loss": 0.535,
      "mean_token_accuracy": 0.828145657479763,
      "num_tokens": 37361789.0,
      "step": 31110
    },
    {
      "entropy": 1.9762079477310182,
      "epoch": 0.09646927715465231,
      "grad_norm": 10.554993629455566,
      "learning_rate": 8.145263628617433e-06,
      "loss": 0.6357,
      "mean_token_accuracy": 0.8193916127085685,
      "num_tokens": 37372191.0,
      "step": 31120
    },
    {
      "entropy": 1.93240787088871,
      "epoch": 0.09650027627970201,
      "grad_norm": 8.852706909179688,
      "learning_rate": 8.143955215580123e-06,
      "loss": 0.6071,
      "mean_token_accuracy": 0.8182915225625038,
      "num_tokens": 37384812.0,
      "step": 31130
    },
    {
      "entropy": 2.0304384171962737,
      "epoch": 0.0965312754047517,
      "grad_norm": 10.083441734313965,
      "learning_rate": 8.142647432870427e-06,
      "loss": 0.6445,
      "mean_token_accuracy": 0.8097123354673386,
      "num_tokens": 37395995.0,
      "step": 31140
    },
    {
      "entropy": 1.9195516496896743,
      "epoch": 0.0965622745298014,
      "grad_norm": 8.50387954711914,
      "learning_rate": 8.141340279982408e-06,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8346219673752785,
      "num_tokens": 37407286.0,
      "step": 31150
    },
    {
      "entropy": 1.9149638175964356,
      "epoch": 0.0965932736548511,
      "grad_norm": 8.920450210571289,
      "learning_rate": 8.140033756410697e-06,
      "loss": 0.5605,
      "mean_token_accuracy": 0.8223188519477844,
      "num_tokens": 37419478.0,
      "step": 31160
    },
    {
      "entropy": 1.8281333968043327,
      "epoch": 0.0966242727799008,
      "grad_norm": 9.24412727355957,
      "learning_rate": 8.138727861650492e-06,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8335327923297882,
      "num_tokens": 37433099.0,
      "step": 31170
    },
    {
      "entropy": 1.8802730202674867,
      "epoch": 0.09665527190495048,
      "grad_norm": 6.880424976348877,
      "learning_rate": 8.137422595197554e-06,
      "loss": 0.588,
      "mean_token_accuracy": 0.8161317735910416,
      "num_tokens": 37445489.0,
      "step": 31180
    },
    {
      "entropy": 1.8878979504108429,
      "epoch": 0.09668627103000017,
      "grad_norm": 5.010068893432617,
      "learning_rate": 8.136117956548222e-06,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8283527597784996,
      "num_tokens": 37458005.0,
      "step": 31190
    },
    {
      "entropy": 1.88038859218359,
      "epoch": 0.09671727015504987,
      "grad_norm": 3.999337673187256,
      "learning_rate": 8.134813945199384e-06,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8222674325108528,
      "num_tokens": 37470320.0,
      "step": 31200
    },
    {
      "entropy": 1.926040168106556,
      "epoch": 0.09674826928009957,
      "grad_norm": 9.107236862182617,
      "learning_rate": 8.133510560648504e-06,
      "loss": 0.592,
      "mean_token_accuracy": 0.8095137163996696,
      "num_tokens": 37482370.0,
      "step": 31210
    },
    {
      "entropy": 1.8264732763171196,
      "epoch": 0.09677926840514926,
      "grad_norm": 8.244362831115723,
      "learning_rate": 8.132207802393603e-06,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8364429995417595,
      "num_tokens": 37495518.0,
      "step": 31220
    },
    {
      "entropy": 1.9302683904767037,
      "epoch": 0.09681026753019896,
      "grad_norm": 7.822775363922119,
      "learning_rate": 8.13090566993327e-06,
      "loss": 0.565,
      "mean_token_accuracy": 0.8201239988207817,
      "num_tokens": 37507302.0,
      "step": 31230
    },
    {
      "entropy": 1.8702651888132096,
      "epoch": 0.09684126665524866,
      "grad_norm": 4.487738132476807,
      "learning_rate": 8.12960416276665e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8344218656420708,
      "num_tokens": 37520315.0,
      "step": 31240
    },
    {
      "entropy": 1.9409600526094437,
      "epoch": 0.09687226578029835,
      "grad_norm": 9.113760948181152,
      "learning_rate": 8.128303280393453e-06,
      "loss": 0.5982,
      "mean_token_accuracy": 0.815177421271801,
      "num_tokens": 37532697.0,
      "step": 31250
    },
    {
      "entropy": 1.8489416658878326,
      "epoch": 0.09690326490534805,
      "grad_norm": 10.157899856567383,
      "learning_rate": 8.12700302231395e-06,
      "loss": 0.4969,
      "mean_token_accuracy": 0.8288843706250191,
      "num_tokens": 37545422.0,
      "step": 31260
    },
    {
      "entropy": 1.9021289080381394,
      "epoch": 0.09693426403039775,
      "grad_norm": 9.482295036315918,
      "learning_rate": 8.125703388028969e-06,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8187751770019531,
      "num_tokens": 37557355.0,
      "step": 31270
    },
    {
      "entropy": 1.8654734954237937,
      "epoch": 0.09696526315544744,
      "grad_norm": 10.133862495422363,
      "learning_rate": 8.124404377039897e-06,
      "loss": 0.5901,
      "mean_token_accuracy": 0.8247566968202591,
      "num_tokens": 37570124.0,
      "step": 31280
    },
    {
      "entropy": 1.9067189067602157,
      "epoch": 0.09699626228049714,
      "grad_norm": 8.202488899230957,
      "learning_rate": 8.123105988848677e-06,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8185949176549911,
      "num_tokens": 37582141.0,
      "step": 31290
    },
    {
      "entropy": 1.9307459384202956,
      "epoch": 0.09702726140554682,
      "grad_norm": 11.123476028442383,
      "learning_rate": 8.121808222957812e-06,
      "loss": 0.6175,
      "mean_token_accuracy": 0.8119331106543541,
      "num_tokens": 37592965.0,
      "step": 31300
    },
    {
      "entropy": 1.8705940045416356,
      "epoch": 0.09705826053059652,
      "grad_norm": 2.6256701946258545,
      "learning_rate": 8.120511078870361e-06,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8242152616381645,
      "num_tokens": 37605456.0,
      "step": 31310
    },
    {
      "entropy": 1.9379503265023232,
      "epoch": 0.09708925965564622,
      "grad_norm": 3.4866576194763184,
      "learning_rate": 8.119214556089939e-06,
      "loss": 0.5883,
      "mean_token_accuracy": 0.819600661098957,
      "num_tokens": 37616988.0,
      "step": 31320
    },
    {
      "entropy": 1.8673909679055214,
      "epoch": 0.09712025878069591,
      "grad_norm": 8.800174713134766,
      "learning_rate": 8.11791865412071e-06,
      "loss": 0.489,
      "mean_token_accuracy": 0.8349416226148605,
      "num_tokens": 37630013.0,
      "step": 31330
    },
    {
      "entropy": 1.9335881680250169,
      "epoch": 0.09715125790574561,
      "grad_norm": 9.899873733520508,
      "learning_rate": 8.1166233724674e-06,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8110765367746353,
      "num_tokens": 37641637.0,
      "step": 31340
    },
    {
      "entropy": 1.8817772254347802,
      "epoch": 0.0971822570307953,
      "grad_norm": 5.787439346313477,
      "learning_rate": 8.115328710635283e-06,
      "loss": 0.5509,
      "mean_token_accuracy": 0.8280811190605164,
      "num_tokens": 37653917.0,
      "step": 31350
    },
    {
      "entropy": 1.8765314996242524,
      "epoch": 0.097213256155845,
      "grad_norm": 8.655739784240723,
      "learning_rate": 8.114034668130184e-06,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8303320676088333,
      "num_tokens": 37665711.0,
      "step": 31360
    },
    {
      "entropy": 1.9583671048283577,
      "epoch": 0.0972442552808947,
      "grad_norm": 8.754752159118652,
      "learning_rate": 8.112741244458482e-06,
      "loss": 0.614,
      "mean_token_accuracy": 0.8211699604988099,
      "num_tokens": 37677018.0,
      "step": 31370
    },
    {
      "entropy": 1.8967312201857567,
      "epoch": 0.0972752544059444,
      "grad_norm": 4.7562575340271,
      "learning_rate": 8.11144843912711e-06,
      "loss": 0.569,
      "mean_token_accuracy": 0.8239043042063713,
      "num_tokens": 37689120.0,
      "step": 31380
    },
    {
      "entropy": 1.952524009346962,
      "epoch": 0.09730625353099409,
      "grad_norm": 9.356344223022461,
      "learning_rate": 8.110156251643543e-06,
      "loss": 0.6207,
      "mean_token_accuracy": 0.8118321433663368,
      "num_tokens": 37700583.0,
      "step": 31390
    },
    {
      "entropy": 1.912900096178055,
      "epoch": 0.09733725265604379,
      "grad_norm": 9.547168731689453,
      "learning_rate": 8.10886468151581e-06,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8243383869528771,
      "num_tokens": 37712109.0,
      "step": 31400
    },
    {
      "entropy": 1.9011776894330978,
      "epoch": 0.09736825178109348,
      "grad_norm": 9.910781860351562,
      "learning_rate": 8.107573728252488e-06,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8336651280522347,
      "num_tokens": 37724208.0,
      "step": 31410
    },
    {
      "entropy": 1.8022936284542084,
      "epoch": 0.09739925090614318,
      "grad_norm": 10.076736450195312,
      "learning_rate": 8.106283391362702e-06,
      "loss": 0.4303,
      "mean_token_accuracy": 0.8385085701942444,
      "num_tokens": 37738243.0,
      "step": 31420
    },
    {
      "entropy": 1.8668938994407653,
      "epoch": 0.09743025003119286,
      "grad_norm": 3.9053683280944824,
      "learning_rate": 8.10499367035612e-06,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8268329188227653,
      "num_tokens": 37750652.0,
      "step": 31430
    },
    {
      "entropy": 1.989102879166603,
      "epoch": 0.09746124915624256,
      "grad_norm": 8.179647445678711,
      "learning_rate": 8.10370456474296e-06,
      "loss": 0.6414,
      "mean_token_accuracy": 0.8093044266104699,
      "num_tokens": 37761475.0,
      "step": 31440
    },
    {
      "entropy": 1.9298017874360085,
      "epoch": 0.09749224828129226,
      "grad_norm": 9.899726867675781,
      "learning_rate": 8.102416074033986e-06,
      "loss": 0.5832,
      "mean_token_accuracy": 0.82451683729887,
      "num_tokens": 37773158.0,
      "step": 31450
    },
    {
      "entropy": 1.926207932829857,
      "epoch": 0.09752324740634195,
      "grad_norm": 10.14471435546875,
      "learning_rate": 8.101128197740498e-06,
      "loss": 0.6074,
      "mean_token_accuracy": 0.8211355611681939,
      "num_tokens": 37784675.0,
      "step": 31460
    },
    {
      "entropy": 1.9524447560310363,
      "epoch": 0.09755424653139165,
      "grad_norm": 10.060158729553223,
      "learning_rate": 8.09984093537435e-06,
      "loss": 0.547,
      "mean_token_accuracy": 0.8268745079636574,
      "num_tokens": 37795692.0,
      "step": 31470
    },
    {
      "entropy": 1.900469544529915,
      "epoch": 0.09758524565644135,
      "grad_norm": 8.439093589782715,
      "learning_rate": 8.098554286447932e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8321187406778335,
      "num_tokens": 37808059.0,
      "step": 31480
    },
    {
      "entropy": 1.931046536564827,
      "epoch": 0.09761624478149104,
      "grad_norm": 9.36986255645752,
      "learning_rate": 8.09726825047418e-06,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8171745136380195,
      "num_tokens": 37819692.0,
      "step": 31490
    },
    {
      "entropy": 1.9444507226347922,
      "epoch": 0.09764724390654074,
      "grad_norm": 10.288396835327148,
      "learning_rate": 8.095982826966572e-06,
      "loss": 0.5723,
      "mean_token_accuracy": 0.835323378443718,
      "num_tokens": 37831392.0,
      "step": 31500
    },
    {
      "entropy": 1.892457826435566,
      "epoch": 0.09767824303159044,
      "grad_norm": 9.328455924987793,
      "learning_rate": 8.094698015439117e-06,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8328243359923363,
      "num_tokens": 37842993.0,
      "step": 31510
    },
    {
      "entropy": 1.9004325211048125,
      "epoch": 0.09770924215664013,
      "grad_norm": 10.092707633972168,
      "learning_rate": 8.093413815406375e-06,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8365309268236161,
      "num_tokens": 37854723.0,
      "step": 31520
    },
    {
      "entropy": 1.8958807915449143,
      "epoch": 0.09774024128168983,
      "grad_norm": 10.082904815673828,
      "learning_rate": 8.092130226383442e-06,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8289808630943298,
      "num_tokens": 37866020.0,
      "step": 31530
    },
    {
      "entropy": 1.8704974979162217,
      "epoch": 0.09777124040673953,
      "grad_norm": 9.165472984313965,
      "learning_rate": 8.090847247885948e-06,
      "loss": 0.5996,
      "mean_token_accuracy": 0.8228276550769806,
      "num_tokens": 37877097.0,
      "step": 31540
    },
    {
      "entropy": 1.9054939955472947,
      "epoch": 0.09780223953178921,
      "grad_norm": 7.8041205406188965,
      "learning_rate": 8.089564879430065e-06,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8295403525233269,
      "num_tokens": 37887835.0,
      "step": 31550
    },
    {
      "entropy": 1.886386200785637,
      "epoch": 0.0978332386568389,
      "grad_norm": 8.552273750305176,
      "learning_rate": 8.088283120532499e-06,
      "loss": 0.5835,
      "mean_token_accuracy": 0.830622710287571,
      "num_tokens": 37898855.0,
      "step": 31560
    },
    {
      "entropy": 1.867349809408188,
      "epoch": 0.0978642377818886,
      "grad_norm": 4.293386936187744,
      "learning_rate": 8.087001970710495e-06,
      "loss": 0.6032,
      "mean_token_accuracy": 0.8121476486325264,
      "num_tokens": 37910879.0,
      "step": 31570
    },
    {
      "entropy": 1.9287435024976731,
      "epoch": 0.0978952369069383,
      "grad_norm": 10.854731559753418,
      "learning_rate": 8.085721429481825e-06,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8168462857604026,
      "num_tokens": 37921745.0,
      "step": 31580
    },
    {
      "entropy": 1.9377621293067933,
      "epoch": 0.097926236031988,
      "grad_norm": 11.060909271240234,
      "learning_rate": 8.084441496364808e-06,
      "loss": 0.6164,
      "mean_token_accuracy": 0.8070833861827851,
      "num_tokens": 37933377.0,
      "step": 31590
    },
    {
      "entropy": 1.854214581847191,
      "epoch": 0.09795723515703769,
      "grad_norm": 8.883403778076172,
      "learning_rate": 8.083162170878286e-06,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8322727754712105,
      "num_tokens": 37945413.0,
      "step": 31600
    },
    {
      "entropy": 1.8605719447135924,
      "epoch": 0.09798823428208739,
      "grad_norm": 8.319462776184082,
      "learning_rate": 8.081883452541636e-06,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8308003038167954,
      "num_tokens": 37956712.0,
      "step": 31610
    },
    {
      "entropy": 1.8450894683599472,
      "epoch": 0.09801923340713709,
      "grad_norm": 4.5528388023376465,
      "learning_rate": 8.080605340874773e-06,
      "loss": 0.6098,
      "mean_token_accuracy": 0.8137845560908318,
      "num_tokens": 37968907.0,
      "step": 31620
    },
    {
      "entropy": 1.864371307194233,
      "epoch": 0.09805023253218678,
      "grad_norm": 10.319393157958984,
      "learning_rate": 8.079327835398136e-06,
      "loss": 0.6086,
      "mean_token_accuracy": 0.8172882586717606,
      "num_tokens": 37980451.0,
      "step": 31630
    },
    {
      "entropy": 1.8554281413555145,
      "epoch": 0.09808123165723648,
      "grad_norm": 8.902338981628418,
      "learning_rate": 8.078050935632698e-06,
      "loss": 0.5693,
      "mean_token_accuracy": 0.831707838177681,
      "num_tokens": 37992222.0,
      "step": 31640
    },
    {
      "entropy": 1.8842545390129088,
      "epoch": 0.09811223078228618,
      "grad_norm": 8.492579460144043,
      "learning_rate": 8.076774641099962e-06,
      "loss": 0.5778,
      "mean_token_accuracy": 0.8209319055080414,
      "num_tokens": 38003647.0,
      "step": 31650
    },
    {
      "entropy": 1.8265250965952873,
      "epoch": 0.09814322990733587,
      "grad_norm": 10.900943756103516,
      "learning_rate": 8.075498951321958e-06,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8293390095233917,
      "num_tokens": 38016248.0,
      "step": 31660
    },
    {
      "entropy": 1.9135718867182732,
      "epoch": 0.09817422903238555,
      "grad_norm": 9.381340026855469,
      "learning_rate": 8.074223865821245e-06,
      "loss": 0.5587,
      "mean_token_accuracy": 0.8256983861327172,
      "num_tokens": 38027814.0,
      "step": 31670
    },
    {
      "entropy": 1.851310819387436,
      "epoch": 0.09820522815743525,
      "grad_norm": 11.365713119506836,
      "learning_rate": 8.072949384120915e-06,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8150163918733597,
      "num_tokens": 38040446.0,
      "step": 31680
    },
    {
      "entropy": 1.9479857064783572,
      "epoch": 0.09823622728248495,
      "grad_norm": 10.72806167602539,
      "learning_rate": 8.071675505744575e-06,
      "loss": 0.6235,
      "mean_token_accuracy": 0.8115170300006866,
      "num_tokens": 38051965.0,
      "step": 31690
    },
    {
      "entropy": 1.9195650681853293,
      "epoch": 0.09826722640753464,
      "grad_norm": 8.423272132873535,
      "learning_rate": 8.070402230216367e-06,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8288802221417427,
      "num_tokens": 38063124.0,
      "step": 31700
    },
    {
      "entropy": 1.8054373532533645,
      "epoch": 0.09829822553258434,
      "grad_norm": 9.05594253540039,
      "learning_rate": 8.06912955706096e-06,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8311830461025238,
      "num_tokens": 38076115.0,
      "step": 31710
    },
    {
      "entropy": 1.80671064555645,
      "epoch": 0.09832922465763404,
      "grad_norm": 7.974825382232666,
      "learning_rate": 8.067857485803538e-06,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8289128035306931,
      "num_tokens": 38089278.0,
      "step": 31720
    },
    {
      "entropy": 1.782078829407692,
      "epoch": 0.09836022378268373,
      "grad_norm": 11.095165252685547,
      "learning_rate": 8.066586015969819e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8267631024122238,
      "num_tokens": 38102671.0,
      "step": 31730
    },
    {
      "entropy": 1.9678689301013947,
      "epoch": 0.09839122290773343,
      "grad_norm": 11.549463272094727,
      "learning_rate": 8.065315147086036e-06,
      "loss": 0.666,
      "mean_token_accuracy": 0.8048723593354226,
      "num_tokens": 38113699.0,
      "step": 31740
    },
    {
      "entropy": 1.836279061436653,
      "epoch": 0.09842222203278313,
      "grad_norm": 9.380285263061523,
      "learning_rate": 8.06404487867895e-06,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8251174956560134,
      "num_tokens": 38127036.0,
      "step": 31750
    },
    {
      "entropy": 1.887456201016903,
      "epoch": 0.09845322115783282,
      "grad_norm": 9.466583251953125,
      "learning_rate": 8.062775210275841e-06,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8264592796564102,
      "num_tokens": 38137940.0,
      "step": 31760
    },
    {
      "entropy": 1.870304611325264,
      "epoch": 0.09848422028288252,
      "grad_norm": 9.614102363586426,
      "learning_rate": 8.061506141404512e-06,
      "loss": 0.5776,
      "mean_token_accuracy": 0.813577763736248,
      "num_tokens": 38150705.0,
      "step": 31770
    },
    {
      "entropy": 1.8906529873609543,
      "epoch": 0.09851521940793222,
      "grad_norm": 4.766526222229004,
      "learning_rate": 8.060237671593283e-06,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8281260386109353,
      "num_tokens": 38162785.0,
      "step": 31780
    },
    {
      "entropy": 1.817509751021862,
      "epoch": 0.09854621853298191,
      "grad_norm": 8.861818313598633,
      "learning_rate": 8.058969800370995e-06,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8299407482147216,
      "num_tokens": 38176287.0,
      "step": 31790
    },
    {
      "entropy": 1.9326122373342514,
      "epoch": 0.0985772176580316,
      "grad_norm": 8.393477439880371,
      "learning_rate": 8.057702527267008e-06,
      "loss": 0.586,
      "mean_token_accuracy": 0.8213211745023727,
      "num_tokens": 38187326.0,
      "step": 31800
    },
    {
      "entropy": 1.9347236022353171,
      "epoch": 0.09860821678308129,
      "grad_norm": 9.152666091918945,
      "learning_rate": 8.056435851811194e-06,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8280371204018593,
      "num_tokens": 38198420.0,
      "step": 31810
    },
    {
      "entropy": 1.9405176371335984,
      "epoch": 0.09863921590813099,
      "grad_norm": 10.188054084777832,
      "learning_rate": 8.055169773533956e-06,
      "loss": 0.6435,
      "mean_token_accuracy": 0.8134154111146927,
      "num_tokens": 38209643.0,
      "step": 31820
    },
    {
      "entropy": 1.8945807382464408,
      "epoch": 0.09867021503318069,
      "grad_norm": 8.68061637878418,
      "learning_rate": 8.053904291966199e-06,
      "loss": 0.5581,
      "mean_token_accuracy": 0.826158694922924,
      "num_tokens": 38221808.0,
      "step": 31830
    },
    {
      "entropy": 1.8253241881728173,
      "epoch": 0.09870121415823038,
      "grad_norm": 10.65565299987793,
      "learning_rate": 8.052639406639352e-06,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8176590099930763,
      "num_tokens": 38235228.0,
      "step": 31840
    },
    {
      "entropy": 1.9344543784856796,
      "epoch": 0.09873221328328008,
      "grad_norm": 8.646658897399902,
      "learning_rate": 8.051375117085356e-06,
      "loss": 0.6081,
      "mean_token_accuracy": 0.8185088708996773,
      "num_tokens": 38246052.0,
      "step": 31850
    },
    {
      "entropy": 1.8455456778407098,
      "epoch": 0.09876321240832978,
      "grad_norm": 8.64867115020752,
      "learning_rate": 8.050111422836666e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8226561039686203,
      "num_tokens": 38258783.0,
      "step": 31860
    },
    {
      "entropy": 1.942963680624962,
      "epoch": 0.09879421153337947,
      "grad_norm": 9.261329650878906,
      "learning_rate": 8.048848323426254e-06,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8249636858701705,
      "num_tokens": 38269528.0,
      "step": 31870
    },
    {
      "entropy": 1.9168901443481445,
      "epoch": 0.09882521065842917,
      "grad_norm": 9.246033668518066,
      "learning_rate": 8.047585818387599e-06,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8281188175082207,
      "num_tokens": 38280554.0,
      "step": 31880
    },
    {
      "entropy": 1.8628426641225815,
      "epoch": 0.09885620978347887,
      "grad_norm": 8.739521026611328,
      "learning_rate": 8.046323907254695e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.8306362301111221,
      "num_tokens": 38292518.0,
      "step": 31890
    },
    {
      "entropy": 1.8454672649502755,
      "epoch": 0.09888720890852856,
      "grad_norm": 12.10193920135498,
      "learning_rate": 8.045062589562051e-06,
      "loss": 0.5709,
      "mean_token_accuracy": 0.827955187857151,
      "num_tokens": 38304577.0,
      "step": 31900
    },
    {
      "entropy": 1.9152088075876237,
      "epoch": 0.09891820803357826,
      "grad_norm": 9.613424301147461,
      "learning_rate": 8.04380186484468e-06,
      "loss": 0.6403,
      "mean_token_accuracy": 0.8144320785999298,
      "num_tokens": 38315941.0,
      "step": 31910
    },
    {
      "entropy": 1.9276456892490388,
      "epoch": 0.09894920715862794,
      "grad_norm": 9.09563159942627,
      "learning_rate": 8.04254173263811e-06,
      "loss": 0.557,
      "mean_token_accuracy": 0.8385657519102097,
      "num_tokens": 38326358.0,
      "step": 31920
    },
    {
      "entropy": 1.8148802295327187,
      "epoch": 0.09898020628367764,
      "grad_norm": 8.845330238342285,
      "learning_rate": 8.041282192478376e-06,
      "loss": 0.5189,
      "mean_token_accuracy": 0.8373483180999756,
      "num_tokens": 38339092.0,
      "step": 31930
    },
    {
      "entropy": 1.8389425054192543,
      "epoch": 0.09901120540872733,
      "grad_norm": 8.750908851623535,
      "learning_rate": 8.040023243902018e-06,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8151543363928795,
      "num_tokens": 38351190.0,
      "step": 31940
    },
    {
      "entropy": 1.8402899622917175,
      "epoch": 0.09904220453377703,
      "grad_norm": 4.322279930114746,
      "learning_rate": 8.038764886446095e-06,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8239884585142135,
      "num_tokens": 38363817.0,
      "step": 31950
    },
    {
      "entropy": 1.8671101585030556,
      "epoch": 0.09907320365882673,
      "grad_norm": 4.702272415161133,
      "learning_rate": 8.037507119648157e-06,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8226954385638237,
      "num_tokens": 38376802.0,
      "step": 31960
    },
    {
      "entropy": 1.9011655792593956,
      "epoch": 0.09910420278387642,
      "grad_norm": 9.228616714477539,
      "learning_rate": 8.036249943046277e-06,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8301041334867477,
      "num_tokens": 38388556.0,
      "step": 31970
    },
    {
      "entropy": 1.854082614183426,
      "epoch": 0.09913520190892612,
      "grad_norm": 4.278061389923096,
      "learning_rate": 8.034993356179019e-06,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8298392608761788,
      "num_tokens": 38400576.0,
      "step": 31980
    },
    {
      "entropy": 1.8935617864131928,
      "epoch": 0.09916620103397582,
      "grad_norm": 8.677916526794434,
      "learning_rate": 8.03373735858546e-06,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8281985089182854,
      "num_tokens": 38412502.0,
      "step": 31990
    },
    {
      "entropy": 1.878693199157715,
      "epoch": 0.09919720015902551,
      "grad_norm": 9.149133682250977,
      "learning_rate": 8.032481949805182e-06,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8354146972298622,
      "num_tokens": 38423360.0,
      "step": 32000
    },
    {
      "entropy": 1.9500204533338548,
      "epoch": 0.09922819928407521,
      "grad_norm": 10.257466316223145,
      "learning_rate": 8.031227129378268e-06,
      "loss": 0.6755,
      "mean_token_accuracy": 0.8080421656370163,
      "num_tokens": 38433951.0,
      "step": 32010
    },
    {
      "entropy": 1.9539231777191162,
      "epoch": 0.09925919840912491,
      "grad_norm": 9.122477531433105,
      "learning_rate": 8.029972896845298e-06,
      "loss": 0.5941,
      "mean_token_accuracy": 0.8169040486216546,
      "num_tokens": 38445236.0,
      "step": 32020
    },
    {
      "entropy": 1.846830153465271,
      "epoch": 0.0992901975341746,
      "grad_norm": 9.854817390441895,
      "learning_rate": 8.028719251747369e-06,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8273584708571434,
      "num_tokens": 38457779.0,
      "step": 32030
    },
    {
      "entropy": 1.9177356496453286,
      "epoch": 0.09932119665922429,
      "grad_norm": 8.14139461517334,
      "learning_rate": 8.027466193626063e-06,
      "loss": 0.563,
      "mean_token_accuracy": 0.8339667573571206,
      "num_tokens": 38469104.0,
      "step": 32040
    },
    {
      "entropy": 1.9175456002354623,
      "epoch": 0.09935219578427398,
      "grad_norm": 4.319037914276123,
      "learning_rate": 8.026213722023473e-06,
      "loss": 0.5978,
      "mean_token_accuracy": 0.8246249735355378,
      "num_tokens": 38480992.0,
      "step": 32050
    },
    {
      "entropy": 1.9631080061197281,
      "epoch": 0.09938319490932368,
      "grad_norm": 9.7236328125,
      "learning_rate": 8.024961836482187e-06,
      "loss": 0.6166,
      "mean_token_accuracy": 0.8274016365408897,
      "num_tokens": 38492002.0,
      "step": 32060
    },
    {
      "entropy": 1.943944238126278,
      "epoch": 0.09941419403437338,
      "grad_norm": 8.304064750671387,
      "learning_rate": 8.023710536545295e-06,
      "loss": 0.6108,
      "mean_token_accuracy": 0.8115967348217964,
      "num_tokens": 38502940.0,
      "step": 32070
    },
    {
      "entropy": 1.9384367182850837,
      "epoch": 0.09944519315942307,
      "grad_norm": 11.586662292480469,
      "learning_rate": 8.022459821756386e-06,
      "loss": 0.6286,
      "mean_token_accuracy": 0.820226113498211,
      "num_tokens": 38514238.0,
      "step": 32080
    },
    {
      "entropy": 1.9123076066374778,
      "epoch": 0.09947619228447277,
      "grad_norm": 9.793305397033691,
      "learning_rate": 8.021209691659546e-06,
      "loss": 0.5969,
      "mean_token_accuracy": 0.8114640265703201,
      "num_tokens": 38525572.0,
      "step": 32090
    },
    {
      "entropy": 1.8171996608376504,
      "epoch": 0.09950719140952247,
      "grad_norm": 4.4752020835876465,
      "learning_rate": 8.019960145799353e-06,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8394015192985534,
      "num_tokens": 38538612.0,
      "step": 32100
    },
    {
      "entropy": 1.8743107169866562,
      "epoch": 0.09953819053457216,
      "grad_norm": 9.059576034545898,
      "learning_rate": 8.01871118372089e-06,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8266194269061089,
      "num_tokens": 38550013.0,
      "step": 32110
    },
    {
      "entropy": 1.8382051154971122,
      "epoch": 0.09956918965962186,
      "grad_norm": 2.9260752201080322,
      "learning_rate": 8.017462804969733e-06,
      "loss": 0.566,
      "mean_token_accuracy": 0.837664969265461,
      "num_tokens": 38562520.0,
      "step": 32120
    },
    {
      "entropy": 1.946621085703373,
      "epoch": 0.09960018878467156,
      "grad_norm": 8.969979286193848,
      "learning_rate": 8.01621500909195e-06,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8193635001778603,
      "num_tokens": 38573702.0,
      "step": 32130
    },
    {
      "entropy": 1.9753321021795274,
      "epoch": 0.09963118790972125,
      "grad_norm": 8.497583389282227,
      "learning_rate": 8.014967795634104e-06,
      "loss": 0.6373,
      "mean_token_accuracy": 0.8141985774040222,
      "num_tokens": 38584725.0,
      "step": 32140
    },
    {
      "entropy": 1.9402645155787468,
      "epoch": 0.09966218703477095,
      "grad_norm": 10.377409934997559,
      "learning_rate": 8.013721164143257e-06,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8230377018451691,
      "num_tokens": 38596403.0,
      "step": 32150
    },
    {
      "entropy": 1.7853017404675484,
      "epoch": 0.09969318615982065,
      "grad_norm": 8.521496772766113,
      "learning_rate": 8.012475114166955e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8455995440483093,
      "num_tokens": 38610309.0,
      "step": 32160
    },
    {
      "entropy": 1.846723584830761,
      "epoch": 0.09972418528487033,
      "grad_norm": 5.919735908508301,
      "learning_rate": 8.011229645253245e-06,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8269700452685356,
      "num_tokens": 38623255.0,
      "step": 32170
    },
    {
      "entropy": 1.8817561469972133,
      "epoch": 0.09975518440992003,
      "grad_norm": 10.0753755569458,
      "learning_rate": 8.009984756950662e-06,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8250563368201256,
      "num_tokens": 38636156.0,
      "step": 32180
    },
    {
      "entropy": 1.8628202617168426,
      "epoch": 0.09978618353496972,
      "grad_norm": 9.534749031066895,
      "learning_rate": 8.008740448808228e-06,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8297192022204399,
      "num_tokens": 38648825.0,
      "step": 32190
    },
    {
      "entropy": 1.8859781965613365,
      "epoch": 0.09981718266001942,
      "grad_norm": 9.754186630249023,
      "learning_rate": 8.007496720375465e-06,
      "loss": 0.554,
      "mean_token_accuracy": 0.8234538331627845,
      "num_tokens": 38660622.0,
      "step": 32200
    },
    {
      "entropy": 1.8697243973612785,
      "epoch": 0.09984818178506912,
      "grad_norm": 5.172030448913574,
      "learning_rate": 8.006253571202375e-06,
      "loss": 0.6183,
      "mean_token_accuracy": 0.8093932062387467,
      "num_tokens": 38672976.0,
      "step": 32210
    },
    {
      "entropy": 1.7175804510712624,
      "epoch": 0.09987918091011881,
      "grad_norm": 8.345678329467773,
      "learning_rate": 8.005011000839453e-06,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8368314996361732,
      "num_tokens": 38687008.0,
      "step": 32220
    },
    {
      "entropy": 1.8780830636620522,
      "epoch": 0.09991018003516851,
      "grad_norm": 4.0767951011657715,
      "learning_rate": 8.003769008837679e-06,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8209434077143669,
      "num_tokens": 38698550.0,
      "step": 32230
    },
    {
      "entropy": 1.9147436633706092,
      "epoch": 0.0999411791602182,
      "grad_norm": 10.640585899353027,
      "learning_rate": 8.00252759474853e-06,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8163613364100456,
      "num_tokens": 38709869.0,
      "step": 32240
    },
    {
      "entropy": 1.8193964034318924,
      "epoch": 0.0999721782852679,
      "grad_norm": 10.641173362731934,
      "learning_rate": 8.001286758123959e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8270770683884621,
      "num_tokens": 38721887.0,
      "step": 32250
    },
    {
      "entropy": 1.7959134474396705,
      "epoch": 0.1000031774103176,
      "grad_norm": 7.74287748336792,
      "learning_rate": 8.000046498516408e-06,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8405986934900284,
      "num_tokens": 38733462.0,
      "step": 32260
    },
    {
      "entropy": 1.851724424213171,
      "epoch": 0.1000341765353673,
      "grad_norm": 10.603741645812988,
      "learning_rate": 7.998806815478807e-06,
      "loss": 0.6026,
      "mean_token_accuracy": 0.8136375203728676,
      "num_tokens": 38744694.0,
      "step": 32270
    },
    {
      "entropy": 1.88350567817688,
      "epoch": 0.10006517566041699,
      "grad_norm": 10.134400367736816,
      "learning_rate": 7.99756770856457e-06,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8215403065085412,
      "num_tokens": 38756425.0,
      "step": 32280
    },
    {
      "entropy": 1.9489238500595092,
      "epoch": 0.10009617478546667,
      "grad_norm": 9.280869483947754,
      "learning_rate": 7.996329177327595e-06,
      "loss": 0.6208,
      "mean_token_accuracy": 0.8188161373138427,
      "num_tokens": 38767890.0,
      "step": 32290
    },
    {
      "entropy": 1.9058944791555406,
      "epoch": 0.10012717391051637,
      "grad_norm": 4.553225994110107,
      "learning_rate": 7.995091221322265e-06,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8262131318449975,
      "num_tokens": 38779858.0,
      "step": 32300
    },
    {
      "entropy": 1.7887984693050385,
      "epoch": 0.10015817303556607,
      "grad_norm": 10.373005867004395,
      "learning_rate": 7.993853840103436e-06,
      "loss": 0.4868,
      "mean_token_accuracy": 0.8316322803497315,
      "num_tokens": 38792827.0,
      "step": 32310
    },
    {
      "entropy": 1.7936713725328446,
      "epoch": 0.10018917216061576,
      "grad_norm": 9.501617431640625,
      "learning_rate": 7.992617033226463e-06,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8299453228712081,
      "num_tokens": 38805462.0,
      "step": 32320
    },
    {
      "entropy": 1.9167312130331993,
      "epoch": 0.10022017128566546,
      "grad_norm": 9.644965171813965,
      "learning_rate": 7.991380800247169e-06,
      "loss": 0.5801,
      "mean_token_accuracy": 0.8267593100667,
      "num_tokens": 38816933.0,
      "step": 32330
    },
    {
      "entropy": 1.8285753324627876,
      "epoch": 0.10025117041071516,
      "grad_norm": 8.455294609069824,
      "learning_rate": 7.990145140721862e-06,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8261840432882309,
      "num_tokens": 38829859.0,
      "step": 32340
    },
    {
      "entropy": 1.89251828789711,
      "epoch": 0.10028216953576485,
      "grad_norm": 8.720006942749023,
      "learning_rate": 7.98891005420733e-06,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8221445247530937,
      "num_tokens": 38841339.0,
      "step": 32350
    },
    {
      "entropy": 1.845874121785164,
      "epoch": 0.10031316866081455,
      "grad_norm": 8.771798133850098,
      "learning_rate": 7.987675540260844e-06,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8211918070912361,
      "num_tokens": 38853479.0,
      "step": 32360
    },
    {
      "entropy": 1.8901546359062196,
      "epoch": 0.10034416778586425,
      "grad_norm": 6.715768337249756,
      "learning_rate": 7.986441598440147e-06,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8205464497208595,
      "num_tokens": 38865059.0,
      "step": 32370
    },
    {
      "entropy": 1.879352556169033,
      "epoch": 0.10037516691091394,
      "grad_norm": 9.13968563079834,
      "learning_rate": 7.985208228303463e-06,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8105901688337326,
      "num_tokens": 38876867.0,
      "step": 32380
    },
    {
      "entropy": 1.84751605540514,
      "epoch": 0.10040616603596364,
      "grad_norm": 8.604146003723145,
      "learning_rate": 7.983975429409497e-06,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8336275666952133,
      "num_tokens": 38889307.0,
      "step": 32390
    },
    {
      "entropy": 1.7912178918719293,
      "epoch": 0.10043716516101334,
      "grad_norm": 9.056184768676758,
      "learning_rate": 7.982743201317426e-06,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8329012975096702,
      "num_tokens": 38902802.0,
      "step": 32400
    },
    {
      "entropy": 1.8672286108136178,
      "epoch": 0.10046816428606302,
      "grad_norm": 8.324599266052246,
      "learning_rate": 7.981511543586906e-06,
      "loss": 0.5679,
      "mean_token_accuracy": 0.8331062749028206,
      "num_tokens": 38914048.0,
      "step": 32410
    },
    {
      "entropy": 1.8162163376808167,
      "epoch": 0.10049916341111272,
      "grad_norm": 2.5847465991973877,
      "learning_rate": 7.980280455778062e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.832746496796608,
      "num_tokens": 38926231.0,
      "step": 32420
    },
    {
      "entropy": 1.8801830425858497,
      "epoch": 0.10053016253616241,
      "grad_norm": 8.904119491577148,
      "learning_rate": 7.979049937451507e-06,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8272658884525299,
      "num_tokens": 38938316.0,
      "step": 32430
    },
    {
      "entropy": 1.9005606561899184,
      "epoch": 0.10056116166121211,
      "grad_norm": 5.724231719970703,
      "learning_rate": 7.977819988168313e-06,
      "loss": 0.5275,
      "mean_token_accuracy": 0.829011881351471,
      "num_tokens": 38950624.0,
      "step": 32440
    },
    {
      "entropy": 1.9463760763406754,
      "epoch": 0.1005921607862618,
      "grad_norm": 8.63336181640625,
      "learning_rate": 7.97659060749004e-06,
      "loss": 0.5908,
      "mean_token_accuracy": 0.82104711830616,
      "num_tokens": 38962016.0,
      "step": 32450
    },
    {
      "entropy": 1.8716831341385842,
      "epoch": 0.1006231599113115,
      "grad_norm": 10.11276912689209,
      "learning_rate": 7.975361794978705e-06,
      "loss": 0.5483,
      "mean_token_accuracy": 0.8303984850645065,
      "num_tokens": 38973912.0,
      "step": 32460
    },
    {
      "entropy": 1.8566598653793336,
      "epoch": 0.1006541590363612,
      "grad_norm": 7.346640586853027,
      "learning_rate": 7.974133550196811e-06,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8243806138634682,
      "num_tokens": 38985487.0,
      "step": 32470
    },
    {
      "entropy": 1.9269026920199395,
      "epoch": 0.1006851581614109,
      "grad_norm": 8.665224075317383,
      "learning_rate": 7.972905872707326e-06,
      "loss": 0.6133,
      "mean_token_accuracy": 0.8166972935199738,
      "num_tokens": 38997045.0,
      "step": 32480
    },
    {
      "entropy": 1.7712698966264724,
      "epoch": 0.10071615728646059,
      "grad_norm": 4.503955364227295,
      "learning_rate": 7.97167876207369e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8519754007458686,
      "num_tokens": 39010592.0,
      "step": 32490
    },
    {
      "entropy": 1.8833463504910468,
      "epoch": 0.10074715641151029,
      "grad_norm": 10.112072944641113,
      "learning_rate": 7.970452217859811e-06,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8291536673903466,
      "num_tokens": 39021828.0,
      "step": 32500
    },
    {
      "entropy": 1.9463322639465332,
      "epoch": 0.10077815553655999,
      "grad_norm": 11.4695405960083,
      "learning_rate": 7.96922623963007e-06,
      "loss": 0.6672,
      "mean_token_accuracy": 0.8071766123175621,
      "num_tokens": 39032954.0,
      "step": 32510
    },
    {
      "entropy": 1.830717845261097,
      "epoch": 0.10080915466160968,
      "grad_norm": 5.045051574707031,
      "learning_rate": 7.968000826949319e-06,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8264394268393517,
      "num_tokens": 39046531.0,
      "step": 32520
    },
    {
      "entropy": 1.8757388591766357,
      "epoch": 0.10084015378665938,
      "grad_norm": 5.511293888092041,
      "learning_rate": 7.96677597938287e-06,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8393940135836602,
      "num_tokens": 39059751.0,
      "step": 32530
    },
    {
      "entropy": 1.9044818967580794,
      "epoch": 0.10087115291170906,
      "grad_norm": 9.242785453796387,
      "learning_rate": 7.965551696496507e-06,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8220992356538772,
      "num_tokens": 39072505.0,
      "step": 32540
    },
    {
      "entropy": 1.9557638376951219,
      "epoch": 0.10090215203675876,
      "grad_norm": 9.517579078674316,
      "learning_rate": 7.964327977856484e-06,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8174139723181725,
      "num_tokens": 39083179.0,
      "step": 32550
    },
    {
      "entropy": 2.0160479575395582,
      "epoch": 0.10093315116180845,
      "grad_norm": 9.149009704589844,
      "learning_rate": 7.963104823029519e-06,
      "loss": 0.644,
      "mean_token_accuracy": 0.8098774880170823,
      "num_tokens": 39093653.0,
      "step": 32560
    },
    {
      "entropy": 1.9425939425826073,
      "epoch": 0.10096415028685815,
      "grad_norm": 8.069141387939453,
      "learning_rate": 7.961882231582794e-06,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8205993890762329,
      "num_tokens": 39105306.0,
      "step": 32570
    },
    {
      "entropy": 1.9585781380534173,
      "epoch": 0.10099514941190785,
      "grad_norm": 8.68270206451416,
      "learning_rate": 7.960660203083954e-06,
      "loss": 0.6058,
      "mean_token_accuracy": 0.8138043344020843,
      "num_tokens": 39117926.0,
      "step": 32580
    },
    {
      "entropy": 1.952934955060482,
      "epoch": 0.10102614853695754,
      "grad_norm": 9.11218547821045,
      "learning_rate": 7.959438737101118e-06,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8283255323767662,
      "num_tokens": 39130134.0,
      "step": 32590
    },
    {
      "entropy": 1.9153422683477401,
      "epoch": 0.10105714766200724,
      "grad_norm": 5.233044147491455,
      "learning_rate": 7.958217833202859e-06,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8188350781798363,
      "num_tokens": 39142127.0,
      "step": 32600
    },
    {
      "entropy": 1.8911504298448563,
      "epoch": 0.10108814678705694,
      "grad_norm": 4.27568244934082,
      "learning_rate": 7.956997490958216e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.8261435091495514,
      "num_tokens": 39154491.0,
      "step": 32610
    },
    {
      "entropy": 1.7859768718481064,
      "epoch": 0.10111914591210663,
      "grad_norm": 4.534674644470215,
      "learning_rate": 7.955777709936692e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8465276822447777,
      "num_tokens": 39168135.0,
      "step": 32620
    },
    {
      "entropy": 1.90359725356102,
      "epoch": 0.10115014503715633,
      "grad_norm": 7.7813496589660645,
      "learning_rate": 7.95455848970825e-06,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8201944395899773,
      "num_tokens": 39180412.0,
      "step": 32630
    },
    {
      "entropy": 1.924836564064026,
      "epoch": 0.10118114416220603,
      "grad_norm": 9.442864418029785,
      "learning_rate": 7.953339829843315e-06,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8264301627874374,
      "num_tokens": 39192257.0,
      "step": 32640
    },
    {
      "entropy": 1.9304316490888596,
      "epoch": 0.10121214328725572,
      "grad_norm": 3.983366012573242,
      "learning_rate": 7.952121729912772e-06,
      "loss": 0.612,
      "mean_token_accuracy": 0.8225965097546577,
      "num_tokens": 39203910.0,
      "step": 32650
    },
    {
      "entropy": 1.9028907686471939,
      "epoch": 0.1012431424123054,
      "grad_norm": 9.803033828735352,
      "learning_rate": 7.95090418948797e-06,
      "loss": 0.6047,
      "mean_token_accuracy": 0.8184835597872734,
      "num_tokens": 39216797.0,
      "step": 32660
    },
    {
      "entropy": 1.8274445042014122,
      "epoch": 0.1012741415373551,
      "grad_norm": 10.706036567687988,
      "learning_rate": 7.949687208140709e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8392621099948883,
      "num_tokens": 39229208.0,
      "step": 32670
    },
    {
      "entropy": 1.983510084450245,
      "epoch": 0.1013051406624048,
      "grad_norm": 9.402242660522461,
      "learning_rate": 7.948470785443254e-06,
      "loss": 0.6156,
      "mean_token_accuracy": 0.8155052870512008,
      "num_tokens": 39240609.0,
      "step": 32680
    },
    {
      "entropy": 1.8997413352131844,
      "epoch": 0.1013361397874545,
      "grad_norm": 11.164144515991211,
      "learning_rate": 7.947254920968327e-06,
      "loss": 0.5142,
      "mean_token_accuracy": 0.823790366947651,
      "num_tokens": 39252794.0,
      "step": 32690
    },
    {
      "entropy": 1.843451727926731,
      "epoch": 0.10136713891250419,
      "grad_norm": 9.909012794494629,
      "learning_rate": 7.946039614289105e-06,
      "loss": 0.546,
      "mean_token_accuracy": 0.8248791456222534,
      "num_tokens": 39265204.0,
      "step": 32700
    },
    {
      "entropy": 1.9523500233888627,
      "epoch": 0.10139813803755389,
      "grad_norm": 8.847643852233887,
      "learning_rate": 7.944824864979225e-06,
      "loss": 0.6042,
      "mean_token_accuracy": 0.8163961425423623,
      "num_tokens": 39276926.0,
      "step": 32710
    },
    {
      "entropy": 1.8452000051736832,
      "epoch": 0.10142913716260359,
      "grad_norm": 8.985699653625488,
      "learning_rate": 7.943610672612779e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8368108674883843,
      "num_tokens": 39288593.0,
      "step": 32720
    },
    {
      "entropy": 1.8509555876255035,
      "epoch": 0.10146013628765328,
      "grad_norm": 8.876476287841797,
      "learning_rate": 7.94239703676431e-06,
      "loss": 0.5477,
      "mean_token_accuracy": 0.8287522226572037,
      "num_tokens": 39301355.0,
      "step": 32730
    },
    {
      "entropy": 1.982967707514763,
      "epoch": 0.10149113541270298,
      "grad_norm": 11.262484550476074,
      "learning_rate": 7.941183957008825e-06,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8227292239665985,
      "num_tokens": 39312750.0,
      "step": 32740
    },
    {
      "entropy": 1.9494792476296425,
      "epoch": 0.10152213453775268,
      "grad_norm": 12.197062492370605,
      "learning_rate": 7.939971432921778e-06,
      "loss": 0.5947,
      "mean_token_accuracy": 0.8156418994069099,
      "num_tokens": 39323923.0,
      "step": 32750
    },
    {
      "entropy": 1.9586309775710107,
      "epoch": 0.10155313366280237,
      "grad_norm": 8.947646141052246,
      "learning_rate": 7.93875946407908e-06,
      "loss": 0.5582,
      "mean_token_accuracy": 0.8243048340082169,
      "num_tokens": 39335909.0,
      "step": 32760
    },
    {
      "entropy": 1.9357258334755898,
      "epoch": 0.10158413278785207,
      "grad_norm": 8.137892723083496,
      "learning_rate": 7.937548050057092e-06,
      "loss": 0.5899,
      "mean_token_accuracy": 0.8267382949590683,
      "num_tokens": 39347350.0,
      "step": 32770
    },
    {
      "entropy": 1.961161696910858,
      "epoch": 0.10161513191290175,
      "grad_norm": 9.436999320983887,
      "learning_rate": 7.936337190432627e-06,
      "loss": 0.5788,
      "mean_token_accuracy": 0.8202362656593323,
      "num_tokens": 39359043.0,
      "step": 32780
    },
    {
      "entropy": 1.9914384290575982,
      "epoch": 0.10164613103795145,
      "grad_norm": 8.315853118896484,
      "learning_rate": 7.935126884782958e-06,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8224820002913475,
      "num_tokens": 39370438.0,
      "step": 32790
    },
    {
      "entropy": 1.9369804859161377,
      "epoch": 0.10167713016300114,
      "grad_norm": 9.922943115234375,
      "learning_rate": 7.9339171326858e-06,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8243139579892158,
      "num_tokens": 39380976.0,
      "step": 32800
    },
    {
      "entropy": 1.9085180804133415,
      "epoch": 0.10170812928805084,
      "grad_norm": 9.743084907531738,
      "learning_rate": 7.932707933719321e-06,
      "loss": 0.614,
      "mean_token_accuracy": 0.8235570520162583,
      "num_tokens": 39392951.0,
      "step": 32810
    },
    {
      "entropy": 1.9496404856443406,
      "epoch": 0.10173912841310054,
      "grad_norm": 8.715404510498047,
      "learning_rate": 7.931499287462138e-06,
      "loss": 0.6393,
      "mean_token_accuracy": 0.811781495809555,
      "num_tokens": 39404058.0,
      "step": 32820
    },
    {
      "entropy": 1.9330283522605896,
      "epoch": 0.10177012753815023,
      "grad_norm": 9.045530319213867,
      "learning_rate": 7.930291193493323e-06,
      "loss": 0.5864,
      "mean_token_accuracy": 0.8231693536043168,
      "num_tokens": 39414589.0,
      "step": 32830
    },
    {
      "entropy": 2.0210474640130998,
      "epoch": 0.10180112666319993,
      "grad_norm": 7.992177486419678,
      "learning_rate": 7.929083651392389e-06,
      "loss": 0.6262,
      "mean_token_accuracy": 0.8037309676408768,
      "num_tokens": 39425755.0,
      "step": 32840
    },
    {
      "entropy": 1.9645818829536439,
      "epoch": 0.10183212578824963,
      "grad_norm": 9.055630683898926,
      "learning_rate": 7.9278766607393e-06,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8236428961157799,
      "num_tokens": 39437178.0,
      "step": 32850
    },
    {
      "entropy": 1.9607673034071922,
      "epoch": 0.10186312491329932,
      "grad_norm": 10.347228050231934,
      "learning_rate": 7.92667022111447e-06,
      "loss": 0.5655,
      "mean_token_accuracy": 0.8204041153192521,
      "num_tokens": 39449415.0,
      "step": 32860
    },
    {
      "entropy": 1.913771566748619,
      "epoch": 0.10189412403834902,
      "grad_norm": 10.000711441040039,
      "learning_rate": 7.92546433209876e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8246194154024125,
      "num_tokens": 39461394.0,
      "step": 32870
    },
    {
      "entropy": 1.9231099665164948,
      "epoch": 0.10192512316339872,
      "grad_norm": 5.153701305389404,
      "learning_rate": 7.92425899327347e-06,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8274278253316879,
      "num_tokens": 39473357.0,
      "step": 32880
    },
    {
      "entropy": 1.9601033926010132,
      "epoch": 0.10195612228844841,
      "grad_norm": 9.463569641113281,
      "learning_rate": 7.923054204220351e-06,
      "loss": 0.6141,
      "mean_token_accuracy": 0.8186487257480621,
      "num_tokens": 39484263.0,
      "step": 32890
    },
    {
      "entropy": 1.989539209008217,
      "epoch": 0.10198712141349811,
      "grad_norm": 11.15012264251709,
      "learning_rate": 7.921849964521603e-06,
      "loss": 0.5973,
      "mean_token_accuracy": 0.8308931365609169,
      "num_tokens": 39495754.0,
      "step": 32900
    },
    {
      "entropy": 1.8660317227244376,
      "epoch": 0.1020181205385478,
      "grad_norm": 9.449760437011719,
      "learning_rate": 7.92064627375986e-06,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8349911078810692,
      "num_tokens": 39508683.0,
      "step": 32910
    },
    {
      "entropy": 1.956910152733326,
      "epoch": 0.10204911966359749,
      "grad_norm": 8.791659355163574,
      "learning_rate": 7.919443131518211e-06,
      "loss": 0.546,
      "mean_token_accuracy": 0.8249326780438423,
      "num_tokens": 39520549.0,
      "step": 32920
    },
    {
      "entropy": 1.9195820301771165,
      "epoch": 0.10208011878864719,
      "grad_norm": 4.308915615081787,
      "learning_rate": 7.91824053738018e-06,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8237533152103425,
      "num_tokens": 39533438.0,
      "step": 32930
    },
    {
      "entropy": 2.001399652659893,
      "epoch": 0.10211111791369688,
      "grad_norm": 11.334473609924316,
      "learning_rate": 7.917038490929737e-06,
      "loss": 0.5865,
      "mean_token_accuracy": 0.8264929071068764,
      "num_tokens": 39544650.0,
      "step": 32940
    },
    {
      "entropy": 1.9738114327192307,
      "epoch": 0.10214211703874658,
      "grad_norm": 4.6155595779418945,
      "learning_rate": 7.915836991751293e-06,
      "loss": 0.6039,
      "mean_token_accuracy": 0.8131748422980308,
      "num_tokens": 39556308.0,
      "step": 32950
    },
    {
      "entropy": 1.884914068877697,
      "epoch": 0.10217311616379628,
      "grad_norm": 9.995161056518555,
      "learning_rate": 7.914636039429701e-06,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8334923848509789,
      "num_tokens": 39568702.0,
      "step": 32960
    },
    {
      "entropy": 1.7988762125372886,
      "epoch": 0.10220411528884597,
      "grad_norm": 9.979853630065918,
      "learning_rate": 7.913435633550255e-06,
      "loss": 0.4959,
      "mean_token_accuracy": 0.83943440169096,
      "num_tokens": 39581966.0,
      "step": 32970
    },
    {
      "entropy": 1.8382903322577477,
      "epoch": 0.10223511441389567,
      "grad_norm": 8.669686317443848,
      "learning_rate": 7.912235773698689e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.838770891726017,
      "num_tokens": 39595357.0,
      "step": 32980
    },
    {
      "entropy": 1.9224609807133675,
      "epoch": 0.10226611353894537,
      "grad_norm": 8.906074523925781,
      "learning_rate": 7.911036459461177e-06,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8308119490742684,
      "num_tokens": 39606657.0,
      "step": 32990
    },
    {
      "entropy": 1.8199217766523361,
      "epoch": 0.10229711266399506,
      "grad_norm": 9.60714340209961,
      "learning_rate": 7.909837690424327e-06,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8385667935013771,
      "num_tokens": 39619596.0,
      "step": 33000
    },
    {
      "entropy": 1.9392257377505302,
      "epoch": 0.10232811178904476,
      "grad_norm": 9.89013671875,
      "learning_rate": 7.908639466175193e-06,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8208078041672706,
      "num_tokens": 39631627.0,
      "step": 33010
    },
    {
      "entropy": 1.9637094184756279,
      "epoch": 0.10235911091409446,
      "grad_norm": 9.878365516662598,
      "learning_rate": 7.907441786301261e-06,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8377602383494377,
      "num_tokens": 39642887.0,
      "step": 33020
    },
    {
      "entropy": 1.8965819612145425,
      "epoch": 0.10239011003914414,
      "grad_norm": 9.686004638671875,
      "learning_rate": 7.906244650390462e-06,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8283202305436135,
      "num_tokens": 39654706.0,
      "step": 33030
    },
    {
      "entropy": 1.8306727975606918,
      "epoch": 0.10242110916419384,
      "grad_norm": 10.983489990234375,
      "learning_rate": 7.905048058031153e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8358713150024414,
      "num_tokens": 39668226.0,
      "step": 33040
    },
    {
      "entropy": 1.898146103322506,
      "epoch": 0.10245210828924353,
      "grad_norm": 8.633929252624512,
      "learning_rate": 7.903852008812132e-06,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8244705319404602,
      "num_tokens": 39680301.0,
      "step": 33050
    },
    {
      "entropy": 1.8473190173506737,
      "epoch": 0.10248310741429323,
      "grad_norm": 4.211243629455566,
      "learning_rate": 7.902656502322633e-06,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8376186162233352,
      "num_tokens": 39693311.0,
      "step": 33060
    },
    {
      "entropy": 1.8591630086302757,
      "epoch": 0.10251410653934293,
      "grad_norm": 4.487063884735107,
      "learning_rate": 7.901461538152326e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8340800628066063,
      "num_tokens": 39706806.0,
      "step": 33070
    },
    {
      "entropy": 1.9278041929006577,
      "epoch": 0.10254510566439262,
      "grad_norm": 9.240703582763672,
      "learning_rate": 7.90026711589131e-06,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8337907418608665,
      "num_tokens": 39719457.0,
      "step": 33080
    },
    {
      "entropy": 1.9978653281927108,
      "epoch": 0.10257610478944232,
      "grad_norm": 10.427108764648438,
      "learning_rate": 7.899073235130122e-06,
      "loss": 0.6744,
      "mean_token_accuracy": 0.8095801830291748,
      "num_tokens": 39730503.0,
      "step": 33090
    },
    {
      "entropy": 1.9866398319602012,
      "epoch": 0.10260710391449201,
      "grad_norm": 8.23255729675293,
      "learning_rate": 7.897879895459734e-06,
      "loss": 0.5904,
      "mean_token_accuracy": 0.8214378640055656,
      "num_tokens": 39741771.0,
      "step": 33100
    },
    {
      "entropy": 1.8491350293159485,
      "epoch": 0.10263810303954171,
      "grad_norm": 3.994368076324463,
      "learning_rate": 7.896687096471543e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8272738143801689,
      "num_tokens": 39754475.0,
      "step": 33110
    },
    {
      "entropy": 1.9193340376019479,
      "epoch": 0.10266910216459141,
      "grad_norm": 8.316622734069824,
      "learning_rate": 7.895494837757387e-06,
      "loss": 0.5405,
      "mean_token_accuracy": 0.82731723934412,
      "num_tokens": 39766516.0,
      "step": 33120
    },
    {
      "entropy": 1.775647784769535,
      "epoch": 0.1027001012896411,
      "grad_norm": 4.390643119812012,
      "learning_rate": 7.894303118909526e-06,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8442358478903771,
      "num_tokens": 39780754.0,
      "step": 33130
    },
    {
      "entropy": 1.9259379595518111,
      "epoch": 0.1027311004146908,
      "grad_norm": 3.8750622272491455,
      "learning_rate": 7.893111939520659e-06,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8207632303237915,
      "num_tokens": 39792005.0,
      "step": 33140
    },
    {
      "entropy": 2.0047280192375183,
      "epoch": 0.10276209953974048,
      "grad_norm": 8.025463104248047,
      "learning_rate": 7.891921299183906e-06,
      "loss": 0.5898,
      "mean_token_accuracy": 0.8211274608969689,
      "num_tokens": 39803614.0,
      "step": 33150
    },
    {
      "entropy": 2.0167581260204317,
      "epoch": 0.10279309866479018,
      "grad_norm": 10.989029884338379,
      "learning_rate": 7.890731197492827e-06,
      "loss": 0.6009,
      "mean_token_accuracy": 0.8194682911038399,
      "num_tokens": 39814772.0,
      "step": 33160
    },
    {
      "entropy": 2.003499576449394,
      "epoch": 0.10282409778983988,
      "grad_norm": 8.563013076782227,
      "learning_rate": 7.889541634041405e-06,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8160374283790588,
      "num_tokens": 39826015.0,
      "step": 33170
    },
    {
      "entropy": 1.9005308762192725,
      "epoch": 0.10285509691488957,
      "grad_norm": 10.718070030212402,
      "learning_rate": 7.888352608424046e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8393619701266288,
      "num_tokens": 39838620.0,
      "step": 33180
    },
    {
      "entropy": 1.9308624148368836,
      "epoch": 0.10288609603993927,
      "grad_norm": 9.094322204589844,
      "learning_rate": 7.887164120235598e-06,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8276225015521049,
      "num_tokens": 39850572.0,
      "step": 33190
    },
    {
      "entropy": 1.9245057985186578,
      "epoch": 0.10291709516498897,
      "grad_norm": 9.67980670928955,
      "learning_rate": 7.885976169071323e-06,
      "loss": 0.5649,
      "mean_token_accuracy": 0.8204725295305252,
      "num_tokens": 39863035.0,
      "step": 33200
    },
    {
      "entropy": 1.8938257083296777,
      "epoch": 0.10294809429003866,
      "grad_norm": 11.41298770904541,
      "learning_rate": 7.884788754526915e-06,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8219778835773468,
      "num_tokens": 39875895.0,
      "step": 33210
    },
    {
      "entropy": 1.9652521327137946,
      "epoch": 0.10297909341508836,
      "grad_norm": 4.608140468597412,
      "learning_rate": 7.883601876198497e-06,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8251558512449264,
      "num_tokens": 39888247.0,
      "step": 33220
    },
    {
      "entropy": 1.9580962508916855,
      "epoch": 0.10301009254013806,
      "grad_norm": 7.537057399749756,
      "learning_rate": 7.882415533682607e-06,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8344349786639214,
      "num_tokens": 39900720.0,
      "step": 33230
    },
    {
      "entropy": 1.9044880703091622,
      "epoch": 0.10304109166518775,
      "grad_norm": 8.43292236328125,
      "learning_rate": 7.88122972657622e-06,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8305587694048882,
      "num_tokens": 39913681.0,
      "step": 33240
    },
    {
      "entropy": 1.8795413970947266,
      "epoch": 0.10307209079023745,
      "grad_norm": 3.7221243381500244,
      "learning_rate": 7.88004445447673e-06,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8384215503931045,
      "num_tokens": 39926055.0,
      "step": 33250
    },
    {
      "entropy": 1.9621031075716018,
      "epoch": 0.10310308991528715,
      "grad_norm": 10.001662254333496,
      "learning_rate": 7.878859716981954e-06,
      "loss": 0.6182,
      "mean_token_accuracy": 0.8217739418148995,
      "num_tokens": 39936665.0,
      "step": 33260
    },
    {
      "entropy": 1.935880383849144,
      "epoch": 0.10313408904033684,
      "grad_norm": 9.233072280883789,
      "learning_rate": 7.87767551369013e-06,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8238107547163963,
      "num_tokens": 39948808.0,
      "step": 33270
    },
    {
      "entropy": 1.9450037762522698,
      "epoch": 0.10316508816538653,
      "grad_norm": 4.693704128265381,
      "learning_rate": 7.876491844199926e-06,
      "loss": 0.6156,
      "mean_token_accuracy": 0.8216011270880699,
      "num_tokens": 39960695.0,
      "step": 33280
    },
    {
      "entropy": 1.910569779574871,
      "epoch": 0.10319608729043622,
      "grad_norm": 10.65746021270752,
      "learning_rate": 7.875308708110426e-06,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8182385757565498,
      "num_tokens": 39973311.0,
      "step": 33290
    },
    {
      "entropy": 1.8974032238125802,
      "epoch": 0.10322708641548592,
      "grad_norm": 8.886612892150879,
      "learning_rate": 7.874126105021134e-06,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8244770526885986,
      "num_tokens": 39985794.0,
      "step": 33300
    },
    {
      "entropy": 1.9022233217954636,
      "epoch": 0.10325808554053562,
      "grad_norm": 4.597454071044922,
      "learning_rate": 7.872944034531982e-06,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8278315708041191,
      "num_tokens": 39998753.0,
      "step": 33310
    },
    {
      "entropy": 1.8010682314634323,
      "epoch": 0.10328908466558531,
      "grad_norm": 2.244396924972534,
      "learning_rate": 7.871762496243318e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.84285439401865,
      "num_tokens": 40013133.0,
      "step": 33320
    },
    {
      "entropy": 1.868261407315731,
      "epoch": 0.10332008379063501,
      "grad_norm": 10.890486717224121,
      "learning_rate": 7.870581489755905e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8372749105095864,
      "num_tokens": 40025402.0,
      "step": 33330
    },
    {
      "entropy": 1.8567996740341186,
      "epoch": 0.1033510829156847,
      "grad_norm": 8.594077110290527,
      "learning_rate": 7.869401014670937e-06,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8362266525626183,
      "num_tokens": 40038004.0,
      "step": 33340
    },
    {
      "entropy": 1.8234371304512025,
      "epoch": 0.1033820820407344,
      "grad_norm": 8.579446792602539,
      "learning_rate": 7.868221070590013e-06,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8359899684786797,
      "num_tokens": 40051320.0,
      "step": 33350
    },
    {
      "entropy": 1.9455517828464508,
      "epoch": 0.1034130811657841,
      "grad_norm": 10.794142723083496,
      "learning_rate": 7.86704165711516e-06,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8300697863101959,
      "num_tokens": 40063005.0,
      "step": 33360
    },
    {
      "entropy": 1.902688279747963,
      "epoch": 0.1034440802908338,
      "grad_norm": 2.90916109085083,
      "learning_rate": 7.865862773848816e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8291781514883041,
      "num_tokens": 40074960.0,
      "step": 33370
    },
    {
      "entropy": 1.8599769324064255,
      "epoch": 0.10347507941588349,
      "grad_norm": 4.288544178009033,
      "learning_rate": 7.864684420393842e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8394434854388237,
      "num_tokens": 40087590.0,
      "step": 33380
    },
    {
      "entropy": 1.8361037239432334,
      "epoch": 0.10350607854093319,
      "grad_norm": 4.094844818115234,
      "learning_rate": 7.863506596353514e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8373026907444,
      "num_tokens": 40101356.0,
      "step": 33390
    },
    {
      "entropy": 1.8626354187726974,
      "epoch": 0.10353707766598287,
      "grad_norm": 9.581649780273438,
      "learning_rate": 7.862329301331517e-06,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8337352395057678,
      "num_tokens": 40113440.0,
      "step": 33400
    },
    {
      "entropy": 1.9167112335562706,
      "epoch": 0.10356807679103257,
      "grad_norm": 12.560930252075195,
      "learning_rate": 7.86115253493196e-06,
      "loss": 0.559,
      "mean_token_accuracy": 0.8210726290941238,
      "num_tokens": 40125063.0,
      "step": 33410
    },
    {
      "entropy": 1.917654138803482,
      "epoch": 0.10359907591608226,
      "grad_norm": 9.083182334899902,
      "learning_rate": 7.859976296759359e-06,
      "loss": 0.5644,
      "mean_token_accuracy": 0.8320072874426842,
      "num_tokens": 40136790.0,
      "step": 33420
    },
    {
      "entropy": 1.884972706437111,
      "epoch": 0.10363007504113196,
      "grad_norm": 9.505833625793457,
      "learning_rate": 7.858800586418653e-06,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8238067865371704,
      "num_tokens": 40148423.0,
      "step": 33430
    },
    {
      "entropy": 1.8900686636567117,
      "epoch": 0.10366107416618166,
      "grad_norm": 9.053071022033691,
      "learning_rate": 7.857625403515186e-06,
      "loss": 0.6058,
      "mean_token_accuracy": 0.8166046008467674,
      "num_tokens": 40160812.0,
      "step": 33440
    },
    {
      "entropy": 1.8167245775461196,
      "epoch": 0.10369207329123135,
      "grad_norm": 10.170108795166016,
      "learning_rate": 7.856450747654719e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8358997702598572,
      "num_tokens": 40174474.0,
      "step": 33450
    },
    {
      "entropy": 1.8785471424460412,
      "epoch": 0.10372307241628105,
      "grad_norm": 10.019652366638184,
      "learning_rate": 7.855276618443426e-06,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8401072070002555,
      "num_tokens": 40186694.0,
      "step": 33460
    },
    {
      "entropy": 1.8747892037034035,
      "epoch": 0.10375407154133075,
      "grad_norm": 9.287095069885254,
      "learning_rate": 7.854103015487889e-06,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8236067086458206,
      "num_tokens": 40199002.0,
      "step": 33470
    },
    {
      "entropy": 1.8964219346642495,
      "epoch": 0.10378507066638044,
      "grad_norm": 8.723387718200684,
      "learning_rate": 7.852929938395108e-06,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8313620582222938,
      "num_tokens": 40211446.0,
      "step": 33480
    },
    {
      "entropy": 1.8906508162617683,
      "epoch": 0.10381606979143014,
      "grad_norm": 9.373579025268555,
      "learning_rate": 7.85175738677249e-06,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8374797239899635,
      "num_tokens": 40222830.0,
      "step": 33490
    },
    {
      "entropy": 1.9234597817063332,
      "epoch": 0.10384706891647984,
      "grad_norm": 7.460415363311768,
      "learning_rate": 7.85058536022785e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8325879499316216,
      "num_tokens": 40234416.0,
      "step": 33500
    },
    {
      "entropy": 1.9918899089097977,
      "epoch": 0.10387806804152953,
      "grad_norm": 8.504613876342773,
      "learning_rate": 7.849413858369415e-06,
      "loss": 0.6361,
      "mean_token_accuracy": 0.8134088665246964,
      "num_tokens": 40245140.0,
      "step": 33510
    },
    {
      "entropy": 1.9241836652159692,
      "epoch": 0.10390906716657922,
      "grad_norm": 8.564291000366211,
      "learning_rate": 7.848242880805818e-06,
      "loss": 0.5701,
      "mean_token_accuracy": 0.8313742846250534,
      "num_tokens": 40257078.0,
      "step": 33520
    },
    {
      "entropy": 1.8901431530714035,
      "epoch": 0.10394006629162891,
      "grad_norm": 4.144465923309326,
      "learning_rate": 7.847072427146111e-06,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8276931583881378,
      "num_tokens": 40269349.0,
      "step": 33530
    },
    {
      "entropy": 1.887324059009552,
      "epoch": 0.10397106541667861,
      "grad_norm": 14.356313705444336,
      "learning_rate": 7.845902496999739e-06,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8178930029273033,
      "num_tokens": 40281901.0,
      "step": 33540
    },
    {
      "entropy": 1.9168416380882263,
      "epoch": 0.1040020645417283,
      "grad_norm": 11.27478313446045,
      "learning_rate": 7.844733089976564e-06,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8334863767027855,
      "num_tokens": 40293978.0,
      "step": 33550
    },
    {
      "entropy": 1.813339551538229,
      "epoch": 0.104033063666778,
      "grad_norm": 8.250176429748535,
      "learning_rate": 7.843564205686856e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.838591480255127,
      "num_tokens": 40306930.0,
      "step": 33560
    },
    {
      "entropy": 1.9164869621396066,
      "epoch": 0.1040640627918277,
      "grad_norm": 8.932500839233398,
      "learning_rate": 7.842395843741287e-06,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8232863306999206,
      "num_tokens": 40318760.0,
      "step": 33570
    },
    {
      "entropy": 1.9515147507190704,
      "epoch": 0.1040950619168774,
      "grad_norm": 8.837969779968262,
      "learning_rate": 7.841228003750933e-06,
      "loss": 0.6046,
      "mean_token_accuracy": 0.8223684683442116,
      "num_tokens": 40330749.0,
      "step": 33580
    },
    {
      "entropy": 1.9510893940925598,
      "epoch": 0.10412606104192709,
      "grad_norm": 10.218021392822266,
      "learning_rate": 7.84006068532728e-06,
      "loss": 0.5715,
      "mean_token_accuracy": 0.8213855206966401,
      "num_tokens": 40342128.0,
      "step": 33590
    },
    {
      "entropy": 1.9364799603819847,
      "epoch": 0.10415706016697679,
      "grad_norm": 9.228703498840332,
      "learning_rate": 7.838893888082218e-06,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8253035977482795,
      "num_tokens": 40353315.0,
      "step": 33600
    },
    {
      "entropy": 1.9496245756745338,
      "epoch": 0.10418805929202649,
      "grad_norm": 9.735555648803711,
      "learning_rate": 7.837727611628043e-06,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8161328807473183,
      "num_tokens": 40365395.0,
      "step": 33610
    },
    {
      "entropy": 1.952901628613472,
      "epoch": 0.10421905841707618,
      "grad_norm": 10.927205085754395,
      "learning_rate": 7.836561855577443e-06,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8294717028737069,
      "num_tokens": 40376818.0,
      "step": 33620
    },
    {
      "entropy": 1.8274588227272033,
      "epoch": 0.10425005754212588,
      "grad_norm": 9.007339477539062,
      "learning_rate": 7.835396619543528e-06,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8340933352708817,
      "num_tokens": 40390071.0,
      "step": 33630
    },
    {
      "entropy": 1.9603443384170531,
      "epoch": 0.10428105666717558,
      "grad_norm": 9.797590255737305,
      "learning_rate": 7.834231903139795e-06,
      "loss": 0.6205,
      "mean_token_accuracy": 0.8226620614528656,
      "num_tokens": 40401203.0,
      "step": 33640
    },
    {
      "entropy": 1.8578498139977455,
      "epoch": 0.10431205579222526,
      "grad_norm": 8.355208396911621,
      "learning_rate": 7.833067705980151e-06,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8306270629167557,
      "num_tokens": 40413367.0,
      "step": 33650
    },
    {
      "entropy": 1.8901911661028863,
      "epoch": 0.10434305491727495,
      "grad_norm": 9.124183654785156,
      "learning_rate": 7.831904027678902e-06,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8174741074442864,
      "num_tokens": 40425631.0,
      "step": 33660
    },
    {
      "entropy": 2.0042662411928176,
      "epoch": 0.10437405404232465,
      "grad_norm": 8.978107452392578,
      "learning_rate": 7.830740867850753e-06,
      "loss": 0.5935,
      "mean_token_accuracy": 0.8254029154777527,
      "num_tokens": 40435603.0,
      "step": 33670
    },
    {
      "entropy": 1.8774489134550094,
      "epoch": 0.10440505316737435,
      "grad_norm": 10.732882499694824,
      "learning_rate": 7.829578226110816e-06,
      "loss": 0.6055,
      "mean_token_accuracy": 0.8271615326404571,
      "num_tokens": 40447665.0,
      "step": 33680
    },
    {
      "entropy": 1.981630663573742,
      "epoch": 0.10443605229242404,
      "grad_norm": 11.340205192565918,
      "learning_rate": 7.828416102074594e-06,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8235415056347847,
      "num_tokens": 40458773.0,
      "step": 33690
    },
    {
      "entropy": 1.8451252445578574,
      "epoch": 0.10446705141747374,
      "grad_norm": 11.530259132385254,
      "learning_rate": 7.827254495357994e-06,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8376246988773346,
      "num_tokens": 40471667.0,
      "step": 33700
    },
    {
      "entropy": 1.950427158176899,
      "epoch": 0.10449805054252344,
      "grad_norm": 10.422981262207031,
      "learning_rate": 7.826093405577326e-06,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8280929684638977,
      "num_tokens": 40483536.0,
      "step": 33710
    },
    {
      "entropy": 1.9090193212032318,
      "epoch": 0.10452904966757313,
      "grad_norm": 8.866836547851562,
      "learning_rate": 7.824932832349292e-06,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8405329346656799,
      "num_tokens": 40494937.0,
      "step": 33720
    },
    {
      "entropy": 1.8627185627818108,
      "epoch": 0.10456004879262283,
      "grad_norm": 9.671921730041504,
      "learning_rate": 7.823772775290993e-06,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8325039520859718,
      "num_tokens": 40507661.0,
      "step": 33730
    },
    {
      "entropy": 1.8910223886370658,
      "epoch": 0.10459104791767253,
      "grad_norm": 4.5752458572387695,
      "learning_rate": 7.822613234019927e-06,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8261340275406838,
      "num_tokens": 40520066.0,
      "step": 33740
    },
    {
      "entropy": 1.9918366223573685,
      "epoch": 0.10462204704272222,
      "grad_norm": 9.358546257019043,
      "learning_rate": 7.821454208153992e-06,
      "loss": 0.6184,
      "mean_token_accuracy": 0.8228957876563072,
      "num_tokens": 40531109.0,
      "step": 33750
    },
    {
      "entropy": 1.9470916539430618,
      "epoch": 0.10465304616777192,
      "grad_norm": 10.189216613769531,
      "learning_rate": 7.82029569731148e-06,
      "loss": 0.6156,
      "mean_token_accuracy": 0.8150023117661476,
      "num_tokens": 40542537.0,
      "step": 33760
    },
    {
      "entropy": 1.9262451082468033,
      "epoch": 0.1046840452928216,
      "grad_norm": 8.381092071533203,
      "learning_rate": 7.819137701111077e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8301904648542404,
      "num_tokens": 40554442.0,
      "step": 33770
    },
    {
      "entropy": 1.910345396399498,
      "epoch": 0.1047150444178713,
      "grad_norm": 8.140871047973633,
      "learning_rate": 7.817980219171866e-06,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8225700587034226,
      "num_tokens": 40566939.0,
      "step": 33780
    },
    {
      "entropy": 1.974452766776085,
      "epoch": 0.104746043542921,
      "grad_norm": 7.459367275238037,
      "learning_rate": 7.816823251113325e-06,
      "loss": 0.5918,
      "mean_token_accuracy": 0.8310072511434555,
      "num_tokens": 40578413.0,
      "step": 33790
    },
    {
      "entropy": 1.9643989235162735,
      "epoch": 0.1047770426679707,
      "grad_norm": 9.469351768493652,
      "learning_rate": 7.815666796555324e-06,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8164726868271828,
      "num_tokens": 40589015.0,
      "step": 33800
    },
    {
      "entropy": 1.7843807175755502,
      "epoch": 0.10480804179302039,
      "grad_norm": 9.5103178024292,
      "learning_rate": 7.814510855118131e-06,
      "loss": 0.4146,
      "mean_token_accuracy": 0.8469722703099251,
      "num_tokens": 40602753.0,
      "step": 33810
    },
    {
      "entropy": 1.786408032476902,
      "epoch": 0.10483904091807009,
      "grad_norm": 8.267829895019531,
      "learning_rate": 7.8133554264224e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8310681402683258,
      "num_tokens": 40616558.0,
      "step": 33820
    },
    {
      "entropy": 1.9000122025609016,
      "epoch": 0.10487004004311978,
      "grad_norm": 8.900775909423828,
      "learning_rate": 7.812200510089185e-06,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8183383151888848,
      "num_tokens": 40629539.0,
      "step": 33830
    },
    {
      "entropy": 1.86521013379097,
      "epoch": 0.10490103916816948,
      "grad_norm": 9.411357879638672,
      "learning_rate": 7.811046105739927e-06,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8357746317982674,
      "num_tokens": 40641726.0,
      "step": 33840
    },
    {
      "entropy": 1.8550844490528107,
      "epoch": 0.10493203829321918,
      "grad_norm": 7.637185096740723,
      "learning_rate": 7.809892212996458e-06,
      "loss": 0.507,
      "mean_token_accuracy": 0.836801141500473,
      "num_tokens": 40654094.0,
      "step": 33850
    },
    {
      "entropy": 1.8382833272218704,
      "epoch": 0.10496303741826887,
      "grad_norm": 9.953614234924316,
      "learning_rate": 7.808738831481007e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8312429085373878,
      "num_tokens": 40667469.0,
      "step": 33860
    },
    {
      "entropy": 1.9534501269459725,
      "epoch": 0.10499403654331857,
      "grad_norm": 11.463193893432617,
      "learning_rate": 7.807585960816184e-06,
      "loss": 0.567,
      "mean_token_accuracy": 0.8186778038740158,
      "num_tokens": 40679414.0,
      "step": 33870
    },
    {
      "entropy": 1.8588482439517975,
      "epoch": 0.10502503566836827,
      "grad_norm": 8.321738243103027,
      "learning_rate": 7.806433600624999e-06,
      "loss": 0.501,
      "mean_token_accuracy": 0.8387341931462288,
      "num_tokens": 40692355.0,
      "step": 33880
    },
    {
      "entropy": 1.949586683511734,
      "epoch": 0.10505603479341795,
      "grad_norm": 8.959894180297852,
      "learning_rate": 7.805281750530844e-06,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8129965797066688,
      "num_tokens": 40703355.0,
      "step": 33890
    },
    {
      "entropy": 1.9277196362614633,
      "epoch": 0.10508703391846765,
      "grad_norm": 4.52024507522583,
      "learning_rate": 7.804130410157503e-06,
      "loss": 0.536,
      "mean_token_accuracy": 0.8277270719408989,
      "num_tokens": 40715053.0,
      "step": 33900
    },
    {
      "entropy": 1.901253941655159,
      "epoch": 0.10511803304351734,
      "grad_norm": 9.572426795959473,
      "learning_rate": 7.802979579129147e-06,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8186801239848137,
      "num_tokens": 40726747.0,
      "step": 33910
    },
    {
      "entropy": 1.9022594541311264,
      "epoch": 0.10514903216856704,
      "grad_norm": 9.995318412780762,
      "learning_rate": 7.801829257070337e-06,
      "loss": 0.5805,
      "mean_token_accuracy": 0.8272980287671089,
      "num_tokens": 40738081.0,
      "step": 33920
    },
    {
      "entropy": 1.8792105168104172,
      "epoch": 0.10518003129361674,
      "grad_norm": 5.292135715484619,
      "learning_rate": 7.800679443606019e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8302574872970581,
      "num_tokens": 40750419.0,
      "step": 33930
    },
    {
      "entropy": 1.8759152442216873,
      "epoch": 0.10521103041866643,
      "grad_norm": 4.318228721618652,
      "learning_rate": 7.799530138361527e-06,
      "loss": 0.455,
      "mean_token_accuracy": 0.8474466159939766,
      "num_tokens": 40762714.0,
      "step": 33940
    },
    {
      "entropy": 1.871192954480648,
      "epoch": 0.10524202954371613,
      "grad_norm": 4.8864336013793945,
      "learning_rate": 7.79838134096258e-06,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8384427651762962,
      "num_tokens": 40775536.0,
      "step": 33950
    },
    {
      "entropy": 1.8584145426750183,
      "epoch": 0.10527302866876583,
      "grad_norm": 7.488039493560791,
      "learning_rate": 7.797233051035284e-06,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8298766538500786,
      "num_tokens": 40789635.0,
      "step": 33960
    },
    {
      "entropy": 1.9506609112024307,
      "epoch": 0.10530402779381552,
      "grad_norm": 9.773181915283203,
      "learning_rate": 7.796085268206132e-06,
      "loss": 0.5754,
      "mean_token_accuracy": 0.8298773288726806,
      "num_tokens": 40800485.0,
      "step": 33970
    },
    {
      "entropy": 1.8901255756616593,
      "epoch": 0.10533502691886522,
      "grad_norm": 3.6833689212799072,
      "learning_rate": 7.794937992102e-06,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8308203518390656,
      "num_tokens": 40812631.0,
      "step": 33980
    },
    {
      "entropy": 1.8272457644343376,
      "epoch": 0.10536602604391491,
      "grad_norm": 5.967985153198242,
      "learning_rate": 7.793791222350145e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8280576094985008,
      "num_tokens": 40825791.0,
      "step": 33990
    },
    {
      "entropy": 1.9373245880007743,
      "epoch": 0.10539702516896461,
      "grad_norm": 8.980064392089844,
      "learning_rate": 7.792644958578212e-06,
      "loss": 0.585,
      "mean_token_accuracy": 0.8223650082945824,
      "num_tokens": 40837496.0,
      "step": 34000
    },
    {
      "entropy": 1.9898645401000976,
      "epoch": 0.10542802429401431,
      "grad_norm": 9.691417694091797,
      "learning_rate": 7.79149920041423e-06,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8341428533196449,
      "num_tokens": 40848939.0,
      "step": 34010
    },
    {
      "entropy": 1.9258566856384278,
      "epoch": 0.10545902341906399,
      "grad_norm": 9.314157485961914,
      "learning_rate": 7.790353947486607e-06,
      "loss": 0.5529,
      "mean_token_accuracy": 0.83617302775383,
      "num_tokens": 40861200.0,
      "step": 34020
    },
    {
      "entropy": 1.9404375448822975,
      "epoch": 0.10549002254411369,
      "grad_norm": 9.493596076965332,
      "learning_rate": 7.789209199424134e-06,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8195881083607673,
      "num_tokens": 40872706.0,
      "step": 34030
    },
    {
      "entropy": 1.8327529534697533,
      "epoch": 0.10552102166916338,
      "grad_norm": 9.407135963439941,
      "learning_rate": 7.788064955855987e-06,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8332900017499923,
      "num_tokens": 40885468.0,
      "step": 34040
    },
    {
      "entropy": 1.8383815258741378,
      "epoch": 0.10555202079421308,
      "grad_norm": 9.750924110412598,
      "learning_rate": 7.78692121641172e-06,
      "loss": 0.4872,
      "mean_token_accuracy": 0.830281549692154,
      "num_tokens": 40898646.0,
      "step": 34050
    },
    {
      "entropy": 1.9488092795014382,
      "epoch": 0.10558301991926278,
      "grad_norm": 8.798650741577148,
      "learning_rate": 7.785777980721267e-06,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8310556679964065,
      "num_tokens": 40909910.0,
      "step": 34060
    },
    {
      "entropy": 1.8724893182516098,
      "epoch": 0.10561401904431247,
      "grad_norm": 8.707452774047852,
      "learning_rate": 7.784635248414945e-06,
      "loss": 0.5413,
      "mean_token_accuracy": 0.825086310505867,
      "num_tokens": 40921391.0,
      "step": 34070
    },
    {
      "entropy": 1.8841823562979698,
      "epoch": 0.10564501816936217,
      "grad_norm": 9.368388175964355,
      "learning_rate": 7.783493019123451e-06,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8284827679395675,
      "num_tokens": 40933106.0,
      "step": 34080
    },
    {
      "entropy": 1.893832103908062,
      "epoch": 0.10567601729441187,
      "grad_norm": 9.662880897521973,
      "learning_rate": 7.78235129247786e-06,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8264197260141373,
      "num_tokens": 40945947.0,
      "step": 34090
    },
    {
      "entropy": 1.959150141477585,
      "epoch": 0.10570701641946156,
      "grad_norm": 11.217707633972168,
      "learning_rate": 7.781210068109623e-06,
      "loss": 0.6322,
      "mean_token_accuracy": 0.8149213716387749,
      "num_tokens": 40957886.0,
      "step": 34100
    },
    {
      "entropy": 1.8911303892731666,
      "epoch": 0.10573801554451126,
      "grad_norm": 9.337475776672363,
      "learning_rate": 7.780069345650573e-06,
      "loss": 0.5306,
      "mean_token_accuracy": 0.8183432757854462,
      "num_tokens": 40969448.0,
      "step": 34110
    },
    {
      "entropy": 1.8776145607233048,
      "epoch": 0.10576901466956096,
      "grad_norm": 9.5222749710083,
      "learning_rate": 7.778929124732918e-06,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8276842266321183,
      "num_tokens": 40981711.0,
      "step": 34120
    },
    {
      "entropy": 1.94348586499691,
      "epoch": 0.10580001379461065,
      "grad_norm": 8.874763488769531,
      "learning_rate": 7.777789404989248e-06,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8309635683894158,
      "num_tokens": 40992978.0,
      "step": 34130
    },
    {
      "entropy": 1.8115389600396157,
      "epoch": 0.10583101291966034,
      "grad_norm": 8.382539749145508,
      "learning_rate": 7.776650186052521e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8349054649472236,
      "num_tokens": 41005680.0,
      "step": 34140
    },
    {
      "entropy": 1.8448505356907845,
      "epoch": 0.10586201204471003,
      "grad_norm": 4.7137451171875,
      "learning_rate": 7.77551146755608e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8491442829370499,
      "num_tokens": 41017977.0,
      "step": 34150
    },
    {
      "entropy": 1.8678378522396089,
      "epoch": 0.10589301116975973,
      "grad_norm": 8.694931983947754,
      "learning_rate": 7.774373249133641e-06,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8227035477757454,
      "num_tokens": 41029814.0,
      "step": 34160
    },
    {
      "entropy": 1.9137464344501496,
      "epoch": 0.10592401029480943,
      "grad_norm": 8.952034950256348,
      "learning_rate": 7.773235530419292e-06,
      "loss": 0.558,
      "mean_token_accuracy": 0.8287974938750267,
      "num_tokens": 41041332.0,
      "step": 34170
    },
    {
      "entropy": 1.872650384902954,
      "epoch": 0.10595500941985912,
      "grad_norm": 9.672776222229004,
      "learning_rate": 7.7720983110475e-06,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8209738954901695,
      "num_tokens": 41054678.0,
      "step": 34180
    },
    {
      "entropy": 1.8565445743501185,
      "epoch": 0.10598600854490882,
      "grad_norm": 2.5345826148986816,
      "learning_rate": 7.770961590653102e-06,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8245491355657577,
      "num_tokens": 41067439.0,
      "step": 34190
    },
    {
      "entropy": 1.997004970908165,
      "epoch": 0.10601700766995852,
      "grad_norm": 10.320369720458984,
      "learning_rate": 7.769825368871312e-06,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8295348644256592,
      "num_tokens": 41077968.0,
      "step": 34200
    },
    {
      "entropy": 1.916537807881832,
      "epoch": 0.10604800679500821,
      "grad_norm": 9.567767143249512,
      "learning_rate": 7.76868964533772e-06,
      "loss": 0.559,
      "mean_token_accuracy": 0.8223750725388527,
      "num_tokens": 41089515.0,
      "step": 34210
    },
    {
      "entropy": 1.8700389847159387,
      "epoch": 0.10607900592005791,
      "grad_norm": 8.427227020263672,
      "learning_rate": 7.767554419688279e-06,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8274742797017097,
      "num_tokens": 41103388.0,
      "step": 34220
    },
    {
      "entropy": 1.920013178884983,
      "epoch": 0.1061100050451076,
      "grad_norm": 9.945453643798828,
      "learning_rate": 7.766419691559324e-06,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8253174960613251,
      "num_tokens": 41116163.0,
      "step": 34230
    },
    {
      "entropy": 1.8836813405156136,
      "epoch": 0.1061410041701573,
      "grad_norm": 7.964671611785889,
      "learning_rate": 7.765285460587557e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8437875971198082,
      "num_tokens": 41128197.0,
      "step": 34240
    },
    {
      "entropy": 1.8750670045614242,
      "epoch": 0.106172003295207,
      "grad_norm": 8.137727737426758,
      "learning_rate": 7.764151726410055e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8270835474133491,
      "num_tokens": 41140601.0,
      "step": 34250
    },
    {
      "entropy": 1.9073902159929275,
      "epoch": 0.10620300242025668,
      "grad_norm": 8.645315170288086,
      "learning_rate": 7.76301848866426e-06,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8317181885242462,
      "num_tokens": 41152349.0,
      "step": 34260
    },
    {
      "entropy": 1.9043108850717545,
      "epoch": 0.10623400154530638,
      "grad_norm": 10.44301700592041,
      "learning_rate": 7.761885746987988e-06,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8260853886604309,
      "num_tokens": 41164311.0,
      "step": 34270
    },
    {
      "entropy": 1.791078907251358,
      "epoch": 0.10626500067035607,
      "grad_norm": 8.425952911376953,
      "learning_rate": 7.760753501019428e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8489831522107124,
      "num_tokens": 41178238.0,
      "step": 34280
    },
    {
      "entropy": 1.8554448202252387,
      "epoch": 0.10629599979540577,
      "grad_norm": 8.66896915435791,
      "learning_rate": 7.759621750397129e-06,
      "loss": 0.5169,
      "mean_token_accuracy": 0.8302797332406044,
      "num_tokens": 41191021.0,
      "step": 34290
    },
    {
      "entropy": 1.9642982304096221,
      "epoch": 0.10632699892045547,
      "grad_norm": 9.48714828491211,
      "learning_rate": 7.758490494760018e-06,
      "loss": 0.5871,
      "mean_token_accuracy": 0.8185561686754227,
      "num_tokens": 41202431.0,
      "step": 34300
    },
    {
      "entropy": 1.8738651275634766,
      "epoch": 0.10635799804550516,
      "grad_norm": 4.17008113861084,
      "learning_rate": 7.757359733747389e-06,
      "loss": 0.5411,
      "mean_token_accuracy": 0.830881142616272,
      "num_tokens": 41215618.0,
      "step": 34310
    },
    {
      "entropy": 1.993169930577278,
      "epoch": 0.10638899717055486,
      "grad_norm": 8.395801544189453,
      "learning_rate": 7.756229466998896e-06,
      "loss": 0.6001,
      "mean_token_accuracy": 0.8117834225296974,
      "num_tokens": 41226702.0,
      "step": 34320
    },
    {
      "entropy": 1.8517363399267197,
      "epoch": 0.10641999629560456,
      "grad_norm": 8.656838417053223,
      "learning_rate": 7.755099694154571e-06,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8280891165137291,
      "num_tokens": 41239449.0,
      "step": 34330
    },
    {
      "entropy": 1.9847760811448096,
      "epoch": 0.10645099542065425,
      "grad_norm": 8.594053268432617,
      "learning_rate": 7.753970414854808e-06,
      "loss": 0.6134,
      "mean_token_accuracy": 0.806470163166523,
      "num_tokens": 41250602.0,
      "step": 34340
    },
    {
      "entropy": 1.9564678460359572,
      "epoch": 0.10648199454570395,
      "grad_norm": 12.032402992248535,
      "learning_rate": 7.752841628740366e-06,
      "loss": 0.6145,
      "mean_token_accuracy": 0.8159340843558311,
      "num_tokens": 41262285.0,
      "step": 34350
    },
    {
      "entropy": 1.9261070042848587,
      "epoch": 0.10651299367075365,
      "grad_norm": 4.049191951751709,
      "learning_rate": 7.751713335452372e-06,
      "loss": 0.5818,
      "mean_token_accuracy": 0.8276063561439514,
      "num_tokens": 41273764.0,
      "step": 34360
    },
    {
      "entropy": 1.879560787975788,
      "epoch": 0.10654399279580334,
      "grad_norm": 8.463150024414062,
      "learning_rate": 7.750585534632318e-06,
      "loss": 0.525,
      "mean_token_accuracy": 0.8304709896445275,
      "num_tokens": 41285963.0,
      "step": 34370
    },
    {
      "entropy": 1.8946872353553772,
      "epoch": 0.10657499192085304,
      "grad_norm": 3.9842400550842285,
      "learning_rate": 7.74945822592206e-06,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8228972956538201,
      "num_tokens": 41297942.0,
      "step": 34380
    },
    {
      "entropy": 1.9021208494901658,
      "epoch": 0.10660599104590272,
      "grad_norm": 8.44839096069336,
      "learning_rate": 7.748331408963822e-06,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8181784346699714,
      "num_tokens": 41309011.0,
      "step": 34390
    },
    {
      "entropy": 1.9210452124476434,
      "epoch": 0.10663699017095242,
      "grad_norm": 7.22066593170166,
      "learning_rate": 7.747205083400192e-06,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8235123857855797,
      "num_tokens": 41320533.0,
      "step": 34400
    },
    {
      "entropy": 1.8867201492190362,
      "epoch": 0.10666798929600212,
      "grad_norm": 5.259980201721191,
      "learning_rate": 7.746079248874114e-06,
      "loss": 0.577,
      "mean_token_accuracy": 0.8186955004930496,
      "num_tokens": 41333184.0,
      "step": 34410
    },
    {
      "entropy": 1.8663445100188256,
      "epoch": 0.10669898842105181,
      "grad_norm": 10.16250228881836,
      "learning_rate": 7.7449539050289e-06,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8328115671873093,
      "num_tokens": 41344781.0,
      "step": 34420
    },
    {
      "entropy": 1.8742147445678712,
      "epoch": 0.10672998754610151,
      "grad_norm": 8.090653419494629,
      "learning_rate": 7.743829051508229e-06,
      "loss": 0.5541,
      "mean_token_accuracy": 0.8289492219686508,
      "num_tokens": 41356329.0,
      "step": 34430
    },
    {
      "entropy": 1.8659478917717933,
      "epoch": 0.1067609866711512,
      "grad_norm": 11.673904418945312,
      "learning_rate": 7.742704687956137e-06,
      "loss": 0.5138,
      "mean_token_accuracy": 0.8437393367290497,
      "num_tokens": 41368494.0,
      "step": 34440
    },
    {
      "entropy": 1.884896233677864,
      "epoch": 0.1067919857962009,
      "grad_norm": 8.397565841674805,
      "learning_rate": 7.741580814017023e-06,
      "loss": 0.5282,
      "mean_token_accuracy": 0.830049929022789,
      "num_tokens": 41380843.0,
      "step": 34450
    },
    {
      "entropy": 1.8628281027078628,
      "epoch": 0.1068229849212506,
      "grad_norm": 9.932318687438965,
      "learning_rate": 7.740457429335646e-06,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8300712257623672,
      "num_tokens": 41393082.0,
      "step": 34460
    },
    {
      "entropy": 1.8951723709702493,
      "epoch": 0.1068539840463003,
      "grad_norm": 9.382803916931152,
      "learning_rate": 7.739334533557126e-06,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8235040470957756,
      "num_tokens": 41404878.0,
      "step": 34470
    },
    {
      "entropy": 1.8519971266388893,
      "epoch": 0.10688498317134999,
      "grad_norm": 9.738779067993164,
      "learning_rate": 7.738212126326949e-06,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8380891785025597,
      "num_tokens": 41416588.0,
      "step": 34480
    },
    {
      "entropy": 1.9151021018624306,
      "epoch": 0.10691598229639969,
      "grad_norm": 9.413227081298828,
      "learning_rate": 7.73709020729095e-06,
      "loss": 0.655,
      "mean_token_accuracy": 0.8125704079866409,
      "num_tokens": 41428429.0,
      "step": 34490
    },
    {
      "entropy": 1.9181360185146332,
      "epoch": 0.10694698142144939,
      "grad_norm": 8.843558311462402,
      "learning_rate": 7.735968776095331e-06,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8319054901599884,
      "num_tokens": 41440103.0,
      "step": 34500
    },
    {
      "entropy": 1.8450947090983392,
      "epoch": 0.10697798054649907,
      "grad_norm": 9.410344123840332,
      "learning_rate": 7.734847832386653e-06,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8249197214841842,
      "num_tokens": 41452461.0,
      "step": 34510
    },
    {
      "entropy": 1.8387939289212227,
      "epoch": 0.10700897967154877,
      "grad_norm": 3.964416980743408,
      "learning_rate": 7.73372737581183e-06,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8225889384746552,
      "num_tokens": 41464924.0,
      "step": 34520
    },
    {
      "entropy": 1.929051786661148,
      "epoch": 0.10703997879659846,
      "grad_norm": 12.12761402130127,
      "learning_rate": 7.73260740601814e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8324173107743263,
      "num_tokens": 41475753.0,
      "step": 34530
    },
    {
      "entropy": 1.9498266860842706,
      "epoch": 0.10707097792164816,
      "grad_norm": 3.5817205905914307,
      "learning_rate": 7.731487922653216e-06,
      "loss": 0.6098,
      "mean_token_accuracy": 0.8192368969321251,
      "num_tokens": 41486719.0,
      "step": 34540
    },
    {
      "entropy": 1.7467003166675568,
      "epoch": 0.10710197704669785,
      "grad_norm": 4.695430278778076,
      "learning_rate": 7.730368925365049e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8428031265735626,
      "num_tokens": 41500593.0,
      "step": 34550
    },
    {
      "entropy": 1.9088023439049722,
      "epoch": 0.10713297617174755,
      "grad_norm": 9.617088317871094,
      "learning_rate": 7.72925041380198e-06,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8277084872126579,
      "num_tokens": 41512514.0,
      "step": 34560
    },
    {
      "entropy": 1.9499268174171447,
      "epoch": 0.10716397529679725,
      "grad_norm": 9.283459663391113,
      "learning_rate": 7.728132387612718e-06,
      "loss": 0.6333,
      "mean_token_accuracy": 0.8034616574645043,
      "num_tokens": 41523613.0,
      "step": 34570
    },
    {
      "entropy": 1.857503816485405,
      "epoch": 0.10719497442184694,
      "grad_norm": 9.62498950958252,
      "learning_rate": 7.727014846446315e-06,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8330330818891525,
      "num_tokens": 41536222.0,
      "step": 34580
    },
    {
      "entropy": 1.927737507224083,
      "epoch": 0.10722597354689664,
      "grad_norm": 7.827761173248291,
      "learning_rate": 7.72589778995219e-06,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8250738069415092,
      "num_tokens": 41547116.0,
      "step": 34590
    },
    {
      "entropy": 1.894035741686821,
      "epoch": 0.10725697267194634,
      "grad_norm": 7.857428073883057,
      "learning_rate": 7.724781217780106e-06,
      "loss": 0.546,
      "mean_token_accuracy": 0.8303033024072647,
      "num_tokens": 41558322.0,
      "step": 34600
    },
    {
      "entropy": 1.8952913254499435,
      "epoch": 0.10728797179699603,
      "grad_norm": 7.348844051361084,
      "learning_rate": 7.723665129580187e-06,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8175556018948555,
      "num_tokens": 41569188.0,
      "step": 34610
    },
    {
      "entropy": 1.8319330915808678,
      "epoch": 0.10731897092204573,
      "grad_norm": 4.2293901443481445,
      "learning_rate": 7.72254952500291e-06,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8246565207839012,
      "num_tokens": 41582070.0,
      "step": 34620
    },
    {
      "entropy": 1.9387992650270462,
      "epoch": 0.10734997004709541,
      "grad_norm": 9.970220565795898,
      "learning_rate": 7.721434403699101e-06,
      "loss": 0.6521,
      "mean_token_accuracy": 0.8120075181126595,
      "num_tokens": 41593014.0,
      "step": 34630
    },
    {
      "entropy": 1.781299701333046,
      "epoch": 0.10738096917214511,
      "grad_norm": 4.431691646575928,
      "learning_rate": 7.720319765319946e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8438036203384399,
      "num_tokens": 41606201.0,
      "step": 34640
    },
    {
      "entropy": 1.902137640118599,
      "epoch": 0.10741196829719481,
      "grad_norm": 8.883857727050781,
      "learning_rate": 7.719205609516975e-06,
      "loss": 0.6318,
      "mean_token_accuracy": 0.8142830148339272,
      "num_tokens": 41616952.0,
      "step": 34650
    },
    {
      "entropy": 1.7906480133533478,
      "epoch": 0.1074429674222445,
      "grad_norm": 8.499350547790527,
      "learning_rate": 7.718091935942078e-06,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8440255209803581,
      "num_tokens": 41629686.0,
      "step": 34660
    },
    {
      "entropy": 1.8703452154994011,
      "epoch": 0.1074739665472942,
      "grad_norm": 9.963065147399902,
      "learning_rate": 7.71697874424749e-06,
      "loss": 0.5693,
      "mean_token_accuracy": 0.8165108144283295,
      "num_tokens": 41641553.0,
      "step": 34670
    },
    {
      "entropy": 1.842675694823265,
      "epoch": 0.1075049656723439,
      "grad_norm": 7.852229118347168,
      "learning_rate": 7.7158660340858e-06,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8368835672736168,
      "num_tokens": 41653222.0,
      "step": 34680
    },
    {
      "entropy": 1.8542825773358345,
      "epoch": 0.1075359647973936,
      "grad_norm": 8.768325805664062,
      "learning_rate": 7.71475380510995e-06,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8340888366103172,
      "num_tokens": 41665255.0,
      "step": 34690
    },
    {
      "entropy": 1.796558152139187,
      "epoch": 0.10756696392244329,
      "grad_norm": 9.535799980163574,
      "learning_rate": 7.713642056973227e-06,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8312591359019279,
      "num_tokens": 41677579.0,
      "step": 34700
    },
    {
      "entropy": 1.8598329350352287,
      "epoch": 0.10759796304749299,
      "grad_norm": 9.506328582763672,
      "learning_rate": 7.71253078932927e-06,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8224905788898468,
      "num_tokens": 41688761.0,
      "step": 34710
    },
    {
      "entropy": 1.8570543482899666,
      "epoch": 0.10762896217254268,
      "grad_norm": 8.487010955810547,
      "learning_rate": 7.711420001832066e-06,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8308738321065903,
      "num_tokens": 41700400.0,
      "step": 34720
    },
    {
      "entropy": 1.8241771847009658,
      "epoch": 0.10765996129759238,
      "grad_norm": 8.443219184875488,
      "learning_rate": 7.710309694135956e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8372260481119156,
      "num_tokens": 41712818.0,
      "step": 34730
    },
    {
      "entropy": 1.8751182019710542,
      "epoch": 0.10769096042264208,
      "grad_norm": 10.146086692810059,
      "learning_rate": 7.709199865895622e-06,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8366193175315857,
      "num_tokens": 41724064.0,
      "step": 34740
    },
    {
      "entropy": 1.8233239591121673,
      "epoch": 0.10772195954769177,
      "grad_norm": 10.28363037109375,
      "learning_rate": 7.708090516766096e-06,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8335344001650811,
      "num_tokens": 41736368.0,
      "step": 34750
    },
    {
      "entropy": 1.8685315743088722,
      "epoch": 0.10775295867274146,
      "grad_norm": 4.088825702667236,
      "learning_rate": 7.706981646402762e-06,
      "loss": 0.5963,
      "mean_token_accuracy": 0.8252598002552987,
      "num_tokens": 41747379.0,
      "step": 34760
    },
    {
      "entropy": 1.8391776710748673,
      "epoch": 0.10778395779779115,
      "grad_norm": 9.287007331848145,
      "learning_rate": 7.705873254461345e-06,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8283318698406219,
      "num_tokens": 41759137.0,
      "step": 34770
    },
    {
      "entropy": 1.825559838116169,
      "epoch": 0.10781495692284085,
      "grad_norm": 8.258102416992188,
      "learning_rate": 7.704765340597917e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8404336258769035,
      "num_tokens": 41771923.0,
      "step": 34780
    },
    {
      "entropy": 1.876462672650814,
      "epoch": 0.10784595604789055,
      "grad_norm": 9.638961791992188,
      "learning_rate": 7.703657904468902e-06,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8340917885303497,
      "num_tokens": 41784775.0,
      "step": 34790
    },
    {
      "entropy": 1.7784705072641374,
      "epoch": 0.10787695517294024,
      "grad_norm": 8.940130233764648,
      "learning_rate": 7.702550945731066e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.831823606789112,
      "num_tokens": 41797943.0,
      "step": 34800
    },
    {
      "entropy": 1.882892769575119,
      "epoch": 0.10790795429798994,
      "grad_norm": 10.979787826538086,
      "learning_rate": 7.701444464041514e-06,
      "loss": 0.5337,
      "mean_token_accuracy": 0.827871498465538,
      "num_tokens": 41810184.0,
      "step": 34810
    },
    {
      "entropy": 1.912169161438942,
      "epoch": 0.10793895342303964,
      "grad_norm": 8.458621978759766,
      "learning_rate": 7.700338459057705e-06,
      "loss": 0.5786,
      "mean_token_accuracy": 0.8240523219108582,
      "num_tokens": 41821501.0,
      "step": 34820
    },
    {
      "entropy": 1.8121132165193559,
      "epoch": 0.10796995254808933,
      "grad_norm": 4.544675350189209,
      "learning_rate": 7.699232930437439e-06,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8329327836632728,
      "num_tokens": 41834757.0,
      "step": 34830
    },
    {
      "entropy": 1.811763161420822,
      "epoch": 0.10800095167313903,
      "grad_norm": 5.41200590133667,
      "learning_rate": 7.698127877838858e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8376516059041024,
      "num_tokens": 41847967.0,
      "step": 34840
    },
    {
      "entropy": 1.9670744597911836,
      "epoch": 0.10803195079818873,
      "grad_norm": 8.704374313354492,
      "learning_rate": 7.69702330092045e-06,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8268567532300949,
      "num_tokens": 41859345.0,
      "step": 34850
    },
    {
      "entropy": 1.9194496154785157,
      "epoch": 0.10806294992323842,
      "grad_norm": 4.117202281951904,
      "learning_rate": 7.695919199341043e-06,
      "loss": 0.5799,
      "mean_token_accuracy": 0.8342431426048279,
      "num_tokens": 41870624.0,
      "step": 34860
    },
    {
      "entropy": 1.8487198695540428,
      "epoch": 0.10809394904828812,
      "grad_norm": 4.074057102203369,
      "learning_rate": 7.694815572759812e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8297583907842636,
      "num_tokens": 41882966.0,
      "step": 34870
    },
    {
      "entropy": 1.8381000190973282,
      "epoch": 0.1081249481733378,
      "grad_norm": 7.816097736358643,
      "learning_rate": 7.693712420836265e-06,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8306377053260803,
      "num_tokens": 41895795.0,
      "step": 34880
    },
    {
      "entropy": 1.9027506560087204,
      "epoch": 0.1081559472983875,
      "grad_norm": 8.619202613830566,
      "learning_rate": 7.692609743230265e-06,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8351558074355125,
      "num_tokens": 41907196.0,
      "step": 34890
    },
    {
      "entropy": 1.937445905804634,
      "epoch": 0.1081869464234372,
      "grad_norm": 8.807971000671387,
      "learning_rate": 7.691507539602005e-06,
      "loss": 0.6096,
      "mean_token_accuracy": 0.8202361524105072,
      "num_tokens": 41918599.0,
      "step": 34900
    },
    {
      "entropy": 1.8574762254953385,
      "epoch": 0.10821794554848689,
      "grad_norm": 8.557307243347168,
      "learning_rate": 7.690405809612025e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8375605642795563,
      "num_tokens": 41930689.0,
      "step": 34910
    },
    {
      "entropy": 1.9088428899645806,
      "epoch": 0.10824894467353659,
      "grad_norm": 8.38049602508545,
      "learning_rate": 7.689304552921199e-06,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8217715606093406,
      "num_tokens": 41941970.0,
      "step": 34920
    },
    {
      "entropy": 1.8612014800310135,
      "epoch": 0.10827994379858628,
      "grad_norm": 10.078642845153809,
      "learning_rate": 7.688203769190748e-06,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8198879033327102,
      "num_tokens": 41954804.0,
      "step": 34930
    },
    {
      "entropy": 1.8768370226025581,
      "epoch": 0.10831094292363598,
      "grad_norm": 10.8411865234375,
      "learning_rate": 7.687103458082228e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8370490476489068,
      "num_tokens": 41965644.0,
      "step": 34940
    },
    {
      "entropy": 1.8967427670955659,
      "epoch": 0.10834194204868568,
      "grad_norm": 9.28530502319336,
      "learning_rate": 7.686003619257535e-06,
      "loss": 0.6235,
      "mean_token_accuracy": 0.8072300642728806,
      "num_tokens": 41977403.0,
      "step": 34950
    },
    {
      "entropy": 1.8946552872657776,
      "epoch": 0.10837294117373537,
      "grad_norm": 4.223282337188721,
      "learning_rate": 7.684904252378904e-06,
      "loss": 0.6063,
      "mean_token_accuracy": 0.8308607801795006,
      "num_tokens": 41989197.0,
      "step": 34960
    },
    {
      "entropy": 1.9117924958467483,
      "epoch": 0.10840394029878507,
      "grad_norm": 12.681775093078613,
      "learning_rate": 7.683805357108907e-06,
      "loss": 0.5557,
      "mean_token_accuracy": 0.823358316719532,
      "num_tokens": 42000840.0,
      "step": 34970
    },
    {
      "entropy": 1.8810236111283303,
      "epoch": 0.10843493942383477,
      "grad_norm": 8.788151741027832,
      "learning_rate": 7.682706933110456e-06,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8217835262417793,
      "num_tokens": 42013194.0,
      "step": 34980
    },
    {
      "entropy": 1.8799246475100517,
      "epoch": 0.10846593854888446,
      "grad_norm": 8.282735824584961,
      "learning_rate": 7.681608980046798e-06,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8285661846399307,
      "num_tokens": 42025091.0,
      "step": 34990
    },
    {
      "entropy": 1.8529492557048797,
      "epoch": 0.10849693767393416,
      "grad_norm": 8.723884582519531,
      "learning_rate": 7.680511497581516e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8336036711931228,
      "num_tokens": 42038406.0,
      "step": 35000
    },
    {
      "entropy": 1.8494236335158347,
      "epoch": 0.10852793679898384,
      "grad_norm": 4.324466705322266,
      "learning_rate": 7.67941448537853e-06,
      "loss": 0.5534,
      "mean_token_accuracy": 0.8200669184327125,
      "num_tokens": 42051116.0,
      "step": 35010
    },
    {
      "entropy": 1.9139706775546075,
      "epoch": 0.10855893592403354,
      "grad_norm": 8.67935562133789,
      "learning_rate": 7.6783179431021e-06,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8277176558971405,
      "num_tokens": 42063002.0,
      "step": 35020
    },
    {
      "entropy": 1.8269087105989457,
      "epoch": 0.10858993504908324,
      "grad_norm": 9.06470775604248,
      "learning_rate": 7.677221870416817e-06,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8408965677022934,
      "num_tokens": 42075421.0,
      "step": 35030
    },
    {
      "entropy": 1.862933087348938,
      "epoch": 0.10862093417413293,
      "grad_norm": 9.062261581420898,
      "learning_rate": 7.676126266987606e-06,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8264860972762108,
      "num_tokens": 42087233.0,
      "step": 35040
    },
    {
      "entropy": 1.874533024430275,
      "epoch": 0.10865193329918263,
      "grad_norm": 8.091785430908203,
      "learning_rate": 7.67503113247973e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8257019862532615,
      "num_tokens": 42099681.0,
      "step": 35050
    },
    {
      "entropy": 1.956665936112404,
      "epoch": 0.10868293242423233,
      "grad_norm": 8.70102310180664,
      "learning_rate": 7.673936466558786e-06,
      "loss": 0.6122,
      "mean_token_accuracy": 0.8198647305369378,
      "num_tokens": 42110874.0,
      "step": 35060
    },
    {
      "entropy": 1.912957863509655,
      "epoch": 0.10871393154928202,
      "grad_norm": 9.428136825561523,
      "learning_rate": 7.672842268890703e-06,
      "loss": 0.6027,
      "mean_token_accuracy": 0.8237558603286743,
      "num_tokens": 42122704.0,
      "step": 35070
    },
    {
      "entropy": 1.9081886559724808,
      "epoch": 0.10874493067433172,
      "grad_norm": 9.511113166809082,
      "learning_rate": 7.671748539141744e-06,
      "loss": 0.6123,
      "mean_token_accuracy": 0.8153191924095153,
      "num_tokens": 42135165.0,
      "step": 35080
    },
    {
      "entropy": 1.8714108362793922,
      "epoch": 0.10877592979938142,
      "grad_norm": 9.29901123046875,
      "learning_rate": 7.670655276978506e-06,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8311486825346946,
      "num_tokens": 42147431.0,
      "step": 35090
    },
    {
      "entropy": 1.8901761874556542,
      "epoch": 0.10880692892443111,
      "grad_norm": 8.597390174865723,
      "learning_rate": 7.669562482067915e-06,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8393839746713638,
      "num_tokens": 42159935.0,
      "step": 35100
    },
    {
      "entropy": 1.8835955768823625,
      "epoch": 0.10883792804948081,
      "grad_norm": 8.558197021484375,
      "learning_rate": 7.668470154077237e-06,
      "loss": 0.594,
      "mean_token_accuracy": 0.8140904232859612,
      "num_tokens": 42172790.0,
      "step": 35110
    },
    {
      "entropy": 1.9337810754776001,
      "epoch": 0.1088689271745305,
      "grad_norm": 8.871868133544922,
      "learning_rate": 7.667378292674056e-06,
      "loss": 0.5931,
      "mean_token_accuracy": 0.8145381569862366,
      "num_tokens": 42184215.0,
      "step": 35120
    },
    {
      "entropy": 1.8214157864451408,
      "epoch": 0.10889992629958019,
      "grad_norm": 4.693521499633789,
      "learning_rate": 7.666286897526304e-06,
      "loss": 0.4763,
      "mean_token_accuracy": 0.8404714584350585,
      "num_tokens": 42197411.0,
      "step": 35130
    },
    {
      "entropy": 1.8844437003135681,
      "epoch": 0.10893092542462988,
      "grad_norm": 9.958372116088867,
      "learning_rate": 7.66519596830223e-06,
      "loss": 0.5385,
      "mean_token_accuracy": 0.822202742099762,
      "num_tokens": 42209406.0,
      "step": 35140
    },
    {
      "entropy": 1.9353954821825028,
      "epoch": 0.10896192454967958,
      "grad_norm": 9.83383560180664,
      "learning_rate": 7.664105504670418e-06,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8054858520627022,
      "num_tokens": 42221328.0,
      "step": 35150
    },
    {
      "entropy": 1.7801556140184402,
      "epoch": 0.10899292367472928,
      "grad_norm": 5.5207061767578125,
      "learning_rate": 7.663015506299786e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.844315542280674,
      "num_tokens": 42234769.0,
      "step": 35160
    },
    {
      "entropy": 1.862704548239708,
      "epoch": 0.10902392279977897,
      "grad_norm": 7.391566276550293,
      "learning_rate": 7.661925972859575e-06,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8368149012327194,
      "num_tokens": 42247060.0,
      "step": 35170
    },
    {
      "entropy": 1.9155489355325699,
      "epoch": 0.10905492192482867,
      "grad_norm": 4.612364768981934,
      "learning_rate": 7.66083690401936e-06,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8262100204825401,
      "num_tokens": 42258503.0,
      "step": 35180
    },
    {
      "entropy": 1.9643469721078872,
      "epoch": 0.10908592104987837,
      "grad_norm": 9.24181079864502,
      "learning_rate": 7.659748299449044e-06,
      "loss": 0.6368,
      "mean_token_accuracy": 0.8126005738973617,
      "num_tokens": 42269513.0,
      "step": 35190
    },
    {
      "entropy": 1.8432982012629509,
      "epoch": 0.10911692017492806,
      "grad_norm": 9.919602394104004,
      "learning_rate": 7.658660158818853e-06,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8412719219923019,
      "num_tokens": 42281589.0,
      "step": 35200
    },
    {
      "entropy": 1.8541820272803307,
      "epoch": 0.10914791929997776,
      "grad_norm": 4.436375617980957,
      "learning_rate": 7.657572481799348e-06,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8266458585858345,
      "num_tokens": 42293563.0,
      "step": 35210
    },
    {
      "entropy": 1.8942555904388427,
      "epoch": 0.10917891842502746,
      "grad_norm": 10.810229301452637,
      "learning_rate": 7.656485268061414e-06,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8252988263964653,
      "num_tokens": 42305355.0,
      "step": 35220
    },
    {
      "entropy": 1.8540264323353768,
      "epoch": 0.10920991755007715,
      "grad_norm": 8.089149475097656,
      "learning_rate": 7.655398517276262e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8406173124909401,
      "num_tokens": 42317042.0,
      "step": 35230
    },
    {
      "entropy": 1.831780730187893,
      "epoch": 0.10924091667512685,
      "grad_norm": 10.136151313781738,
      "learning_rate": 7.654312229115433e-06,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8417727887630463,
      "num_tokens": 42329730.0,
      "step": 35240
    },
    {
      "entropy": 1.8968460232019424,
      "epoch": 0.10927191580017653,
      "grad_norm": 10.715241432189941,
      "learning_rate": 7.65322640325079e-06,
      "loss": 0.6043,
      "mean_token_accuracy": 0.8156336963176727,
      "num_tokens": 42341271.0,
      "step": 35250
    },
    {
      "entropy": 1.8117874696850778,
      "epoch": 0.10930291492522623,
      "grad_norm": 8.665305137634277,
      "learning_rate": 7.652141039354524e-06,
      "loss": 0.4908,
      "mean_token_accuracy": 0.8276491865515709,
      "num_tokens": 42354850.0,
      "step": 35260
    },
    {
      "entropy": 1.85145965218544,
      "epoch": 0.10933391405027593,
      "grad_norm": 10.74636173248291,
      "learning_rate": 7.651056137099154e-06,
      "loss": 0.512,
      "mean_token_accuracy": 0.8369231000542641,
      "num_tokens": 42367039.0,
      "step": 35270
    },
    {
      "entropy": 1.818227480351925,
      "epoch": 0.10936491317532562,
      "grad_norm": 10.717218399047852,
      "learning_rate": 7.649971696157518e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8339720323681832,
      "num_tokens": 42380069.0,
      "step": 35280
    },
    {
      "entropy": 1.8332565039396287,
      "epoch": 0.10939591230037532,
      "grad_norm": 8.914962768554688,
      "learning_rate": 7.648887716202781e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8378045499324799,
      "num_tokens": 42392532.0,
      "step": 35290
    },
    {
      "entropy": 1.8315134570002556,
      "epoch": 0.10942691142542502,
      "grad_norm": 8.588201522827148,
      "learning_rate": 7.647804196908435e-06,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8481560617685318,
      "num_tokens": 42404545.0,
      "step": 35300
    },
    {
      "entropy": 1.9187032520771026,
      "epoch": 0.10945791055047471,
      "grad_norm": 7.530204772949219,
      "learning_rate": 7.646721137948292e-06,
      "loss": 0.6584,
      "mean_token_accuracy": 0.8151246398687363,
      "num_tokens": 42416517.0,
      "step": 35310
    },
    {
      "entropy": 1.8617155611515046,
      "epoch": 0.10948890967552441,
      "grad_norm": 9.000489234924316,
      "learning_rate": 7.64563853899649e-06,
      "loss": 0.5521,
      "mean_token_accuracy": 0.837635052204132,
      "num_tokens": 42429370.0,
      "step": 35320
    },
    {
      "entropy": 1.9121372044086455,
      "epoch": 0.1095199088005741,
      "grad_norm": 11.23980712890625,
      "learning_rate": 7.644556399727486e-06,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8292974248528481,
      "num_tokens": 42440341.0,
      "step": 35330
    },
    {
      "entropy": 1.9182278335094451,
      "epoch": 0.1095509079256238,
      "grad_norm": 10.349918365478516,
      "learning_rate": 7.643474719816064e-06,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8224563241004944,
      "num_tokens": 42451755.0,
      "step": 35340
    },
    {
      "entropy": 1.9242574140429496,
      "epoch": 0.1095819070506735,
      "grad_norm": 8.09676742553711,
      "learning_rate": 7.642393498937326e-06,
      "loss": 0.5653,
      "mean_token_accuracy": 0.8311538890004158,
      "num_tokens": 42463039.0,
      "step": 35350
    },
    {
      "entropy": 1.9706586122512817,
      "epoch": 0.1096129061757232,
      "grad_norm": 10.670557022094727,
      "learning_rate": 7.6413127367667e-06,
      "loss": 0.6166,
      "mean_token_accuracy": 0.817748686671257,
      "num_tokens": 42473647.0,
      "step": 35360
    },
    {
      "entropy": 1.8795671597123147,
      "epoch": 0.10964390530077289,
      "grad_norm": 4.104644775390625,
      "learning_rate": 7.640232432979932e-06,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8274060249328613,
      "num_tokens": 42485582.0,
      "step": 35370
    },
    {
      "entropy": 1.9466561555862427,
      "epoch": 0.10967490442582258,
      "grad_norm": 11.086281776428223,
      "learning_rate": 7.639152587253087e-06,
      "loss": 0.6171,
      "mean_token_accuracy": 0.8145519152283669,
      "num_tokens": 42496181.0,
      "step": 35380
    },
    {
      "entropy": 1.8248122036457062,
      "epoch": 0.10970590355087227,
      "grad_norm": 5.816926956176758,
      "learning_rate": 7.638073199262556e-06,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8266668051481247,
      "num_tokens": 42509157.0,
      "step": 35390
    },
    {
      "entropy": 1.8483053863048553,
      "epoch": 0.10973690267592197,
      "grad_norm": 9.291295051574707,
      "learning_rate": 7.636994268685048e-06,
      "loss": 0.5601,
      "mean_token_accuracy": 0.8281493291258812,
      "num_tokens": 42520936.0,
      "step": 35400
    },
    {
      "entropy": 1.906061513721943,
      "epoch": 0.10976790180097167,
      "grad_norm": 9.213872909545898,
      "learning_rate": 7.635915795197586e-06,
      "loss": 0.577,
      "mean_token_accuracy": 0.8235763564705849,
      "num_tokens": 42531678.0,
      "step": 35410
    },
    {
      "entropy": 1.9281126588582993,
      "epoch": 0.10979890092602136,
      "grad_norm": 9.515471458435059,
      "learning_rate": 7.634837778477519e-06,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8304232120513916,
      "num_tokens": 42542312.0,
      "step": 35420
    },
    {
      "entropy": 1.8663022175431252,
      "epoch": 0.10982990005107106,
      "grad_norm": 4.878605842590332,
      "learning_rate": 7.633760218202513e-06,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8294124737381935,
      "num_tokens": 42554279.0,
      "step": 35430
    },
    {
      "entropy": 1.774543423950672,
      "epoch": 0.10986089917612075,
      "grad_norm": 5.2295002937316895,
      "learning_rate": 7.632683114050551e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8427405998110771,
      "num_tokens": 42567273.0,
      "step": 35440
    },
    {
      "entropy": 1.8794532790780067,
      "epoch": 0.10989189830117045,
      "grad_norm": 9.379169464111328,
      "learning_rate": 7.631606465699934e-06,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8301954373717308,
      "num_tokens": 42579668.0,
      "step": 35450
    },
    {
      "entropy": 1.878871650993824,
      "epoch": 0.10992289742622015,
      "grad_norm": 4.546285152435303,
      "learning_rate": 7.630530272829285e-06,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8306925192475318,
      "num_tokens": 42591107.0,
      "step": 35460
    },
    {
      "entropy": 1.7232867300510406,
      "epoch": 0.10995389655126984,
      "grad_norm": 4.568423748016357,
      "learning_rate": 7.629454535117535e-06,
      "loss": 0.3853,
      "mean_token_accuracy": 0.8541397914290428,
      "num_tokens": 42604850.0,
      "step": 35470
    },
    {
      "entropy": 1.8509867563843727,
      "epoch": 0.10998489567631954,
      "grad_norm": 9.155996322631836,
      "learning_rate": 7.6283792522439415e-06,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8339457020163537,
      "num_tokens": 42616914.0,
      "step": 35480
    },
    {
      "entropy": 1.9788624405860902,
      "epoch": 0.11001589480136924,
      "grad_norm": 9.914673805236816,
      "learning_rate": 7.6273044238880745e-06,
      "loss": 0.6481,
      "mean_token_accuracy": 0.8120816603302956,
      "num_tokens": 42628068.0,
      "step": 35490
    },
    {
      "entropy": 1.8431833282113075,
      "epoch": 0.11004689392641892,
      "grad_norm": 8.730440139770508,
      "learning_rate": 7.626230049729815e-06,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8376294538378716,
      "num_tokens": 42640526.0,
      "step": 35500
    },
    {
      "entropy": 1.8701015904545784,
      "epoch": 0.11007789305146862,
      "grad_norm": 9.417694091796875,
      "learning_rate": 7.625156129449368e-06,
      "loss": 0.5799,
      "mean_token_accuracy": 0.8195733115077019,
      "num_tokens": 42652641.0,
      "step": 35510
    },
    {
      "entropy": 1.9821488350629806,
      "epoch": 0.11010889217651831,
      "grad_norm": 7.322350025177002,
      "learning_rate": 7.624082662727249e-06,
      "loss": 0.6099,
      "mean_token_accuracy": 0.8140304252505303,
      "num_tokens": 42664155.0,
      "step": 35520
    },
    {
      "entropy": 1.8935043781995773,
      "epoch": 0.11013989130156801,
      "grad_norm": 9.178773880004883,
      "learning_rate": 7.623009649244287e-06,
      "loss": 0.513,
      "mean_token_accuracy": 0.8334458202123642,
      "num_tokens": 42676188.0,
      "step": 35530
    },
    {
      "entropy": 1.891274669766426,
      "epoch": 0.1101708904266177,
      "grad_norm": 10.043092727661133,
      "learning_rate": 7.62193708868163e-06,
      "loss": 0.5871,
      "mean_token_accuracy": 0.816723170876503,
      "num_tokens": 42688395.0,
      "step": 35540
    },
    {
      "entropy": 1.8267503723502159,
      "epoch": 0.1102018895516674,
      "grad_norm": 4.575459957122803,
      "learning_rate": 7.620864980720736e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8234075903892517,
      "num_tokens": 42701643.0,
      "step": 35550
    },
    {
      "entropy": 1.956119680404663,
      "epoch": 0.1102328886767171,
      "grad_norm": 7.073038101196289,
      "learning_rate": 7.619793325043378e-06,
      "loss": 0.6057,
      "mean_token_accuracy": 0.8232446178793907,
      "num_tokens": 42713319.0,
      "step": 35560
    },
    {
      "entropy": 1.9920398950576783,
      "epoch": 0.1102638878017668,
      "grad_norm": 8.376012802124023,
      "learning_rate": 7.618722121331642e-06,
      "loss": 0.598,
      "mean_token_accuracy": 0.8307427033782006,
      "num_tokens": 42724006.0,
      "step": 35570
    },
    {
      "entropy": 1.906696754693985,
      "epoch": 0.1102948869268165,
      "grad_norm": 4.264830589294434,
      "learning_rate": 7.617651369267926e-06,
      "loss": 0.591,
      "mean_token_accuracy": 0.8193379983305931,
      "num_tokens": 42735888.0,
      "step": 35580
    },
    {
      "entropy": 1.8427229791879653,
      "epoch": 0.11032588605186619,
      "grad_norm": 9.698928833007812,
      "learning_rate": 7.6165810685349415e-06,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8317845165729523,
      "num_tokens": 42747806.0,
      "step": 35590
    },
    {
      "entropy": 1.8134034514427184,
      "epoch": 0.11035688517691589,
      "grad_norm": 3.2644388675689697,
      "learning_rate": 7.615511218815713e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8382085859775543,
      "num_tokens": 42761313.0,
      "step": 35600
    },
    {
      "entropy": 1.8178164064884186,
      "epoch": 0.11038788430196558,
      "grad_norm": 8.92131233215332,
      "learning_rate": 7.614441819793575e-06,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8241424784064293,
      "num_tokens": 42774872.0,
      "step": 35610
    },
    {
      "entropy": 1.953722095489502,
      "epoch": 0.11041888342701527,
      "grad_norm": 10.026144981384277,
      "learning_rate": 7.61337287115217e-06,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8248757779598236,
      "num_tokens": 42787444.0,
      "step": 35620
    },
    {
      "entropy": 1.8925233513116837,
      "epoch": 0.11044988255206496,
      "grad_norm": 9.82945728302002,
      "learning_rate": 7.612304372575457e-06,
      "loss": 0.5397,
      "mean_token_accuracy": 0.8335311338305473,
      "num_tokens": 42800219.0,
      "step": 35630
    },
    {
      "entropy": 1.9320378288626672,
      "epoch": 0.11048088167711466,
      "grad_norm": 8.172740936279297,
      "learning_rate": 7.611236323747706e-06,
      "loss": 0.567,
      "mean_token_accuracy": 0.8279871761798858,
      "num_tokens": 42811737.0,
      "step": 35640
    },
    {
      "entropy": 1.8768427297472954,
      "epoch": 0.11051188080216436,
      "grad_norm": 4.6713972091674805,
      "learning_rate": 7.610168724353488e-06,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8223695576190948,
      "num_tokens": 42823969.0,
      "step": 35650
    },
    {
      "entropy": 1.8219685062766076,
      "epoch": 0.11054287992721405,
      "grad_norm": 8.601825714111328,
      "learning_rate": 7.6091015740776955e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.8326170966029167,
      "num_tokens": 42835925.0,
      "step": 35660
    },
    {
      "entropy": 1.9021424800157547,
      "epoch": 0.11057387905226375,
      "grad_norm": 8.71493148803711,
      "learning_rate": 7.608034872605521e-06,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8241175979375839,
      "num_tokens": 42847533.0,
      "step": 35670
    },
    {
      "entropy": 1.8227483302354812,
      "epoch": 0.11060487817731345,
      "grad_norm": 9.012811660766602,
      "learning_rate": 7.606968619622469e-06,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8344050481915474,
      "num_tokens": 42860072.0,
      "step": 35680
    },
    {
      "entropy": 1.9262940838932991,
      "epoch": 0.11063587730236314,
      "grad_norm": 8.894078254699707,
      "learning_rate": 7.605902814814354e-06,
      "loss": 0.586,
      "mean_token_accuracy": 0.8217165872454644,
      "num_tokens": 42871958.0,
      "step": 35690
    },
    {
      "entropy": 1.9645220905542373,
      "epoch": 0.11066687642741284,
      "grad_norm": 6.617010116577148,
      "learning_rate": 7.604837457867298e-06,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8164415255188942,
      "num_tokens": 42882974.0,
      "step": 35700
    },
    {
      "entropy": 1.8369232729077338,
      "epoch": 0.11069787555246254,
      "grad_norm": 4.062877178192139,
      "learning_rate": 7.603772548467727e-06,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8306289702653885,
      "num_tokens": 42895928.0,
      "step": 35710
    },
    {
      "entropy": 1.9046246379613876,
      "epoch": 0.11072887467751223,
      "grad_norm": 9.858223915100098,
      "learning_rate": 7.6027080863023806e-06,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8280208811163903,
      "num_tokens": 42908519.0,
      "step": 35720
    },
    {
      "entropy": 1.861171054840088,
      "epoch": 0.11075987380256193,
      "grad_norm": 4.55070686340332,
      "learning_rate": 7.601644071058299e-06,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8228657454252243,
      "num_tokens": 42921353.0,
      "step": 35730
    },
    {
      "entropy": 1.8522720783948898,
      "epoch": 0.11079087292761162,
      "grad_norm": 9.031878471374512,
      "learning_rate": 7.600580502422833e-06,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8284451842308045,
      "num_tokens": 42934541.0,
      "step": 35740
    },
    {
      "entropy": 1.965764506161213,
      "epoch": 0.11082187205266131,
      "grad_norm": 9.1340970993042,
      "learning_rate": 7.59951738008364e-06,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8256940588355064,
      "num_tokens": 42946350.0,
      "step": 35750
    },
    {
      "entropy": 1.9360377460718154,
      "epoch": 0.110852871177711,
      "grad_norm": 9.9710693359375,
      "learning_rate": 7.598454703728679e-06,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8291813790798187,
      "num_tokens": 42957478.0,
      "step": 35760
    },
    {
      "entropy": 1.898517556488514,
      "epoch": 0.1108838703027607,
      "grad_norm": 11.122152328491211,
      "learning_rate": 7.597392473046215e-06,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8223585531115531,
      "num_tokens": 42969607.0,
      "step": 35770
    },
    {
      "entropy": 1.9774446800351142,
      "epoch": 0.1109148694278104,
      "grad_norm": 8.738960266113281,
      "learning_rate": 7.596330687724825e-06,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8237465664744377,
      "num_tokens": 42981808.0,
      "step": 35780
    },
    {
      "entropy": 2.0006524354219435,
      "epoch": 0.1109458685528601,
      "grad_norm": 10.508487701416016,
      "learning_rate": 7.595269347453383e-06,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8199486985802651,
      "num_tokens": 42993002.0,
      "step": 35790
    },
    {
      "entropy": 1.9162476733326912,
      "epoch": 0.11097686767790979,
      "grad_norm": 8.760506629943848,
      "learning_rate": 7.5942084519210665e-06,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8336918845772743,
      "num_tokens": 43005023.0,
      "step": 35800
    },
    {
      "entropy": 1.9015693604946136,
      "epoch": 0.11100786680295949,
      "grad_norm": 3.9503958225250244,
      "learning_rate": 7.5931480008173654e-06,
      "loss": 0.5675,
      "mean_token_accuracy": 0.8282639935612679,
      "num_tokens": 43017177.0,
      "step": 35810
    },
    {
      "entropy": 1.8436180412769319,
      "epoch": 0.11103886592800918,
      "grad_norm": 5.058619022369385,
      "learning_rate": 7.592087993832064e-06,
      "loss": 0.556,
      "mean_token_accuracy": 0.8285689920186996,
      "num_tokens": 43029713.0,
      "step": 35820
    },
    {
      "entropy": 1.934212613105774,
      "epoch": 0.11106986505305888,
      "grad_norm": 9.86316967010498,
      "learning_rate": 7.591028430655252e-06,
      "loss": 0.587,
      "mean_token_accuracy": 0.8228744760155677,
      "num_tokens": 43041006.0,
      "step": 35830
    },
    {
      "entropy": 1.9125287413597107,
      "epoch": 0.11110086417810858,
      "grad_norm": 11.178047180175781,
      "learning_rate": 7.589969310977325e-06,
      "loss": 0.5717,
      "mean_token_accuracy": 0.8275589749217034,
      "num_tokens": 43053262.0,
      "step": 35840
    },
    {
      "entropy": 1.890585133433342,
      "epoch": 0.11113186330315827,
      "grad_norm": 8.734784126281738,
      "learning_rate": 7.588910634488981e-06,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8315401718020439,
      "num_tokens": 43064804.0,
      "step": 35850
    },
    {
      "entropy": 1.9079217493534089,
      "epoch": 0.11116286242820797,
      "grad_norm": 8.155106544494629,
      "learning_rate": 7.587852400881212e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8348406001925468,
      "num_tokens": 43076576.0,
      "step": 35860
    },
    {
      "entropy": 1.9703772380948066,
      "epoch": 0.11119386155325765,
      "grad_norm": 9.1405611038208,
      "learning_rate": 7.586794609845321e-06,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8214395239949226,
      "num_tokens": 43087904.0,
      "step": 35870
    },
    {
      "entropy": 2.0060620248317718,
      "epoch": 0.11122486067830735,
      "grad_norm": 8.357717514038086,
      "learning_rate": 7.58573726107291e-06,
      "loss": 0.6403,
      "mean_token_accuracy": 0.8090036302804947,
      "num_tokens": 43098746.0,
      "step": 35880
    },
    {
      "entropy": 1.9552163541316987,
      "epoch": 0.11125585980335705,
      "grad_norm": 12.635717391967773,
      "learning_rate": 7.584680354255878e-06,
      "loss": 0.6179,
      "mean_token_accuracy": 0.809755727648735,
      "num_tokens": 43109605.0,
      "step": 35890
    },
    {
      "entropy": 1.9294922351837158,
      "epoch": 0.11128685892840674,
      "grad_norm": 7.567071437835693,
      "learning_rate": 7.583623889086426e-06,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8201198145747185,
      "num_tokens": 43121734.0,
      "step": 35900
    },
    {
      "entropy": 1.8198191657662393,
      "epoch": 0.11131785805345644,
      "grad_norm": 10.164730072021484,
      "learning_rate": 7.5825678652570565e-06,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8393372297286987,
      "num_tokens": 43135018.0,
      "step": 35910
    },
    {
      "entropy": 1.9463165909051896,
      "epoch": 0.11134885717850614,
      "grad_norm": 10.353556632995605,
      "learning_rate": 7.58151228246057e-06,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8151269420981407,
      "num_tokens": 43146922.0,
      "step": 35920
    },
    {
      "entropy": 1.8669220060110092,
      "epoch": 0.11137985630355583,
      "grad_norm": 7.747506618499756,
      "learning_rate": 7.5804571403900685e-06,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8381304860115051,
      "num_tokens": 43159179.0,
      "step": 35930
    },
    {
      "entropy": 1.975856500864029,
      "epoch": 0.11141085542860553,
      "grad_norm": 8.58327579498291,
      "learning_rate": 7.579402438738951e-06,
      "loss": 0.6303,
      "mean_token_accuracy": 0.8211771547794342,
      "num_tokens": 43170565.0,
      "step": 35940
    },
    {
      "entropy": 1.8953873470425606,
      "epoch": 0.11144185455365523,
      "grad_norm": 3.883213520050049,
      "learning_rate": 7.578348177200915e-06,
      "loss": 0.548,
      "mean_token_accuracy": 0.8355704694986343,
      "num_tokens": 43182660.0,
      "step": 35950
    },
    {
      "entropy": 1.9930355623364449,
      "epoch": 0.11147285367870492,
      "grad_norm": 7.887125015258789,
      "learning_rate": 7.577294355469956e-06,
      "loss": 0.5559,
      "mean_token_accuracy": 0.8245319411158561,
      "num_tokens": 43194697.0,
      "step": 35960
    },
    {
      "entropy": 1.9415101364254952,
      "epoch": 0.11150385280375462,
      "grad_norm": 9.421870231628418,
      "learning_rate": 7.576240973240371e-06,
      "loss": 0.5932,
      "mean_token_accuracy": 0.8234505504369736,
      "num_tokens": 43206210.0,
      "step": 35970
    },
    {
      "entropy": 1.9392582058906556,
      "epoch": 0.11153485192880432,
      "grad_norm": 8.467351913452148,
      "learning_rate": 7.575188030206747e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8313940614461899,
      "num_tokens": 43218118.0,
      "step": 35980
    },
    {
      "entropy": 1.9223637744784354,
      "epoch": 0.111565851053854,
      "grad_norm": 9.47497272491455,
      "learning_rate": 7.574135526063976e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.840501819550991,
      "num_tokens": 43229669.0,
      "step": 35990
    },
    {
      "entropy": 2.0074471473693847,
      "epoch": 0.1115968501789037,
      "grad_norm": 10.357573509216309,
      "learning_rate": 7.5730834605072416e-06,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8227631166577339,
      "num_tokens": 43240593.0,
      "step": 36000
    },
    {
      "entropy": 1.859610728919506,
      "epoch": 0.11162784930395339,
      "grad_norm": 9.257423400878906,
      "learning_rate": 7.5720318332320255e-06,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8329069197177887,
      "num_tokens": 43253611.0,
      "step": 36010
    },
    {
      "entropy": 1.9028906911611556,
      "epoch": 0.11165884842900309,
      "grad_norm": 8.12290096282959,
      "learning_rate": 7.570980643934104e-06,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8339066132903099,
      "num_tokens": 43265093.0,
      "step": 36020
    },
    {
      "entropy": 1.8995389580726623,
      "epoch": 0.11168984755405278,
      "grad_norm": 9.205475807189941,
      "learning_rate": 7.56992989230955e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8334958225488662,
      "num_tokens": 43277585.0,
      "step": 36030
    },
    {
      "entropy": 1.9100613698363305,
      "epoch": 0.11172084667910248,
      "grad_norm": 7.445013046264648,
      "learning_rate": 7.5688795780547335e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.842551352083683,
      "num_tokens": 43289300.0,
      "step": 36040
    },
    {
      "entropy": 1.9307687520980834,
      "epoch": 0.11175184580415218,
      "grad_norm": 9.54047966003418,
      "learning_rate": 7.5678297008663135e-06,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8196006685495376,
      "num_tokens": 43300995.0,
      "step": 36050
    },
    {
      "entropy": 1.9431654140353203,
      "epoch": 0.11178284492920187,
      "grad_norm": 8.615473747253418,
      "learning_rate": 7.566780260441252e-06,
      "loss": 0.6086,
      "mean_token_accuracy": 0.8174616977572441,
      "num_tokens": 43312218.0,
      "step": 36060
    },
    {
      "entropy": 1.8542766660451888,
      "epoch": 0.11181384405425157,
      "grad_norm": 8.529784202575684,
      "learning_rate": 7.565731256476797e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8297783643007278,
      "num_tokens": 43325617.0,
      "step": 36070
    },
    {
      "entropy": 1.9352862015366554,
      "epoch": 0.11184484317930127,
      "grad_norm": 8.217283248901367,
      "learning_rate": 7.564682688670496e-06,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8212688997387886,
      "num_tokens": 43337245.0,
      "step": 36080
    },
    {
      "entropy": 1.972224122285843,
      "epoch": 0.11187584230435096,
      "grad_norm": 10.360859870910645,
      "learning_rate": 7.563634556720185e-06,
      "loss": 0.567,
      "mean_token_accuracy": 0.8349591135978699,
      "num_tokens": 43348686.0,
      "step": 36090
    },
    {
      "entropy": 1.9579823553562163,
      "epoch": 0.11190684142940066,
      "grad_norm": 9.387046813964844,
      "learning_rate": 7.562586860323996e-06,
      "loss": 0.5886,
      "mean_token_accuracy": 0.8227823153138161,
      "num_tokens": 43359689.0,
      "step": 36100
    },
    {
      "entropy": 1.9274040132761001,
      "epoch": 0.11193784055445036,
      "grad_norm": 4.451748371124268,
      "learning_rate": 7.561539599180354e-06,
      "loss": 0.5332,
      "mean_token_accuracy": 0.8205605849623681,
      "num_tokens": 43372242.0,
      "step": 36110
    },
    {
      "entropy": 1.8405972003936768,
      "epoch": 0.11196883967950004,
      "grad_norm": 4.114645481109619,
      "learning_rate": 7.560492772987975e-06,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8403662115335464,
      "num_tokens": 43385797.0,
      "step": 36120
    },
    {
      "entropy": 1.9486818492412568,
      "epoch": 0.11199983880454974,
      "grad_norm": 4.341696739196777,
      "learning_rate": 7.5594463814458676e-06,
      "loss": 0.5247,
      "mean_token_accuracy": 0.839286656677723,
      "num_tokens": 43396906.0,
      "step": 36130
    },
    {
      "entropy": 1.9541378676891328,
      "epoch": 0.11203083792959943,
      "grad_norm": 9.982046127319336,
      "learning_rate": 7.558400424253328e-06,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8157087191939354,
      "num_tokens": 43409073.0,
      "step": 36140
    },
    {
      "entropy": 1.9935852527618407,
      "epoch": 0.11206183705464913,
      "grad_norm": 8.649558067321777,
      "learning_rate": 7.557354901109952e-06,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8184923201799392,
      "num_tokens": 43420300.0,
      "step": 36150
    },
    {
      "entropy": 1.9860820427536965,
      "epoch": 0.11209283617969883,
      "grad_norm": 9.3584566116333,
      "learning_rate": 7.556309811715618e-06,
      "loss": 0.5715,
      "mean_token_accuracy": 0.8259601920843125,
      "num_tokens": 43431655.0,
      "step": 36160
    },
    {
      "entropy": 1.9509647816419602,
      "epoch": 0.11212383530474852,
      "grad_norm": 10.400931358337402,
      "learning_rate": 7.5552651557705e-06,
      "loss": 0.598,
      "mean_token_accuracy": 0.8289371073246002,
      "num_tokens": 43443051.0,
      "step": 36170
    },
    {
      "entropy": 1.8806715980172157,
      "epoch": 0.11215483442979822,
      "grad_norm": 11.046425819396973,
      "learning_rate": 7.55422093297506e-06,
      "loss": 0.547,
      "mean_token_accuracy": 0.8374142736196518,
      "num_tokens": 43455604.0,
      "step": 36180
    },
    {
      "entropy": 1.9038266450166703,
      "epoch": 0.11218583355484792,
      "grad_norm": 4.295387268066406,
      "learning_rate": 7.553177143030047e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8337746739387513,
      "num_tokens": 43468108.0,
      "step": 36190
    },
    {
      "entropy": 1.8910601362586021,
      "epoch": 0.11221683267989761,
      "grad_norm": 8.46925163269043,
      "learning_rate": 7.5521337856365064e-06,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8458377867937088,
      "num_tokens": 43480308.0,
      "step": 36200
    },
    {
      "entropy": 1.9406689956784249,
      "epoch": 0.11224783180494731,
      "grad_norm": 8.312734603881836,
      "learning_rate": 7.551090860495766e-06,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8199841871857643,
      "num_tokens": 43492821.0,
      "step": 36210
    },
    {
      "entropy": 1.9295011416077614,
      "epoch": 0.112278830929997,
      "grad_norm": 4.657339096069336,
      "learning_rate": 7.550048367309445e-06,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8267869040369987,
      "num_tokens": 43504911.0,
      "step": 36220
    },
    {
      "entropy": 1.8883660644292832,
      "epoch": 0.1123098300550467,
      "grad_norm": 12.443599700927734,
      "learning_rate": 7.54900630577945e-06,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8197856619954109,
      "num_tokens": 43517464.0,
      "step": 36230
    },
    {
      "entropy": 1.9811781361699103,
      "epoch": 0.11234082918009639,
      "grad_norm": 11.399888038635254,
      "learning_rate": 7.547964675607977e-06,
      "loss": 0.6267,
      "mean_token_accuracy": 0.8182986229658127,
      "num_tokens": 43528891.0,
      "step": 36240
    },
    {
      "entropy": 1.847057183086872,
      "epoch": 0.11237182830514608,
      "grad_norm": 7.622710704803467,
      "learning_rate": 7.546923476497509e-06,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8390681058168411,
      "num_tokens": 43541805.0,
      "step": 36250
    },
    {
      "entropy": 1.9104607298970222,
      "epoch": 0.11240282743019578,
      "grad_norm": 9.49293041229248,
      "learning_rate": 7.545882708150815e-06,
      "loss": 0.5786,
      "mean_token_accuracy": 0.8211572468280792,
      "num_tokens": 43552709.0,
      "step": 36260
    },
    {
      "entropy": 1.7649886459112167,
      "epoch": 0.11243382655524548,
      "grad_norm": 4.7445807456970215,
      "learning_rate": 7.544842370270952e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.841453991830349,
      "num_tokens": 43566226.0,
      "step": 36270
    },
    {
      "entropy": 1.9194318518042564,
      "epoch": 0.11246482568029517,
      "grad_norm": 4.042590618133545,
      "learning_rate": 7.543802462561263e-06,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8197874516248703,
      "num_tokens": 43578124.0,
      "step": 36280
    },
    {
      "entropy": 1.8747497126460075,
      "epoch": 0.11249582480534487,
      "grad_norm": 8.346511840820312,
      "learning_rate": 7.5427629847253766e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8315855219960213,
      "num_tokens": 43590273.0,
      "step": 36290
    },
    {
      "entropy": 1.962376557290554,
      "epoch": 0.11252682393039456,
      "grad_norm": 7.562712669372559,
      "learning_rate": 7.541723936467211e-06,
      "loss": 0.541,
      "mean_token_accuracy": 0.8308800607919693,
      "num_tokens": 43601348.0,
      "step": 36300
    },
    {
      "entropy": 1.9383685559034347,
      "epoch": 0.11255782305544426,
      "grad_norm": 4.1252875328063965,
      "learning_rate": 7.540685317490964e-06,
      "loss": 0.6044,
      "mean_token_accuracy": 0.8207226291298866,
      "num_tokens": 43613086.0,
      "step": 36310
    },
    {
      "entropy": 1.9785435765981674,
      "epoch": 0.11258882218049396,
      "grad_norm": 9.106582641601562,
      "learning_rate": 7.539647127501121e-06,
      "loss": 0.6021,
      "mean_token_accuracy": 0.8204389974474907,
      "num_tokens": 43623726.0,
      "step": 36320
    },
    {
      "entropy": 1.9137952119112014,
      "epoch": 0.11261982130554365,
      "grad_norm": 8.635375022888184,
      "learning_rate": 7.5386093662024515e-06,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8232575729489326,
      "num_tokens": 43635777.0,
      "step": 36330
    },
    {
      "entropy": 2.0330847591161727,
      "epoch": 0.11265082043059335,
      "grad_norm": 10.423239707946777,
      "learning_rate": 7.537572033300013e-06,
      "loss": 0.6766,
      "mean_token_accuracy": 0.8112093076109886,
      "num_tokens": 43646664.0,
      "step": 36340
    },
    {
      "entropy": 2.00158928334713,
      "epoch": 0.11268181955564305,
      "grad_norm": 8.458577156066895,
      "learning_rate": 7.536535128499144e-06,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8329016759991645,
      "num_tokens": 43657222.0,
      "step": 36350
    },
    {
      "entropy": 1.9344168439507485,
      "epoch": 0.11271281868069273,
      "grad_norm": 5.806880950927734,
      "learning_rate": 7.535498651505465e-06,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8308494806289672,
      "num_tokens": 43669133.0,
      "step": 36360
    },
    {
      "entropy": 1.9564279466867447,
      "epoch": 0.11274381780574243,
      "grad_norm": 7.9867329597473145,
      "learning_rate": 7.5344626020248825e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8347235858440399,
      "num_tokens": 43680521.0,
      "step": 36370
    },
    {
      "entropy": 1.9638302087783814,
      "epoch": 0.11277481693079212,
      "grad_norm": 9.185903549194336,
      "learning_rate": 7.533426979763585e-06,
      "loss": 0.6052,
      "mean_token_accuracy": 0.8131253823637963,
      "num_tokens": 43691642.0,
      "step": 36380
    },
    {
      "entropy": 1.910866443812847,
      "epoch": 0.11280581605584182,
      "grad_norm": 8.041609764099121,
      "learning_rate": 7.532391784428045e-06,
      "loss": 0.5606,
      "mean_token_accuracy": 0.8212843969464302,
      "num_tokens": 43703753.0,
      "step": 36390
    },
    {
      "entropy": 1.7988116875290872,
      "epoch": 0.11283681518089152,
      "grad_norm": 8.769583702087402,
      "learning_rate": 7.531357015725014e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8505406931042672,
      "num_tokens": 43716887.0,
      "step": 36400
    },
    {
      "entropy": 1.8762123107910156,
      "epoch": 0.11286781430594121,
      "grad_norm": 7.339711666107178,
      "learning_rate": 7.53032267336153e-06,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8283303081989288,
      "num_tokens": 43728905.0,
      "step": 36410
    },
    {
      "entropy": 2.0162509113550184,
      "epoch": 0.11289881343099091,
      "grad_norm": 8.925753593444824,
      "learning_rate": 7.529288757044908e-06,
      "loss": 0.6202,
      "mean_token_accuracy": 0.8204985901713371,
      "num_tokens": 43739992.0,
      "step": 36420
    },
    {
      "entropy": 1.8500737398862839,
      "epoch": 0.1129298125560406,
      "grad_norm": 4.230498790740967,
      "learning_rate": 7.528255266482748e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8447869628667831,
      "num_tokens": 43752806.0,
      "step": 36430
    },
    {
      "entropy": 1.9196542382240296,
      "epoch": 0.1129608116810903,
      "grad_norm": 10.557394981384277,
      "learning_rate": 7.527222201382927e-06,
      "loss": 0.56,
      "mean_token_accuracy": 0.8287433817982673,
      "num_tokens": 43764530.0,
      "step": 36440
    },
    {
      "entropy": 1.937133614718914,
      "epoch": 0.11299181080614,
      "grad_norm": 9.187378883361816,
      "learning_rate": 7.526189561453605e-06,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8356496065855026,
      "num_tokens": 43775958.0,
      "step": 36450
    },
    {
      "entropy": 1.8305316418409348,
      "epoch": 0.1130228099311897,
      "grad_norm": 10.253772735595703,
      "learning_rate": 7.525157346403224e-06,
      "loss": 0.4758,
      "mean_token_accuracy": 0.846756660938263,
      "num_tokens": 43788156.0,
      "step": 36460
    },
    {
      "entropy": 1.8645455896854402,
      "epoch": 0.1130538090562394,
      "grad_norm": 2.7411298751831055,
      "learning_rate": 7.5241255559405015e-06,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8278805449604988,
      "num_tokens": 43800701.0,
      "step": 36470
    },
    {
      "entropy": 1.974195511639118,
      "epoch": 0.11308480818128909,
      "grad_norm": 4.0369110107421875,
      "learning_rate": 7.523094189774437e-06,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8239115789532662,
      "num_tokens": 43811490.0,
      "step": 36480
    },
    {
      "entropy": 1.8882897064089774,
      "epoch": 0.11311580730633877,
      "grad_norm": 9.004681587219238,
      "learning_rate": 7.5220632476143095e-06,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8223536923527718,
      "num_tokens": 43823579.0,
      "step": 36490
    },
    {
      "entropy": 1.8579727187752724,
      "epoch": 0.11314680643138847,
      "grad_norm": 4.422814846038818,
      "learning_rate": 7.521032729169676e-06,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8264896929264068,
      "num_tokens": 43835758.0,
      "step": 36500
    },
    {
      "entropy": 1.8324081644415855,
      "epoch": 0.11317780555643817,
      "grad_norm": 4.702271461486816,
      "learning_rate": 7.520002634150373e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8371421471238136,
      "num_tokens": 43848630.0,
      "step": 36510
    },
    {
      "entropy": 1.8566824480891229,
      "epoch": 0.11320880468148786,
      "grad_norm": 4.19157075881958,
      "learning_rate": 7.518972962266511e-06,
      "loss": 0.6098,
      "mean_token_accuracy": 0.8109669283032417,
      "num_tokens": 43861241.0,
      "step": 36520
    },
    {
      "entropy": 1.8918891534209252,
      "epoch": 0.11323980380653756,
      "grad_norm": 4.325606346130371,
      "learning_rate": 7.517943713228485e-06,
      "loss": 0.5829,
      "mean_token_accuracy": 0.8150859504938126,
      "num_tokens": 43872873.0,
      "step": 36530
    },
    {
      "entropy": 1.8818762391805648,
      "epoch": 0.11327080293158726,
      "grad_norm": 8.26830768585205,
      "learning_rate": 7.516914886746964e-06,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8375229358673095,
      "num_tokens": 43885092.0,
      "step": 36540
    },
    {
      "entropy": 1.8862257033586503,
      "epoch": 0.11330180205663695,
      "grad_norm": 4.080634117126465,
      "learning_rate": 7.515886482532891e-06,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8279654592275619,
      "num_tokens": 43897930.0,
      "step": 36550
    },
    {
      "entropy": 1.9121008217334747,
      "epoch": 0.11333280118168665,
      "grad_norm": 7.579226970672607,
      "learning_rate": 7.5148585002974895e-06,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8319680899381637,
      "num_tokens": 43910576.0,
      "step": 36560
    },
    {
      "entropy": 1.8942013755440712,
      "epoch": 0.11336380030673635,
      "grad_norm": 10.820945739746094,
      "learning_rate": 7.513830939752261e-06,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8288922995328903,
      "num_tokens": 43922364.0,
      "step": 36570
    },
    {
      "entropy": 1.8880238860845566,
      "epoch": 0.11339479943178604,
      "grad_norm": 4.12148904800415,
      "learning_rate": 7.512803800608977e-06,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8186680763959885,
      "num_tokens": 43934444.0,
      "step": 36580
    },
    {
      "entropy": 1.8842089757323266,
      "epoch": 0.11342579855683574,
      "grad_norm": 7.816810131072998,
      "learning_rate": 7.511777082579692e-06,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8386390700936317,
      "num_tokens": 43946333.0,
      "step": 36590
    },
    {
      "entropy": 1.905991567671299,
      "epoch": 0.11345679768188544,
      "grad_norm": 9.459228515625,
      "learning_rate": 7.5107507853767304e-06,
      "loss": 0.556,
      "mean_token_accuracy": 0.8341964855790138,
      "num_tokens": 43958369.0,
      "step": 36600
    },
    {
      "entropy": 1.9278974682092667,
      "epoch": 0.11348779680693512,
      "grad_norm": 9.090117454528809,
      "learning_rate": 7.509724908712693e-06,
      "loss": 0.531,
      "mean_token_accuracy": 0.8365562587976456,
      "num_tokens": 43970116.0,
      "step": 36610
    },
    {
      "entropy": 1.95464196652174,
      "epoch": 0.11351879593198481,
      "grad_norm": 8.458623886108398,
      "learning_rate": 7.508699452300459e-06,
      "loss": 0.5787,
      "mean_token_accuracy": 0.830603589117527,
      "num_tokens": 43981286.0,
      "step": 36620
    },
    {
      "entropy": 1.839498682320118,
      "epoch": 0.11354979505703451,
      "grad_norm": 8.516286849975586,
      "learning_rate": 7.507674415853176e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8414850428700447,
      "num_tokens": 43993909.0,
      "step": 36630
    },
    {
      "entropy": 1.936448486149311,
      "epoch": 0.11358079418208421,
      "grad_norm": 4.347235202789307,
      "learning_rate": 7.506649799084268e-06,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8339558869600296,
      "num_tokens": 44005833.0,
      "step": 36640
    },
    {
      "entropy": 1.884600205719471,
      "epoch": 0.1136117933071339,
      "grad_norm": 7.8523712158203125,
      "learning_rate": 7.505625601707435e-06,
      "loss": 0.517,
      "mean_token_accuracy": 0.8333307653665543,
      "num_tokens": 44018334.0,
      "step": 36650
    },
    {
      "entropy": 1.8594166815280915,
      "epoch": 0.1136427924321836,
      "grad_norm": 8.999347686767578,
      "learning_rate": 7.504601823436648e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8310374036431313,
      "num_tokens": 44030326.0,
      "step": 36660
    },
    {
      "entropy": 1.9014796167612076,
      "epoch": 0.1136737915572333,
      "grad_norm": 7.825422763824463,
      "learning_rate": 7.503578463986152e-06,
      "loss": 0.579,
      "mean_token_accuracy": 0.816353191435337,
      "num_tokens": 44042199.0,
      "step": 36670
    },
    {
      "entropy": 1.7826032146811486,
      "epoch": 0.113704790682283,
      "grad_norm": 9.484929084777832,
      "learning_rate": 7.502555523070463e-06,
      "loss": 0.4162,
      "mean_token_accuracy": 0.851135803759098,
      "num_tokens": 44055210.0,
      "step": 36680
    },
    {
      "entropy": 1.8730618610978127,
      "epoch": 0.11373578980733269,
      "grad_norm": 10.11883544921875,
      "learning_rate": 7.5015330004043705e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8271143138408661,
      "num_tokens": 44066956.0,
      "step": 36690
    },
    {
      "entropy": 1.9547926247119904,
      "epoch": 0.11376678893238239,
      "grad_norm": 9.232172012329102,
      "learning_rate": 7.500510895702939e-06,
      "loss": 0.6011,
      "mean_token_accuracy": 0.8220591425895691,
      "num_tokens": 44078247.0,
      "step": 36700
    },
    {
      "entropy": 1.8781561613082887,
      "epoch": 0.11379778805743208,
      "grad_norm": 7.583634376525879,
      "learning_rate": 7.499489208681497e-06,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8380724370479584,
      "num_tokens": 44089827.0,
      "step": 36710
    },
    {
      "entropy": 1.822493526339531,
      "epoch": 0.11382878718248178,
      "grad_norm": 4.266534328460693,
      "learning_rate": 7.498467939055656e-06,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8341230794787406,
      "num_tokens": 44102605.0,
      "step": 36720
    },
    {
      "entropy": 1.8786309957504272,
      "epoch": 0.11385978630753146,
      "grad_norm": 9.360684394836426,
      "learning_rate": 7.497447086541285e-06,
      "loss": 0.5763,
      "mean_token_accuracy": 0.8306545913219452,
      "num_tokens": 44114636.0,
      "step": 36730
    },
    {
      "entropy": 1.904089206457138,
      "epoch": 0.11389078543258116,
      "grad_norm": 5.562136650085449,
      "learning_rate": 7.496426650854535e-06,
      "loss": 0.6014,
      "mean_token_accuracy": 0.8137870237231255,
      "num_tokens": 44127313.0,
      "step": 36740
    },
    {
      "entropy": 1.9380560591816902,
      "epoch": 0.11392178455763086,
      "grad_norm": 9.176344871520996,
      "learning_rate": 7.4954066317118205e-06,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8196625456213951,
      "num_tokens": 44138997.0,
      "step": 36750
    },
    {
      "entropy": 1.881463260948658,
      "epoch": 0.11395278368268055,
      "grad_norm": 8.139053344726562,
      "learning_rate": 7.494387028829828e-06,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8382967382669448,
      "num_tokens": 44151672.0,
      "step": 36760
    },
    {
      "entropy": 1.8621976956725121,
      "epoch": 0.11398378280773025,
      "grad_norm": 4.298003673553467,
      "learning_rate": 7.493367841925514e-06,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8231415241956711,
      "num_tokens": 44164318.0,
      "step": 36770
    },
    {
      "entropy": 1.9737728282809257,
      "epoch": 0.11401478193277995,
      "grad_norm": 10.217061996459961,
      "learning_rate": 7.492349070716108e-06,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8322141289710998,
      "num_tokens": 44175114.0,
      "step": 36780
    },
    {
      "entropy": 1.9328984022140503,
      "epoch": 0.11404578105782964,
      "grad_norm": 4.168199062347412,
      "learning_rate": 7.4913307149191e-06,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8284803554415703,
      "num_tokens": 44187241.0,
      "step": 36790
    },
    {
      "entropy": 1.9318128556013108,
      "epoch": 0.11407678018287934,
      "grad_norm": 9.854540824890137,
      "learning_rate": 7.490312774252257e-06,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8278173923492431,
      "num_tokens": 44197838.0,
      "step": 36800
    },
    {
      "entropy": 1.8671427622437478,
      "epoch": 0.11410777930792904,
      "grad_norm": 8.820489883422852,
      "learning_rate": 7.489295248433609e-06,
      "loss": 0.6118,
      "mean_token_accuracy": 0.8205173119902611,
      "num_tokens": 44209989.0,
      "step": 36810
    },
    {
      "entropy": 1.9128638491034509,
      "epoch": 0.11413877843297873,
      "grad_norm": 10.370189666748047,
      "learning_rate": 7.488278137181456e-06,
      "loss": 0.593,
      "mean_token_accuracy": 0.8210515394806862,
      "num_tokens": 44221247.0,
      "step": 36820
    },
    {
      "entropy": 1.8790946841239928,
      "epoch": 0.11416977755802843,
      "grad_norm": 8.685145378112793,
      "learning_rate": 7.48726144021437e-06,
      "loss": 0.512,
      "mean_token_accuracy": 0.8406370490789413,
      "num_tokens": 44233155.0,
      "step": 36830
    },
    {
      "entropy": 1.973842205107212,
      "epoch": 0.11420077668307813,
      "grad_norm": 9.793169021606445,
      "learning_rate": 7.48624515725118e-06,
      "loss": 0.5913,
      "mean_token_accuracy": 0.8233802810311317,
      "num_tokens": 44243970.0,
      "step": 36840
    },
    {
      "entropy": 1.800303579866886,
      "epoch": 0.11423177580812782,
      "grad_norm": 9.683552742004395,
      "learning_rate": 7.485229288010991e-06,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8449695229530334,
      "num_tokens": 44257962.0,
      "step": 36850
    },
    {
      "entropy": 1.8852539584040642,
      "epoch": 0.1142627749331775,
      "grad_norm": 8.617185592651367,
      "learning_rate": 7.484213832213174e-06,
      "loss": 0.531,
      "mean_token_accuracy": 0.8190362945199012,
      "num_tokens": 44270821.0,
      "step": 36860
    },
    {
      "entropy": 1.8611651435494423,
      "epoch": 0.1142937740582272,
      "grad_norm": 3.9128634929656982,
      "learning_rate": 7.483198789577362e-06,
      "loss": 0.5256,
      "mean_token_accuracy": 0.825941427052021,
      "num_tokens": 44283425.0,
      "step": 36870
    },
    {
      "entropy": 1.863780789077282,
      "epoch": 0.1143247731832769,
      "grad_norm": 9.122457504272461,
      "learning_rate": 7.482184159823459e-06,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8322040572762489,
      "num_tokens": 44295820.0,
      "step": 36880
    },
    {
      "entropy": 1.8578864350914954,
      "epoch": 0.1143557723083266,
      "grad_norm": 4.551950931549072,
      "learning_rate": 7.481169942671628e-06,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8270832076668739,
      "num_tokens": 44308102.0,
      "step": 36890
    },
    {
      "entropy": 1.9780358895659447,
      "epoch": 0.11438677143337629,
      "grad_norm": 9.209576606750488,
      "learning_rate": 7.480156137842306e-06,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8193370506167412,
      "num_tokens": 44319023.0,
      "step": 36900
    },
    {
      "entropy": 1.9365268610417843,
      "epoch": 0.11441777055842599,
      "grad_norm": 9.44648551940918,
      "learning_rate": 7.479142745056188e-06,
      "loss": 0.5477,
      "mean_token_accuracy": 0.8305159211158752,
      "num_tokens": 44330798.0,
      "step": 36910
    },
    {
      "entropy": 1.9427412554621697,
      "epoch": 0.11444876968347568,
      "grad_norm": 9.822259902954102,
      "learning_rate": 7.478129764034238e-06,
      "loss": 0.5697,
      "mean_token_accuracy": 0.8291716650128365,
      "num_tokens": 44341817.0,
      "step": 36920
    },
    {
      "entropy": 1.884063209593296,
      "epoch": 0.11447976880852538,
      "grad_norm": 9.40356731414795,
      "learning_rate": 7.477117194497685e-06,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8296123817563057,
      "num_tokens": 44354019.0,
      "step": 36930
    },
    {
      "entropy": 1.8861080020666123,
      "epoch": 0.11451076793357508,
      "grad_norm": 10.153528213500977,
      "learning_rate": 7.476105036168018e-06,
      "loss": 0.548,
      "mean_token_accuracy": 0.8213342532515526,
      "num_tokens": 44366926.0,
      "step": 36940
    },
    {
      "entropy": 1.8933350324630738,
      "epoch": 0.11454176705862477,
      "grad_norm": 9.459976196289062,
      "learning_rate": 7.475093288766992e-06,
      "loss": 0.5634,
      "mean_token_accuracy": 0.828109310567379,
      "num_tokens": 44378877.0,
      "step": 36950
    },
    {
      "entropy": 1.8734344974160195,
      "epoch": 0.11457276618367447,
      "grad_norm": 8.441810607910156,
      "learning_rate": 7.474081952016626e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8315728649497032,
      "num_tokens": 44392038.0,
      "step": 36960
    },
    {
      "entropy": 1.874902254343033,
      "epoch": 0.11460376530872417,
      "grad_norm": 4.568936347961426,
      "learning_rate": 7.473071025639202e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8361835688352585,
      "num_tokens": 44404425.0,
      "step": 36970
    },
    {
      "entropy": 1.9008426293730736,
      "epoch": 0.11463476443377385,
      "grad_norm": 9.730356216430664,
      "learning_rate": 7.4720605093572664e-06,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8256875947117805,
      "num_tokens": 44416109.0,
      "step": 36980
    },
    {
      "entropy": 1.925182616710663,
      "epoch": 0.11466576355882355,
      "grad_norm": 3.8875186443328857,
      "learning_rate": 7.471050402893625e-06,
      "loss": 0.611,
      "mean_token_accuracy": 0.8156601503491402,
      "num_tokens": 44427762.0,
      "step": 36990
    },
    {
      "entropy": 1.8984755888581275,
      "epoch": 0.11469676268387324,
      "grad_norm": 8.992522239685059,
      "learning_rate": 7.470040705971346e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8337834537029266,
      "num_tokens": 44439999.0,
      "step": 37000
    },
    {
      "entropy": 1.9663574412465095,
      "epoch": 0.11472776180892294,
      "grad_norm": 8.80916690826416,
      "learning_rate": 7.469031418313763e-06,
      "loss": 0.571,
      "mean_token_accuracy": 0.8257024556398391,
      "num_tokens": 44451506.0,
      "step": 37010
    },
    {
      "entropy": 1.8491823315620421,
      "epoch": 0.11475876093397264,
      "grad_norm": 4.779745101928711,
      "learning_rate": 7.46802253964447e-06,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8266303971409797,
      "num_tokens": 44464222.0,
      "step": 37020
    },
    {
      "entropy": 1.9856320530176164,
      "epoch": 0.11478976005902233,
      "grad_norm": 7.643355846405029,
      "learning_rate": 7.46701406968732e-06,
      "loss": 0.6223,
      "mean_token_accuracy": 0.8207634121179581,
      "num_tokens": 44475276.0,
      "step": 37030
    },
    {
      "entropy": 1.940791991353035,
      "epoch": 0.11482075918407203,
      "grad_norm": 5.0710954666137695,
      "learning_rate": 7.466006008166425e-06,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8188979029655457,
      "num_tokens": 44487103.0,
      "step": 37040
    },
    {
      "entropy": 1.9016083896160125,
      "epoch": 0.11485175830912173,
      "grad_norm": 4.698379039764404,
      "learning_rate": 7.464998354806166e-06,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8199516862630845,
      "num_tokens": 44498694.0,
      "step": 37050
    },
    {
      "entropy": 1.9881937965750693,
      "epoch": 0.11488275743417142,
      "grad_norm": 9.964612007141113,
      "learning_rate": 7.463991109331177e-06,
      "loss": 0.618,
      "mean_token_accuracy": 0.8096556261181831,
      "num_tokens": 44509990.0,
      "step": 37060
    },
    {
      "entropy": 1.884983916580677,
      "epoch": 0.11491375655922112,
      "grad_norm": 9.16238021850586,
      "learning_rate": 7.462984271466356e-06,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8483946248888969,
      "num_tokens": 44521793.0,
      "step": 37070
    },
    {
      "entropy": 1.9347106859087944,
      "epoch": 0.11494475568427082,
      "grad_norm": 8.995933532714844,
      "learning_rate": 7.461977840936856e-06,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8224361136555671,
      "num_tokens": 44533297.0,
      "step": 37080
    },
    {
      "entropy": 1.9060018703341484,
      "epoch": 0.11497575480932051,
      "grad_norm": 8.693288803100586,
      "learning_rate": 7.460971817468093e-06,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8281079009175301,
      "num_tokens": 44545609.0,
      "step": 37090
    },
    {
      "entropy": 1.8859413847327233,
      "epoch": 0.1150067539343702,
      "grad_norm": 7.988814353942871,
      "learning_rate": 7.459966200785744e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8470205634832382,
      "num_tokens": 44557728.0,
      "step": 37100
    },
    {
      "entropy": 1.8287798702716827,
      "epoch": 0.11503775305941989,
      "grad_norm": 7.7615437507629395,
      "learning_rate": 7.45896099061574e-06,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8320494025945664,
      "num_tokens": 44571340.0,
      "step": 37110
    },
    {
      "entropy": 1.91844123005867,
      "epoch": 0.11506875218446959,
      "grad_norm": 11.6615629196167,
      "learning_rate": 7.457956186684274e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8266859829425812,
      "num_tokens": 44582865.0,
      "step": 37120
    },
    {
      "entropy": 1.7767401605844497,
      "epoch": 0.11509975130951929,
      "grad_norm": 7.414669513702393,
      "learning_rate": 7.4569517887177935e-06,
      "loss": 0.4635,
      "mean_token_accuracy": 0.8381605073809624,
      "num_tokens": 44596679.0,
      "step": 37130
    },
    {
      "entropy": 1.9388007283210755,
      "epoch": 0.11513075043456898,
      "grad_norm": 11.721390724182129,
      "learning_rate": 7.455947796443009e-06,
      "loss": 0.6072,
      "mean_token_accuracy": 0.8257848516106605,
      "num_tokens": 44608325.0,
      "step": 37140
    },
    {
      "entropy": 1.9356591030955315,
      "epoch": 0.11516174955961868,
      "grad_norm": 7.786122798919678,
      "learning_rate": 7.454944209586882e-06,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8272521525621415,
      "num_tokens": 44619842.0,
      "step": 37150
    },
    {
      "entropy": 1.8918299242854117,
      "epoch": 0.11519274868466838,
      "grad_norm": 10.668729782104492,
      "learning_rate": 7.453941027876637e-06,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8250093191862107,
      "num_tokens": 44631902.0,
      "step": 37160
    },
    {
      "entropy": 1.7994364276528358,
      "epoch": 0.11522374780971807,
      "grad_norm": 6.118994235992432,
      "learning_rate": 7.452938251039754e-06,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8353979960083961,
      "num_tokens": 44645016.0,
      "step": 37170
    },
    {
      "entropy": 1.9420592486858368,
      "epoch": 0.11525474693476777,
      "grad_norm": 9.136016845703125,
      "learning_rate": 7.451935878803968e-06,
      "loss": 0.5985,
      "mean_token_accuracy": 0.8215920001268386,
      "num_tokens": 44656784.0,
      "step": 37180
    },
    {
      "entropy": 1.9028725042939185,
      "epoch": 0.11528574605981746,
      "grad_norm": 7.933960914611816,
      "learning_rate": 7.45093391089727e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8299235314130783,
      "num_tokens": 44669372.0,
      "step": 37190
    },
    {
      "entropy": 1.8886778056621552,
      "epoch": 0.11531674518486716,
      "grad_norm": 8.652473449707031,
      "learning_rate": 7.4499323470479075e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.835668683052063,
      "num_tokens": 44681290.0,
      "step": 37200
    },
    {
      "entropy": 1.902733063697815,
      "epoch": 0.11534774430991686,
      "grad_norm": 9.28559684753418,
      "learning_rate": 7.448931186984387e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8212042555212975,
      "num_tokens": 44692941.0,
      "step": 37210
    },
    {
      "entropy": 1.900286576151848,
      "epoch": 0.11537874343496655,
      "grad_norm": 8.697643280029297,
      "learning_rate": 7.447930430435463e-06,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8290207460522652,
      "num_tokens": 44704646.0,
      "step": 37220
    },
    {
      "entropy": 1.8271199837327003,
      "epoch": 0.11540974256001624,
      "grad_norm": 10.892705917358398,
      "learning_rate": 7.446930077130154e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8422973319888115,
      "num_tokens": 44717316.0,
      "step": 37230
    },
    {
      "entropy": 1.875792995095253,
      "epoch": 0.11544074168506593,
      "grad_norm": 8.272143363952637,
      "learning_rate": 7.445930126797723e-06,
      "loss": 0.5986,
      "mean_token_accuracy": 0.8243017882108689,
      "num_tokens": 44729226.0,
      "step": 37240
    },
    {
      "entropy": 1.8794658452272415,
      "epoch": 0.11547174081011563,
      "grad_norm": 8.320490837097168,
      "learning_rate": 7.444930579167699e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8193329513072968,
      "num_tokens": 44740879.0,
      "step": 37250
    },
    {
      "entropy": 1.8461437433958054,
      "epoch": 0.11550273993516533,
      "grad_norm": 7.777599334716797,
      "learning_rate": 7.443931433969854e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.8361176624894142,
      "num_tokens": 44753425.0,
      "step": 37260
    },
    {
      "entropy": 1.9898195713758469,
      "epoch": 0.11553373906021502,
      "grad_norm": 12.342960357666016,
      "learning_rate": 7.442932690934222e-06,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8236858904361725,
      "num_tokens": 44764029.0,
      "step": 37270
    },
    {
      "entropy": 1.850753267109394,
      "epoch": 0.11556473818526472,
      "grad_norm": 10.538897514343262,
      "learning_rate": 7.441934349791088e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.8256380185484886,
      "num_tokens": 44775779.0,
      "step": 37280
    },
    {
      "entropy": 1.7188558861613275,
      "epoch": 0.11559573731031442,
      "grad_norm": 4.442847728729248,
      "learning_rate": 7.440936410270987e-06,
      "loss": 0.3958,
      "mean_token_accuracy": 0.8477207824587822,
      "num_tokens": 44789900.0,
      "step": 37290
    },
    {
      "entropy": 1.854294976592064,
      "epoch": 0.11562673643536411,
      "grad_norm": 9.337111473083496,
      "learning_rate": 7.439938872104712e-06,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8315454825758934,
      "num_tokens": 44801588.0,
      "step": 37300
    },
    {
      "entropy": 1.7844763696193695,
      "epoch": 0.11565773556041381,
      "grad_norm": 8.090426445007324,
      "learning_rate": 7.438941735023301e-06,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8366570591926574,
      "num_tokens": 44813896.0,
      "step": 37310
    },
    {
      "entropy": 1.808639107644558,
      "epoch": 0.1156887346854635,
      "grad_norm": 4.602088451385498,
      "learning_rate": 7.437944998758055e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8374237030744552,
      "num_tokens": 44826692.0,
      "step": 37320
    },
    {
      "entropy": 1.9111674383282662,
      "epoch": 0.1157197338105132,
      "grad_norm": 9.890421867370605,
      "learning_rate": 7.436948663040519e-06,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8278648778796196,
      "num_tokens": 44838639.0,
      "step": 37330
    },
    {
      "entropy": 1.9008503317832948,
      "epoch": 0.1157507329355629,
      "grad_norm": 10.691512107849121,
      "learning_rate": 7.435952727602491e-06,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8256294175982475,
      "num_tokens": 44850123.0,
      "step": 37340
    },
    {
      "entropy": 1.9054113239049912,
      "epoch": 0.11578173206061258,
      "grad_norm": 9.177732467651367,
      "learning_rate": 7.434957192176021e-06,
      "loss": 0.5876,
      "mean_token_accuracy": 0.8166389659047126,
      "num_tokens": 44862112.0,
      "step": 37350
    },
    {
      "entropy": 1.7511047348380089,
      "epoch": 0.11581273118566228,
      "grad_norm": 9.275190353393555,
      "learning_rate": 7.4339620564934135e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8530907735228539,
      "num_tokens": 44875516.0,
      "step": 37360
    },
    {
      "entropy": 1.7619803249835968,
      "epoch": 0.11584373031071198,
      "grad_norm": 5.685603618621826,
      "learning_rate": 7.432967320287217e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.834882402420044,
      "num_tokens": 44889047.0,
      "step": 37370
    },
    {
      "entropy": 1.8839653521776198,
      "epoch": 0.11587472943576167,
      "grad_norm": 9.889922142028809,
      "learning_rate": 7.431972983290233e-06,
      "loss": 0.5696,
      "mean_token_accuracy": 0.820878466963768,
      "num_tokens": 44901808.0,
      "step": 37380
    },
    {
      "entropy": 1.9028914406895638,
      "epoch": 0.11590572856081137,
      "grad_norm": 9.42833423614502,
      "learning_rate": 7.430979045235518e-06,
      "loss": 0.5525,
      "mean_token_accuracy": 0.8282787010073662,
      "num_tokens": 44912993.0,
      "step": 37390
    },
    {
      "entropy": 1.9422337591648102,
      "epoch": 0.11593672768586107,
      "grad_norm": 10.66053295135498,
      "learning_rate": 7.429985505856372e-06,
      "loss": 0.6107,
      "mean_token_accuracy": 0.813813716173172,
      "num_tokens": 44924256.0,
      "step": 37400
    },
    {
      "entropy": 1.8587273508310318,
      "epoch": 0.11596772681091076,
      "grad_norm": 7.039865493774414,
      "learning_rate": 7.428992364886347e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8369274318218232,
      "num_tokens": 44935852.0,
      "step": 37410
    },
    {
      "entropy": 1.8579163044691085,
      "epoch": 0.11599872593596046,
      "grad_norm": 3.244483470916748,
      "learning_rate": 7.427999622059245e-06,
      "loss": 0.5921,
      "mean_token_accuracy": 0.8229958653450012,
      "num_tokens": 44948437.0,
      "step": 37420
    },
    {
      "entropy": 1.8869281217455864,
      "epoch": 0.11602972506101016,
      "grad_norm": 8.914671897888184,
      "learning_rate": 7.427007277109115e-06,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8428304120898247,
      "num_tokens": 44959542.0,
      "step": 37430
    },
    {
      "entropy": 1.88766710460186,
      "epoch": 0.11606072418605985,
      "grad_norm": 8.622590065002441,
      "learning_rate": 7.426015329770258e-06,
      "loss": 0.5914,
      "mean_token_accuracy": 0.8276036903262138,
      "num_tokens": 44970414.0,
      "step": 37440
    },
    {
      "entropy": 1.8154839545488357,
      "epoch": 0.11609172331110955,
      "grad_norm": 9.341155052185059,
      "learning_rate": 7.425023779777217e-06,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8210044398903846,
      "num_tokens": 44983628.0,
      "step": 37450
    },
    {
      "entropy": 1.8794740170240403,
      "epoch": 0.11612272243615925,
      "grad_norm": 8.086356163024902,
      "learning_rate": 7.424032626864791e-06,
      "loss": 0.6234,
      "mean_token_accuracy": 0.8183172300457955,
      "num_tokens": 44994394.0,
      "step": 37460
    },
    {
      "entropy": 1.8632339894771577,
      "epoch": 0.11615372156120893,
      "grad_norm": 7.764867782592773,
      "learning_rate": 7.423041870768022e-06,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8289612337946892,
      "num_tokens": 45005298.0,
      "step": 37470
    },
    {
      "entropy": 1.8341542795300483,
      "epoch": 0.11618472068625862,
      "grad_norm": 8.701674461364746,
      "learning_rate": 7.422051511222199e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.8334753260016441,
      "num_tokens": 45018299.0,
      "step": 37480
    },
    {
      "entropy": 1.915319959819317,
      "epoch": 0.11621571981130832,
      "grad_norm": 9.27292251586914,
      "learning_rate": 7.4210615479628625e-06,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8271935701370239,
      "num_tokens": 45029787.0,
      "step": 37490
    },
    {
      "entropy": 1.8192658558487893,
      "epoch": 0.11624671893635802,
      "grad_norm": 5.670734882354736,
      "learning_rate": 7.420071980725793e-06,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8343614637851715,
      "num_tokens": 45042647.0,
      "step": 37500
    },
    {
      "entropy": 1.9440273195505142,
      "epoch": 0.11627771806140771,
      "grad_norm": 10.800583839416504,
      "learning_rate": 7.419082809247022e-06,
      "loss": 0.6296,
      "mean_token_accuracy": 0.8177393227815628,
      "num_tokens": 45053626.0,
      "step": 37510
    },
    {
      "entropy": 1.8665451392531396,
      "epoch": 0.11630871718645741,
      "grad_norm": 11.033953666687012,
      "learning_rate": 7.41809403326283e-06,
      "loss": 0.5682,
      "mean_token_accuracy": 0.822401012480259,
      "num_tokens": 45064923.0,
      "step": 37520
    },
    {
      "entropy": 1.899769589304924,
      "epoch": 0.11633971631150711,
      "grad_norm": 8.622315406799316,
      "learning_rate": 7.417105652509737e-06,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8258567243814469,
      "num_tokens": 45076477.0,
      "step": 37530
    },
    {
      "entropy": 1.951041680574417,
      "epoch": 0.1163707154365568,
      "grad_norm": 12.631427764892578,
      "learning_rate": 7.4161176667245125e-06,
      "loss": 0.6739,
      "mean_token_accuracy": 0.8053397119045258,
      "num_tokens": 45087117.0,
      "step": 37540
    },
    {
      "entropy": 1.740734612941742,
      "epoch": 0.1164017145616065,
      "grad_norm": 4.5630879402160645,
      "learning_rate": 7.415130075644172e-06,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8315686360001564,
      "num_tokens": 45101656.0,
      "step": 37550
    },
    {
      "entropy": 1.8748736679553986,
      "epoch": 0.1164327136866562,
      "grad_norm": 8.166940689086914,
      "learning_rate": 7.414142879005973e-06,
      "loss": 0.5907,
      "mean_token_accuracy": 0.8158686429262161,
      "num_tokens": 45113754.0,
      "step": 37560
    },
    {
      "entropy": 1.8461486741900444,
      "epoch": 0.1164637128117059,
      "grad_norm": 10.343168258666992,
      "learning_rate": 7.4131560765474195e-06,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8233461812138557,
      "num_tokens": 45126280.0,
      "step": 37570
    },
    {
      "entropy": 1.8095831125974655,
      "epoch": 0.11649471193675559,
      "grad_norm": 9.836362838745117,
      "learning_rate": 7.4121696680062626e-06,
      "loss": 0.531,
      "mean_token_accuracy": 0.8389960706233979,
      "num_tokens": 45138548.0,
      "step": 37580
    },
    {
      "entropy": 1.8525902017951013,
      "epoch": 0.11652571106180529,
      "grad_norm": 2.9441442489624023,
      "learning_rate": 7.411183653120493e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8308546274900437,
      "num_tokens": 45150311.0,
      "step": 37590
    },
    {
      "entropy": 1.8115593805909156,
      "epoch": 0.11655671018685497,
      "grad_norm": 10.99833869934082,
      "learning_rate": 7.410198031628346e-06,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8222774997353554,
      "num_tokens": 45161956.0,
      "step": 37600
    },
    {
      "entropy": 1.7148562341928482,
      "epoch": 0.11658770931190467,
      "grad_norm": 9.413897514343262,
      "learning_rate": 7.409212803268305e-06,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8297837525606155,
      "num_tokens": 45175439.0,
      "step": 37610
    },
    {
      "entropy": 1.8933440566062927,
      "epoch": 0.11661870843695436,
      "grad_norm": 9.255085945129395,
      "learning_rate": 7.408227967779092e-06,
      "loss": 0.6599,
      "mean_token_accuracy": 0.8186682641506196,
      "num_tokens": 45187297.0,
      "step": 37620
    },
    {
      "entropy": 1.8789986670017242,
      "epoch": 0.11664970756200406,
      "grad_norm": 7.9891180992126465,
      "learning_rate": 7.407243524899674e-06,
      "loss": 0.5768,
      "mean_token_accuracy": 0.82134408056736,
      "num_tokens": 45198638.0,
      "step": 37630
    },
    {
      "entropy": 1.9211965501308441,
      "epoch": 0.11668070668705376,
      "grad_norm": 7.830356597900391,
      "learning_rate": 7.40625947436926e-06,
      "loss": 0.6321,
      "mean_token_accuracy": 0.813792322576046,
      "num_tokens": 45209458.0,
      "step": 37640
    },
    {
      "entropy": 1.853388948738575,
      "epoch": 0.11671170581210345,
      "grad_norm": 9.149991989135742,
      "learning_rate": 7.405275815927302e-06,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8271188646554947,
      "num_tokens": 45220490.0,
      "step": 37650
    },
    {
      "entropy": 1.8560687810182572,
      "epoch": 0.11674270493715315,
      "grad_norm": 10.992624282836914,
      "learning_rate": 7.404292549313496e-06,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8299055308103561,
      "num_tokens": 45231538.0,
      "step": 37660
    },
    {
      "entropy": 1.8339884772896766,
      "epoch": 0.11677370406220285,
      "grad_norm": 9.377613067626953,
      "learning_rate": 7.403309674267774e-06,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8145007371902466,
      "num_tokens": 45243273.0,
      "step": 37670
    },
    {
      "entropy": 1.8108773604035378,
      "epoch": 0.11680470318725254,
      "grad_norm": 8.911115646362305,
      "learning_rate": 7.402327190530316e-06,
      "loss": 0.5862,
      "mean_token_accuracy": 0.8245181262493133,
      "num_tokens": 45255185.0,
      "step": 37680
    },
    {
      "entropy": 1.7524414852261543,
      "epoch": 0.11683570231230224,
      "grad_norm": 8.754755973815918,
      "learning_rate": 7.40134509784154e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.8447420582175255,
      "num_tokens": 45268461.0,
      "step": 37690
    },
    {
      "entropy": 1.853088989853859,
      "epoch": 0.11686670143735194,
      "grad_norm": 8.089972496032715,
      "learning_rate": 7.400363395942107e-06,
      "loss": 0.5872,
      "mean_token_accuracy": 0.8264397710561753,
      "num_tokens": 45279403.0,
      "step": 37700
    },
    {
      "entropy": 1.8071050971746445,
      "epoch": 0.11689770056240163,
      "grad_norm": 4.270401954650879,
      "learning_rate": 7.399382084572917e-06,
      "loss": 0.5712,
      "mean_token_accuracy": 0.829101575911045,
      "num_tokens": 45292068.0,
      "step": 37710
    },
    {
      "entropy": 1.878752313554287,
      "epoch": 0.11692869968745132,
      "grad_norm": 5.220165252685547,
      "learning_rate": 7.39840116347511e-06,
      "loss": 0.5356,
      "mean_token_accuracy": 0.828040985763073,
      "num_tokens": 45303427.0,
      "step": 37720
    },
    {
      "entropy": 1.860033529996872,
      "epoch": 0.11695969881250101,
      "grad_norm": 8.818233489990234,
      "learning_rate": 7.397420632390068e-06,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8269825220108032,
      "num_tokens": 45314769.0,
      "step": 37730
    },
    {
      "entropy": 1.88105977922678,
      "epoch": 0.11699069793755071,
      "grad_norm": 8.774882316589355,
      "learning_rate": 7.396440491059412e-06,
      "loss": 0.5873,
      "mean_token_accuracy": 0.82732093334198,
      "num_tokens": 45326083.0,
      "step": 37740
    },
    {
      "entropy": 1.9004520326852798,
      "epoch": 0.1170216970626004,
      "grad_norm": 7.999516487121582,
      "learning_rate": 7.395460739225003e-06,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8362064957618713,
      "num_tokens": 45337413.0,
      "step": 37750
    },
    {
      "entropy": 1.858451707661152,
      "epoch": 0.1170526961876501,
      "grad_norm": 10.978395462036133,
      "learning_rate": 7.39448137662894e-06,
      "loss": 0.583,
      "mean_token_accuracy": 0.8314261004328728,
      "num_tokens": 45349411.0,
      "step": 37760
    },
    {
      "entropy": 1.7877505451440812,
      "epoch": 0.1170836953126998,
      "grad_norm": 8.580302238464355,
      "learning_rate": 7.393502403013563e-06,
      "loss": 0.539,
      "mean_token_accuracy": 0.8316345691680909,
      "num_tokens": 45362113.0,
      "step": 37770
    },
    {
      "entropy": 1.7756628662347793,
      "epoch": 0.1171146944377495,
      "grad_norm": 2.5867841243743896,
      "learning_rate": 7.3925238181214465e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8359642580151558,
      "num_tokens": 45375493.0,
      "step": 37780
    },
    {
      "entropy": 1.9025995954871178,
      "epoch": 0.11714569356279919,
      "grad_norm": 8.61524486541748,
      "learning_rate": 7.39154562169541e-06,
      "loss": 0.6269,
      "mean_token_accuracy": 0.8146417796611786,
      "num_tokens": 45386898.0,
      "step": 37790
    },
    {
      "entropy": 1.8143214404582977,
      "epoch": 0.11717669268784889,
      "grad_norm": 4.412146091461182,
      "learning_rate": 7.390567813478508e-06,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8260921567678452,
      "num_tokens": 45398636.0,
      "step": 37800
    },
    {
      "entropy": 1.8221836000680924,
      "epoch": 0.11720769181289858,
      "grad_norm": 9.057048797607422,
      "learning_rate": 7.38959039321403e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8306093484163284,
      "num_tokens": 45411214.0,
      "step": 37810
    },
    {
      "entropy": 1.7029403105378151,
      "epoch": 0.11723869093794828,
      "grad_norm": 3.986116886138916,
      "learning_rate": 7.388613360645508e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8498943701386452,
      "num_tokens": 45424321.0,
      "step": 37820
    },
    {
      "entropy": 1.8506017461419106,
      "epoch": 0.11726969006299798,
      "grad_norm": 11.799568176269531,
      "learning_rate": 7.387636715516706e-06,
      "loss": 0.5875,
      "mean_token_accuracy": 0.8202611520886421,
      "num_tokens": 45435300.0,
      "step": 37830
    },
    {
      "entropy": 1.7843811869621278,
      "epoch": 0.11730068918804766,
      "grad_norm": 9.326491355895996,
      "learning_rate": 7.38666045757163e-06,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8335476636886596,
      "num_tokens": 45448436.0,
      "step": 37840
    },
    {
      "entropy": 1.7802442833781242,
      "epoch": 0.11733168831309736,
      "grad_norm": 9.365975379943848,
      "learning_rate": 7.38568458655452e-06,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8346278235316277,
      "num_tokens": 45461982.0,
      "step": 37850
    },
    {
      "entropy": 1.774220162630081,
      "epoch": 0.11736268743814705,
      "grad_norm": 5.482513427734375,
      "learning_rate": 7.384709102209855e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8305883064866066,
      "num_tokens": 45474506.0,
      "step": 37860
    },
    {
      "entropy": 1.7526094153523446,
      "epoch": 0.11739368656319675,
      "grad_norm": 4.42122220993042,
      "learning_rate": 7.383734004282347e-06,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8322111248970032,
      "num_tokens": 45487584.0,
      "step": 37870
    },
    {
      "entropy": 1.7954672902822495,
      "epoch": 0.11742468568824645,
      "grad_norm": 9.287115097045898,
      "learning_rate": 7.382759292516944e-06,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8465468645095825,
      "num_tokens": 45499890.0,
      "step": 37880
    },
    {
      "entropy": 1.8831429094076158,
      "epoch": 0.11745568481329614,
      "grad_norm": 8.810885429382324,
      "learning_rate": 7.3817849666588325e-06,
      "loss": 0.5603,
      "mean_token_accuracy": 0.8357957974076271,
      "num_tokens": 45511132.0,
      "step": 37890
    },
    {
      "entropy": 1.8124101281166076,
      "epoch": 0.11748668393834584,
      "grad_norm": 8.837002754211426,
      "learning_rate": 7.3808110264534325e-06,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8318711072206497,
      "num_tokens": 45523253.0,
      "step": 37900
    },
    {
      "entropy": 1.8693040862679482,
      "epoch": 0.11751768306339554,
      "grad_norm": 7.558449745178223,
      "learning_rate": 7.379837471646401e-06,
      "loss": 0.6194,
      "mean_token_accuracy": 0.817514568567276,
      "num_tokens": 45534537.0,
      "step": 37910
    },
    {
      "entropy": 1.8211430594325067,
      "epoch": 0.11754868218844523,
      "grad_norm": 9.2406005859375,
      "learning_rate": 7.378864301983624e-06,
      "loss": 0.521,
      "mean_token_accuracy": 0.8348601713776589,
      "num_tokens": 45547222.0,
      "step": 37920
    },
    {
      "entropy": 1.788476151227951,
      "epoch": 0.11757968131349493,
      "grad_norm": 8.003228187561035,
      "learning_rate": 7.3778915172112294e-06,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8349843844771385,
      "num_tokens": 45558966.0,
      "step": 37930
    },
    {
      "entropy": 1.8203987568616866,
      "epoch": 0.11761068043854463,
      "grad_norm": 8.81309700012207,
      "learning_rate": 7.376919117075574e-06,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8295062810182572,
      "num_tokens": 45571715.0,
      "step": 37940
    },
    {
      "entropy": 1.8281734257936477,
      "epoch": 0.11764167956359432,
      "grad_norm": 9.268831253051758,
      "learning_rate": 7.375947101323252e-06,
      "loss": 0.5534,
      "mean_token_accuracy": 0.8323431298136711,
      "num_tokens": 45583585.0,
      "step": 37950
    },
    {
      "entropy": 1.8576543658971787,
      "epoch": 0.11767267868864402,
      "grad_norm": 9.705360412597656,
      "learning_rate": 7.37497546970109e-06,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8329675674438477,
      "num_tokens": 45595651.0,
      "step": 37960
    },
    {
      "entropy": 1.7942541763186455,
      "epoch": 0.1177036778136937,
      "grad_norm": 8.475988388061523,
      "learning_rate": 7.374004221956146e-06,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8307255417108536,
      "num_tokens": 45608005.0,
      "step": 37970
    },
    {
      "entropy": 1.7157593086361884,
      "epoch": 0.1177346769387434,
      "grad_norm": 10.300793647766113,
      "learning_rate": 7.373033357835715e-06,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8378100886940956,
      "num_tokens": 45621783.0,
      "step": 37980
    },
    {
      "entropy": 1.8599382385611534,
      "epoch": 0.1177656760637931,
      "grad_norm": 9.142414093017578,
      "learning_rate": 7.372062877087321e-06,
      "loss": 0.5541,
      "mean_token_accuracy": 0.8253407448530197,
      "num_tokens": 45632804.0,
      "step": 37990
    },
    {
      "entropy": 1.8708630874752998,
      "epoch": 0.11779667518884279,
      "grad_norm": 9.671748161315918,
      "learning_rate": 7.371092779458723e-06,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8188276901841164,
      "num_tokens": 45644415.0,
      "step": 38000
    },
    {
      "entropy": 1.8491891831159593,
      "epoch": 0.11782767431389249,
      "grad_norm": 8.546015739440918,
      "learning_rate": 7.3701230646979126e-06,
      "loss": 0.551,
      "mean_token_accuracy": 0.8339832827448845,
      "num_tokens": 45655755.0,
      "step": 38010
    },
    {
      "entropy": 1.8837746277451515,
      "epoch": 0.11785867343894219,
      "grad_norm": 10.270685195922852,
      "learning_rate": 7.369153732553109e-06,
      "loss": 0.575,
      "mean_token_accuracy": 0.8240560159087181,
      "num_tokens": 45666750.0,
      "step": 38020
    },
    {
      "entropy": 1.8285693258047104,
      "epoch": 0.11788967256399188,
      "grad_norm": 8.460458755493164,
      "learning_rate": 7.368184782772773e-06,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8289884492754936,
      "num_tokens": 45678928.0,
      "step": 38030
    },
    {
      "entropy": 1.903639057278633,
      "epoch": 0.11792067168904158,
      "grad_norm": 9.557387351989746,
      "learning_rate": 7.367216215105582e-06,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8232555508613586,
      "num_tokens": 45689650.0,
      "step": 38040
    },
    {
      "entropy": 1.840640588104725,
      "epoch": 0.11795167081409128,
      "grad_norm": 10.47343635559082,
      "learning_rate": 7.366248029300462e-06,
      "loss": 0.551,
      "mean_token_accuracy": 0.8262113586068154,
      "num_tokens": 45701186.0,
      "step": 38050
    },
    {
      "entropy": 1.6772912368178368,
      "epoch": 0.11798266993914097,
      "grad_norm": 5.621872425079346,
      "learning_rate": 7.365280225106553e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8406716391444207,
      "num_tokens": 45715025.0,
      "step": 38060
    },
    {
      "entropy": 1.674746471643448,
      "epoch": 0.11801366906419067,
      "grad_norm": 8.922220230102539,
      "learning_rate": 7.364312802273238e-06,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8508239656686782,
      "num_tokens": 45729652.0,
      "step": 38070
    },
    {
      "entropy": 1.8886992782354355,
      "epoch": 0.11804466818924036,
      "grad_norm": 9.37268352508545,
      "learning_rate": 7.3633457605501245e-06,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8332405045628548,
      "num_tokens": 45740799.0,
      "step": 38080
    },
    {
      "entropy": 1.7671217247843742,
      "epoch": 0.11807566731429005,
      "grad_norm": 3.940173387527466,
      "learning_rate": 7.362379099687053e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8395871177315712,
      "num_tokens": 45753706.0,
      "step": 38090
    },
    {
      "entropy": 1.83549507856369,
      "epoch": 0.11810666643933974,
      "grad_norm": 10.712289810180664,
      "learning_rate": 7.3614128194340895e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8255318030714989,
      "num_tokens": 45765958.0,
      "step": 38100
    },
    {
      "entropy": 1.775768581032753,
      "epoch": 0.11813766556438944,
      "grad_norm": 9.007753372192383,
      "learning_rate": 7.360446919541537e-06,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8286249712109566,
      "num_tokens": 45778973.0,
      "step": 38110
    },
    {
      "entropy": 1.8200758472085,
      "epoch": 0.11816866468943914,
      "grad_norm": 9.196110725402832,
      "learning_rate": 7.359481399759919e-06,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8254531383514404,
      "num_tokens": 45791157.0,
      "step": 38120
    },
    {
      "entropy": 1.8400531589984894,
      "epoch": 0.11819966381448883,
      "grad_norm": 8.93841552734375,
      "learning_rate": 7.358516259839993e-06,
      "loss": 0.5626,
      "mean_token_accuracy": 0.8274215057492256,
      "num_tokens": 45802910.0,
      "step": 38130
    },
    {
      "entropy": 1.8769369035959245,
      "epoch": 0.11823066293953853,
      "grad_norm": 8.658340454101562,
      "learning_rate": 7.3575514995327465e-06,
      "loss": 0.59,
      "mean_token_accuracy": 0.8198636502027512,
      "num_tokens": 45814938.0,
      "step": 38140
    },
    {
      "entropy": 1.8544082716107368,
      "epoch": 0.11826166206458823,
      "grad_norm": 10.661956787109375,
      "learning_rate": 7.3565871185893936e-06,
      "loss": 0.5401,
      "mean_token_accuracy": 0.833048839867115,
      "num_tokens": 45826509.0,
      "step": 38150
    },
    {
      "entropy": 1.8468938201665879,
      "epoch": 0.11829266118963792,
      "grad_norm": 8.67770004272461,
      "learning_rate": 7.3556231167613724e-06,
      "loss": 0.5409,
      "mean_token_accuracy": 0.829650467634201,
      "num_tokens": 45838046.0,
      "step": 38160
    },
    {
      "entropy": 1.8621126562356949,
      "epoch": 0.11832366031468762,
      "grad_norm": 6.933115005493164,
      "learning_rate": 7.354659493800356e-06,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8264819413423539,
      "num_tokens": 45850028.0,
      "step": 38170
    },
    {
      "entropy": 1.8032435089349748,
      "epoch": 0.11835465943973732,
      "grad_norm": 7.068675994873047,
      "learning_rate": 7.353696249458242e-06,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8219858318567276,
      "num_tokens": 45862980.0,
      "step": 38180
    },
    {
      "entropy": 1.8914562717080117,
      "epoch": 0.11838565856478701,
      "grad_norm": 8.82200813293457,
      "learning_rate": 7.352733383487156e-06,
      "loss": 0.6047,
      "mean_token_accuracy": 0.8125669702887535,
      "num_tokens": 45875364.0,
      "step": 38190
    },
    {
      "entropy": 1.77557475566864,
      "epoch": 0.11841665768983671,
      "grad_norm": 4.266529083251953,
      "learning_rate": 7.3517708956394485e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.843745194375515,
      "num_tokens": 45887863.0,
      "step": 38200
    },
    {
      "entropy": 1.8870860502123832,
      "epoch": 0.1184476568148864,
      "grad_norm": 9.665167808532715,
      "learning_rate": 7.3508087856677e-06,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8245103910565377,
      "num_tokens": 45899658.0,
      "step": 38210
    },
    {
      "entropy": 1.7931729927659035,
      "epoch": 0.11847865593993609,
      "grad_norm": 10.355236053466797,
      "learning_rate": 7.3498470533247175e-06,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8335598543286323,
      "num_tokens": 45912392.0,
      "step": 38220
    },
    {
      "entropy": 1.739191673696041,
      "epoch": 0.11850965506498579,
      "grad_norm": 4.354625225067139,
      "learning_rate": 7.34888569836353e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8427866518497467,
      "num_tokens": 45925460.0,
      "step": 38230
    },
    {
      "entropy": 1.832131990790367,
      "epoch": 0.11854065419003548,
      "grad_norm": 11.056679725646973,
      "learning_rate": 7.347924720537399e-06,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8355686992406846,
      "num_tokens": 45937998.0,
      "step": 38240
    },
    {
      "entropy": 1.8358628578484057,
      "epoch": 0.11857165331508518,
      "grad_norm": 2.7926924228668213,
      "learning_rate": 7.346964119599805e-06,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8267253056168556,
      "num_tokens": 45950095.0,
      "step": 38250
    },
    {
      "entropy": 1.8598995104432106,
      "epoch": 0.11860265244013488,
      "grad_norm": 9.030755996704102,
      "learning_rate": 7.346003895304459e-06,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8271134316921234,
      "num_tokens": 45962187.0,
      "step": 38260
    },
    {
      "entropy": 1.8761465281248093,
      "epoch": 0.11863365156518457,
      "grad_norm": 4.900582313537598,
      "learning_rate": 7.345044047405296e-06,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8255822539329529,
      "num_tokens": 45974742.0,
      "step": 38270
    },
    {
      "entropy": 1.8882175594568253,
      "epoch": 0.11866465069023427,
      "grad_norm": 9.869363784790039,
      "learning_rate": 7.344084575656477e-06,
      "loss": 0.578,
      "mean_token_accuracy": 0.818389254808426,
      "num_tokens": 45986513.0,
      "step": 38280
    },
    {
      "entropy": 1.9118270069360732,
      "epoch": 0.11869564981528397,
      "grad_norm": 10.015856742858887,
      "learning_rate": 7.343125479812384e-06,
      "loss": 0.626,
      "mean_token_accuracy": 0.8135067865252494,
      "num_tokens": 45998695.0,
      "step": 38290
    },
    {
      "entropy": 1.7809939235448837,
      "epoch": 0.11872664894033366,
      "grad_norm": 2.5885379314422607,
      "learning_rate": 7.342166759627627e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8359186798334122,
      "num_tokens": 46011982.0,
      "step": 38300
    },
    {
      "entropy": 1.90637346804142,
      "epoch": 0.11875764806538336,
      "grad_norm": 9.184483528137207,
      "learning_rate": 7.341208414857039e-06,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8383510798215866,
      "num_tokens": 46023222.0,
      "step": 38310
    },
    {
      "entropy": 1.9249227464199066,
      "epoch": 0.11878864719043306,
      "grad_norm": 8.180423736572266,
      "learning_rate": 7.340250445255678e-06,
      "loss": 0.591,
      "mean_token_accuracy": 0.822523207962513,
      "num_tokens": 46033818.0,
      "step": 38320
    },
    {
      "entropy": 1.8500513106584549,
      "epoch": 0.11881964631548275,
      "grad_norm": 8.05856990814209,
      "learning_rate": 7.3392928505788245e-06,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8258534908294678,
      "num_tokens": 46045985.0,
      "step": 38330
    },
    {
      "entropy": 1.891994585096836,
      "epoch": 0.11885064544053243,
      "grad_norm": 10.122121810913086,
      "learning_rate": 7.338335630581982e-06,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8260172605514526,
      "num_tokens": 46057552.0,
      "step": 38340
    },
    {
      "entropy": 1.9085305988788606,
      "epoch": 0.11888164456558213,
      "grad_norm": 8.372586250305176,
      "learning_rate": 7.3373787850208775e-06,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8344064444303513,
      "num_tokens": 46068444.0,
      "step": 38350
    },
    {
      "entropy": 1.8459767132997513,
      "epoch": 0.11891264369063183,
      "grad_norm": 9.715620040893555,
      "learning_rate": 7.336422313651464e-06,
      "loss": 0.5169,
      "mean_token_accuracy": 0.83453588783741,
      "num_tokens": 46080267.0,
      "step": 38360
    },
    {
      "entropy": 1.966032502055168,
      "epoch": 0.11894364281568152,
      "grad_norm": 8.732338905334473,
      "learning_rate": 7.33546621622991e-06,
      "loss": 0.6502,
      "mean_token_accuracy": 0.8121734812855721,
      "num_tokens": 46091339.0,
      "step": 38370
    },
    {
      "entropy": 1.8556820943951606,
      "epoch": 0.11897464194073122,
      "grad_norm": 10.176673889160156,
      "learning_rate": 7.334510492512614e-06,
      "loss": 0.5158,
      "mean_token_accuracy": 0.837830537557602,
      "num_tokens": 46102953.0,
      "step": 38380
    },
    {
      "entropy": 1.883063942193985,
      "epoch": 0.11900564106578092,
      "grad_norm": 7.522932052612305,
      "learning_rate": 7.3335551422561916e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8345891013741493,
      "num_tokens": 46114625.0,
      "step": 38390
    },
    {
      "entropy": 1.8097529754042625,
      "epoch": 0.11903664019083061,
      "grad_norm": 3.6872365474700928,
      "learning_rate": 7.3326001652174846e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.8357144206762314,
      "num_tokens": 46126642.0,
      "step": 38400
    },
    {
      "entropy": 1.8187772005796432,
      "epoch": 0.11906763931588031,
      "grad_norm": 11.781440734863281,
      "learning_rate": 7.331645561153551e-06,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8383165881037712,
      "num_tokens": 46138142.0,
      "step": 38410
    },
    {
      "entropy": 1.8450339302420615,
      "epoch": 0.11909863844093001,
      "grad_norm": 12.367788314819336,
      "learning_rate": 7.330691329821676e-06,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8337818294763565,
      "num_tokens": 46149774.0,
      "step": 38420
    },
    {
      "entropy": 1.8153291314840316,
      "epoch": 0.1191296375659797,
      "grad_norm": 8.434037208557129,
      "learning_rate": 7.329737470979359e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8328280210494995,
      "num_tokens": 46162425.0,
      "step": 38430
    },
    {
      "entropy": 1.8206452041864396,
      "epoch": 0.1191606366910294,
      "grad_norm": 8.897926330566406,
      "learning_rate": 7.328783984384326e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8331076070666313,
      "num_tokens": 46175025.0,
      "step": 38440
    },
    {
      "entropy": 1.8406914666295051,
      "epoch": 0.1191916358160791,
      "grad_norm": 9.005738258361816,
      "learning_rate": 7.327830869794524e-06,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8288857981562614,
      "num_tokens": 46187094.0,
      "step": 38450
    },
    {
      "entropy": 1.8507864832878114,
      "epoch": 0.11922263494112878,
      "grad_norm": 5.744321823120117,
      "learning_rate": 7.326878126968114e-06,
      "loss": 0.5862,
      "mean_token_accuracy": 0.815381346642971,
      "num_tokens": 46199554.0,
      "step": 38460
    },
    {
      "entropy": 1.8325456768274306,
      "epoch": 0.11925363406617848,
      "grad_norm": 10.185477256774902,
      "learning_rate": 7.325925755663483e-06,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8304981887340546,
      "num_tokens": 46211540.0,
      "step": 38470
    },
    {
      "entropy": 1.8854450479149818,
      "epoch": 0.11928463319122817,
      "grad_norm": 8.560661315917969,
      "learning_rate": 7.324973755639235e-06,
      "loss": 0.5288,
      "mean_token_accuracy": 0.823890820145607,
      "num_tokens": 46223017.0,
      "step": 38480
    },
    {
      "entropy": 1.892309932410717,
      "epoch": 0.11931563231627787,
      "grad_norm": 9.574992179870605,
      "learning_rate": 7.324022126654195e-06,
      "loss": 0.5751,
      "mean_token_accuracy": 0.823742826282978,
      "num_tokens": 46235073.0,
      "step": 38490
    },
    {
      "entropy": 1.814597336947918,
      "epoch": 0.11934663144132757,
      "grad_norm": 8.49984359741211,
      "learning_rate": 7.323070868467407e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8360254690051079,
      "num_tokens": 46247594.0,
      "step": 38500
    },
    {
      "entropy": 1.8429415673017502,
      "epoch": 0.11937763056637726,
      "grad_norm": 8.082172393798828,
      "learning_rate": 7.322119980838131e-06,
      "loss": 0.5642,
      "mean_token_accuracy": 0.8176757171750069,
      "num_tokens": 46259339.0,
      "step": 38510
    },
    {
      "entropy": 1.8391290426254272,
      "epoch": 0.11940862969142696,
      "grad_norm": 8.323386192321777,
      "learning_rate": 7.32116946352585e-06,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8202460646629334,
      "num_tokens": 46271303.0,
      "step": 38520
    },
    {
      "entropy": 1.8703968793153762,
      "epoch": 0.11943962881647666,
      "grad_norm": 4.209524631500244,
      "learning_rate": 7.3202193162902655e-06,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8387380838394165,
      "num_tokens": 46282368.0,
      "step": 38530
    },
    {
      "entropy": 1.8230261772871017,
      "epoch": 0.11947062794152635,
      "grad_norm": 9.72610855102539,
      "learning_rate": 7.319269538891293e-06,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8248476162552834,
      "num_tokens": 46294921.0,
      "step": 38540
    },
    {
      "entropy": 1.827505610883236,
      "epoch": 0.11950162706657605,
      "grad_norm": 8.594587326049805,
      "learning_rate": 7.318320131089069e-06,
      "loss": 0.5674,
      "mean_token_accuracy": 0.826340651512146,
      "num_tokens": 46307519.0,
      "step": 38550
    },
    {
      "entropy": 1.8293501362204552,
      "epoch": 0.11953262619162575,
      "grad_norm": 7.893844127655029,
      "learning_rate": 7.31737109264395e-06,
      "loss": 0.4868,
      "mean_token_accuracy": 0.8477142170071602,
      "num_tokens": 46319880.0,
      "step": 38560
    },
    {
      "entropy": 1.967301172018051,
      "epoch": 0.11956362531667544,
      "grad_norm": 10.387765884399414,
      "learning_rate": 7.316422423316503e-06,
      "loss": 0.6318,
      "mean_token_accuracy": 0.8172219157218933,
      "num_tokens": 46331125.0,
      "step": 38570
    },
    {
      "entropy": 1.9015961229801177,
      "epoch": 0.11959462444172514,
      "grad_norm": 8.525230407714844,
      "learning_rate": 7.315474122867519e-06,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8258678615093231,
      "num_tokens": 46343547.0,
      "step": 38580
    },
    {
      "entropy": 1.877706202864647,
      "epoch": 0.11962562356677482,
      "grad_norm": 8.298994064331055,
      "learning_rate": 7.314526191058002e-06,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8354897379875184,
      "num_tokens": 46354401.0,
      "step": 38590
    },
    {
      "entropy": 1.7901566669344902,
      "epoch": 0.11965662269182452,
      "grad_norm": 8.73811149597168,
      "learning_rate": 7.313578627649177e-06,
      "loss": 0.4969,
      "mean_token_accuracy": 0.8303673461079597,
      "num_tokens": 46367136.0,
      "step": 38600
    },
    {
      "entropy": 1.855716420710087,
      "epoch": 0.11968762181687422,
      "grad_norm": 8.627557754516602,
      "learning_rate": 7.31263143240248e-06,
      "loss": 0.552,
      "mean_token_accuracy": 0.8318748503923417,
      "num_tokens": 46378695.0,
      "step": 38610
    },
    {
      "entropy": 1.8308327570557594,
      "epoch": 0.11971862094192391,
      "grad_norm": 4.684299468994141,
      "learning_rate": 7.3116846050795675e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8276481464505195,
      "num_tokens": 46390539.0,
      "step": 38620
    },
    {
      "entropy": 1.7808781787753105,
      "epoch": 0.11974962006697361,
      "grad_norm": 9.359618186950684,
      "learning_rate": 7.3107381454423095e-06,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8304883912205696,
      "num_tokens": 46403361.0,
      "step": 38630
    },
    {
      "entropy": 1.832536792755127,
      "epoch": 0.1197806191920233,
      "grad_norm": 9.308186531066895,
      "learning_rate": 7.3097920532527925e-06,
      "loss": 0.5477,
      "mean_token_accuracy": 0.8254012614488602,
      "num_tokens": 46415320.0,
      "step": 38640
    },
    {
      "entropy": 1.875816921889782,
      "epoch": 0.119811618317073,
      "grad_norm": 6.979348182678223,
      "learning_rate": 7.3088463282733195e-06,
      "loss": 0.6144,
      "mean_token_accuracy": 0.8188940063118935,
      "num_tokens": 46426785.0,
      "step": 38650
    },
    {
      "entropy": 1.896639221906662,
      "epoch": 0.1198426174421227,
      "grad_norm": 8.80232048034668,
      "learning_rate": 7.307900970266406e-06,
      "loss": 0.5687,
      "mean_token_accuracy": 0.831016306579113,
      "num_tokens": 46437545.0,
      "step": 38660
    },
    {
      "entropy": 1.8298520892858505,
      "epoch": 0.1198736165671724,
      "grad_norm": 6.179274559020996,
      "learning_rate": 7.3069559789947875e-06,
      "loss": 0.6193,
      "mean_token_accuracy": 0.8189263239502906,
      "num_tokens": 46449253.0,
      "step": 38670
    },
    {
      "entropy": 1.8082061618566514,
      "epoch": 0.11990461569222209,
      "grad_norm": 7.518819808959961,
      "learning_rate": 7.306011354221407e-06,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8381061047315598,
      "num_tokens": 46461547.0,
      "step": 38680
    },
    {
      "entropy": 1.8741240479052066,
      "epoch": 0.11993561481727179,
      "grad_norm": 8.276564598083496,
      "learning_rate": 7.305067095709427e-06,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8233344674110412,
      "num_tokens": 46473612.0,
      "step": 38690
    },
    {
      "entropy": 1.82129565179348,
      "epoch": 0.11996661394232148,
      "grad_norm": 3.840449571609497,
      "learning_rate": 7.3041232032222255e-06,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8374928280711174,
      "num_tokens": 46485577.0,
      "step": 38700
    },
    {
      "entropy": 1.7987656205892564,
      "epoch": 0.11999761306737117,
      "grad_norm": 9.687265396118164,
      "learning_rate": 7.303179676523391e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8323082059621811,
      "num_tokens": 46497845.0,
      "step": 38710
    },
    {
      "entropy": 1.8387595310807228,
      "epoch": 0.12002861219242086,
      "grad_norm": 7.756543159484863,
      "learning_rate": 7.302236515376725e-06,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8210761576890946,
      "num_tokens": 46510088.0,
      "step": 38720
    },
    {
      "entropy": 1.7564123407006265,
      "epoch": 0.12005961131747056,
      "grad_norm": 8.537627220153809,
      "learning_rate": 7.301293719546245e-06,
      "loss": 0.5524,
      "mean_token_accuracy": 0.831873744726181,
      "num_tokens": 46522634.0,
      "step": 38730
    },
    {
      "entropy": 1.8440282315015792,
      "epoch": 0.12009061044252026,
      "grad_norm": 7.22578763961792,
      "learning_rate": 7.300351288796182e-06,
      "loss": 0.575,
      "mean_token_accuracy": 0.8209284529089927,
      "num_tokens": 46534334.0,
      "step": 38740
    },
    {
      "entropy": 1.8955597847700119,
      "epoch": 0.12012160956756995,
      "grad_norm": 8.206884384155273,
      "learning_rate": 7.299409222890979e-06,
      "loss": 0.6,
      "mean_token_accuracy": 0.8230985984206199,
      "num_tokens": 46546015.0,
      "step": 38750
    },
    {
      "entropy": 1.831144355237484,
      "epoch": 0.12015260869261965,
      "grad_norm": 9.138228416442871,
      "learning_rate": 7.298467521595293e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8403099969029426,
      "num_tokens": 46558715.0,
      "step": 38760
    },
    {
      "entropy": 1.7825349926948548,
      "epoch": 0.12018360781766935,
      "grad_norm": 8.191915512084961,
      "learning_rate": 7.297526184673988e-06,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8280392870306968,
      "num_tokens": 46572867.0,
      "step": 38770
    },
    {
      "entropy": 1.8217300280928612,
      "epoch": 0.12021460694271904,
      "grad_norm": 8.535613059997559,
      "learning_rate": 7.296585211892146e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8408209979534149,
      "num_tokens": 46585902.0,
      "step": 38780
    },
    {
      "entropy": 1.8220679610967636,
      "epoch": 0.12024560606776874,
      "grad_norm": 4.572232246398926,
      "learning_rate": 7.295644603015063e-06,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8344119265675545,
      "num_tokens": 46599076.0,
      "step": 38790
    },
    {
      "entropy": 1.8718759939074516,
      "epoch": 0.12027660519281844,
      "grad_norm": 9.355148315429688,
      "learning_rate": 7.294704357808237e-06,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8285117372870445,
      "num_tokens": 46611131.0,
      "step": 38800
    },
    {
      "entropy": 1.901232734322548,
      "epoch": 0.12030760431786813,
      "grad_norm": 10.37480354309082,
      "learning_rate": 7.2937644760373896e-06,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8172040551900863,
      "num_tokens": 46623342.0,
      "step": 38810
    },
    {
      "entropy": 1.8803485922515393,
      "epoch": 0.12033860344291783,
      "grad_norm": 9.32126235961914,
      "learning_rate": 7.292824957468444e-06,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8315271034836769,
      "num_tokens": 46634924.0,
      "step": 38820
    },
    {
      "entropy": 1.9120069742202759,
      "epoch": 0.12036960256796751,
      "grad_norm": 10.770748138427734,
      "learning_rate": 7.29188580186754e-06,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8313893929123879,
      "num_tokens": 46647007.0,
      "step": 38830
    },
    {
      "entropy": 1.7526630774140357,
      "epoch": 0.12040060169301721,
      "grad_norm": 4.222184181213379,
      "learning_rate": 7.290947009001024e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8405937299132347,
      "num_tokens": 46660402.0,
      "step": 38840
    },
    {
      "entropy": 1.8250176429748535,
      "epoch": 0.1204316008180669,
      "grad_norm": 8.69330883026123,
      "learning_rate": 7.290008578635457e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.82737468034029,
      "num_tokens": 46673041.0,
      "step": 38850
    },
    {
      "entropy": 1.8118784487247468,
      "epoch": 0.1204625999431166,
      "grad_norm": 9.943561553955078,
      "learning_rate": 7.289070510537608e-06,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8418055370450019,
      "num_tokens": 46686648.0,
      "step": 38860
    },
    {
      "entropy": 1.9078863650560378,
      "epoch": 0.1204935990681663,
      "grad_norm": 9.226347923278809,
      "learning_rate": 7.288132804474457e-06,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8267140224575996,
      "num_tokens": 46698591.0,
      "step": 38870
    },
    {
      "entropy": 1.8343891605734826,
      "epoch": 0.120524598193216,
      "grad_norm": 5.343838691711426,
      "learning_rate": 7.28719546021319e-06,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8361982107162476,
      "num_tokens": 46711162.0,
      "step": 38880
    },
    {
      "entropy": 1.8952967941761016,
      "epoch": 0.12055559731826569,
      "grad_norm": 7.90724515914917,
      "learning_rate": 7.286258477521211e-06,
      "loss": 0.5974,
      "mean_token_accuracy": 0.8240769699215889,
      "num_tokens": 46723517.0,
      "step": 38890
    },
    {
      "entropy": 1.8919255122542382,
      "epoch": 0.12058659644331539,
      "grad_norm": 8.763474464416504,
      "learning_rate": 7.285321856166125e-06,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8284026741981506,
      "num_tokens": 46734522.0,
      "step": 38900
    },
    {
      "entropy": 1.9270798504352569,
      "epoch": 0.12061759556836509,
      "grad_norm": 8.891312599182129,
      "learning_rate": 7.284385595915748e-06,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8333567604422569,
      "num_tokens": 46745723.0,
      "step": 38910
    },
    {
      "entropy": 1.904827606678009,
      "epoch": 0.12064859469341478,
      "grad_norm": 9.994085311889648,
      "learning_rate": 7.283449696538109e-06,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8297219768166542,
      "num_tokens": 46757162.0,
      "step": 38920
    },
    {
      "entropy": 1.9497202724218368,
      "epoch": 0.12067959381846448,
      "grad_norm": 10.245466232299805,
      "learning_rate": 7.282514157801443e-06,
      "loss": 0.6198,
      "mean_token_accuracy": 0.8121256932616234,
      "num_tokens": 46768342.0,
      "step": 38930
    },
    {
      "entropy": 1.9423091277480125,
      "epoch": 0.12071059294351418,
      "grad_norm": 8.347733497619629,
      "learning_rate": 7.2815789794741885e-06,
      "loss": 0.5971,
      "mean_token_accuracy": 0.8201627746224404,
      "num_tokens": 46779722.0,
      "step": 38940
    },
    {
      "entropy": 1.9126493602991104,
      "epoch": 0.12074159206856387,
      "grad_norm": 8.225603103637695,
      "learning_rate": 7.280644161324999e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8218429237604141,
      "num_tokens": 46791464.0,
      "step": 38950
    },
    {
      "entropy": 1.8790960252285003,
      "epoch": 0.12077259119361355,
      "grad_norm": 8.749955177307129,
      "learning_rate": 7.2797097031227335e-06,
      "loss": 0.5397,
      "mean_token_accuracy": 0.8350474625825882,
      "num_tokens": 46804115.0,
      "step": 38960
    },
    {
      "entropy": 1.8715863823890686,
      "epoch": 0.12080359031866325,
      "grad_norm": 11.107318878173828,
      "learning_rate": 7.278775604636458e-06,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8246800765395165,
      "num_tokens": 46815723.0,
      "step": 38970
    },
    {
      "entropy": 1.8465174466371537,
      "epoch": 0.12083458944371295,
      "grad_norm": 6.929959774017334,
      "learning_rate": 7.277841865635446e-06,
      "loss": 0.5764,
      "mean_token_accuracy": 0.8237011566758156,
      "num_tokens": 46828326.0,
      "step": 38980
    },
    {
      "entropy": 1.9157025456428527,
      "epoch": 0.12086558856876264,
      "grad_norm": 7.555674076080322,
      "learning_rate": 7.27690848588918e-06,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8201544851064682,
      "num_tokens": 46840929.0,
      "step": 38990
    },
    {
      "entropy": 1.855295367538929,
      "epoch": 0.12089658769381234,
      "grad_norm": 9.253195762634277,
      "learning_rate": 7.275975465167346e-06,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8400891706347465,
      "num_tokens": 46853363.0,
      "step": 39000
    },
    {
      "entropy": 1.9134544879198074,
      "epoch": 0.12092758681886204,
      "grad_norm": 9.56458568572998,
      "learning_rate": 7.27504280323984e-06,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8285094171762466,
      "num_tokens": 46865204.0,
      "step": 39010
    },
    {
      "entropy": 1.8941469311714172,
      "epoch": 0.12095858594391173,
      "grad_norm": 8.927152633666992,
      "learning_rate": 7.274110499876761e-06,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8449266985058784,
      "num_tokens": 46876368.0,
      "step": 39020
    },
    {
      "entropy": 1.8193779736757278,
      "epoch": 0.12098958506896143,
      "grad_norm": 3.3484020233154297,
      "learning_rate": 7.273178554848418e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8299607783555984,
      "num_tokens": 46889571.0,
      "step": 39030
    },
    {
      "entropy": 1.8515197828412056,
      "epoch": 0.12102058419401113,
      "grad_norm": 11.17811393737793,
      "learning_rate": 7.272246967925323e-06,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8318181455135345,
      "num_tokens": 46901011.0,
      "step": 39040
    },
    {
      "entropy": 1.9295257300138473,
      "epoch": 0.12105158331906082,
      "grad_norm": 9.795385360717773,
      "learning_rate": 7.271315738878194e-06,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8381862804293633,
      "num_tokens": 46911801.0,
      "step": 39050
    },
    {
      "entropy": 1.955726158618927,
      "epoch": 0.12108258244411052,
      "grad_norm": 11.714072227478027,
      "learning_rate": 7.270384867477956e-06,
      "loss": 0.5593,
      "mean_token_accuracy": 0.8203748896718025,
      "num_tokens": 46923433.0,
      "step": 39060
    },
    {
      "entropy": 1.9114267766475677,
      "epoch": 0.12111358156916022,
      "grad_norm": 9.219902992248535,
      "learning_rate": 7.269454353495741e-06,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8343056261539459,
      "num_tokens": 46934777.0,
      "step": 39070
    },
    {
      "entropy": 1.9058862313628198,
      "epoch": 0.1211445806942099,
      "grad_norm": 9.145793914794922,
      "learning_rate": 7.26852419670288e-06,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8218263626098633,
      "num_tokens": 46947035.0,
      "step": 39080
    },
    {
      "entropy": 1.8588959291577338,
      "epoch": 0.1211755798192596,
      "grad_norm": 9.202445983886719,
      "learning_rate": 7.267594396870911e-06,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8362350627779961,
      "num_tokens": 46959557.0,
      "step": 39090
    },
    {
      "entropy": 1.928936019539833,
      "epoch": 0.12120657894430929,
      "grad_norm": 11.063957214355469,
      "learning_rate": 7.2666649537715814e-06,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8108921408653259,
      "num_tokens": 46971699.0,
      "step": 39100
    },
    {
      "entropy": 1.9692992329597474,
      "epoch": 0.12123757806935899,
      "grad_norm": 11.699572563171387,
      "learning_rate": 7.265735867176837e-06,
      "loss": 0.609,
      "mean_token_accuracy": 0.8138938039541245,
      "num_tokens": 46983377.0,
      "step": 39110
    },
    {
      "entropy": 1.942272038757801,
      "epoch": 0.12126857719440869,
      "grad_norm": 8.434643745422363,
      "learning_rate": 7.264807136858832e-06,
      "loss": 0.6463,
      "mean_token_accuracy": 0.8089325189590454,
      "num_tokens": 46996376.0,
      "step": 39120
    },
    {
      "entropy": 1.8001896619796753,
      "epoch": 0.12129957631945838,
      "grad_norm": 3.925816059112549,
      "learning_rate": 7.2638787625899185e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8355321779847145,
      "num_tokens": 47010314.0,
      "step": 39130
    },
    {
      "entropy": 1.9362492710351944,
      "epoch": 0.12133057544450808,
      "grad_norm": 7.411516189575195,
      "learning_rate": 7.26295074414266e-06,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8266261234879494,
      "num_tokens": 47022279.0,
      "step": 39140
    },
    {
      "entropy": 1.892134927213192,
      "epoch": 0.12136157456955778,
      "grad_norm": 7.244060039520264,
      "learning_rate": 7.262023081289816e-06,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8465871244668961,
      "num_tokens": 47034156.0,
      "step": 39150
    },
    {
      "entropy": 1.9123440355062484,
      "epoch": 0.12139257369460747,
      "grad_norm": 8.467802047729492,
      "learning_rate": 7.261095773804354e-06,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8312965899705886,
      "num_tokens": 47045539.0,
      "step": 39160
    },
    {
      "entropy": 1.909021759033203,
      "epoch": 0.12142357281965717,
      "grad_norm": 8.491313934326172,
      "learning_rate": 7.260168821459445e-06,
      "loss": 0.5967,
      "mean_token_accuracy": 0.809444610774517,
      "num_tokens": 47057991.0,
      "step": 39170
    },
    {
      "entropy": 1.8988661482930183,
      "epoch": 0.12145457194470687,
      "grad_norm": 8.673091888427734,
      "learning_rate": 7.259242224028456e-06,
      "loss": 0.5633,
      "mean_token_accuracy": 0.824192288517952,
      "num_tokens": 47069487.0,
      "step": 39180
    },
    {
      "entropy": 1.8936348468065263,
      "epoch": 0.12148557106975656,
      "grad_norm": 7.909213066101074,
      "learning_rate": 7.258315981284962e-06,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8279350861907006,
      "num_tokens": 47081241.0,
      "step": 39190
    },
    {
      "entropy": 1.887278787791729,
      "epoch": 0.12151657019480624,
      "grad_norm": 9.825845718383789,
      "learning_rate": 7.2573900930027416e-06,
      "loss": 0.5607,
      "mean_token_accuracy": 0.82880117893219,
      "num_tokens": 47093264.0,
      "step": 39200
    },
    {
      "entropy": 1.8978556409478187,
      "epoch": 0.12154756931985594,
      "grad_norm": 9.518089294433594,
      "learning_rate": 7.25646455895577e-06,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8237947776913643,
      "num_tokens": 47105376.0,
      "step": 39210
    },
    {
      "entropy": 1.8898191466927527,
      "epoch": 0.12157856844490564,
      "grad_norm": 3.807635545730591,
      "learning_rate": 7.255539378918229e-06,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8275874778628349,
      "num_tokens": 47117645.0,
      "step": 39220
    },
    {
      "entropy": 1.8876211807131766,
      "epoch": 0.12160956756995533,
      "grad_norm": 5.313508033752441,
      "learning_rate": 7.254614552664499e-06,
      "loss": 0.5445,
      "mean_token_accuracy": 0.815770597755909,
      "num_tokens": 47129805.0,
      "step": 39230
    },
    {
      "entropy": 1.89053615629673,
      "epoch": 0.12164056669500503,
      "grad_norm": 9.875513076782227,
      "learning_rate": 7.253690079969162e-06,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8238019242882728,
      "num_tokens": 47142055.0,
      "step": 39240
    },
    {
      "entropy": 1.900741669535637,
      "epoch": 0.12167156582005473,
      "grad_norm": 4.167596340179443,
      "learning_rate": 7.252765960607002e-06,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8360105812549591,
      "num_tokens": 47154609.0,
      "step": 39250
    },
    {
      "entropy": 1.8273233488202094,
      "epoch": 0.12170256494510442,
      "grad_norm": 9.690666198730469,
      "learning_rate": 7.251842194353004e-06,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8299174129962921,
      "num_tokens": 47167381.0,
      "step": 39260
    },
    {
      "entropy": 1.9333147034049034,
      "epoch": 0.12173356407015412,
      "grad_norm": 3.9693877696990967,
      "learning_rate": 7.25091878098235e-06,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8288700088858605,
      "num_tokens": 47179138.0,
      "step": 39270
    },
    {
      "entropy": 1.8680218786001206,
      "epoch": 0.12176456319520382,
      "grad_norm": 3.812345504760742,
      "learning_rate": 7.249995720270428e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8331614285707474,
      "num_tokens": 47192026.0,
      "step": 39280
    },
    {
      "entropy": 1.8831014022231103,
      "epoch": 0.12179556232025351,
      "grad_norm": 9.065089225769043,
      "learning_rate": 7.249073011992822e-06,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8253825202584266,
      "num_tokens": 47204601.0,
      "step": 39290
    },
    {
      "entropy": 1.878520594537258,
      "epoch": 0.12182656144530321,
      "grad_norm": 8.888299942016602,
      "learning_rate": 7.248150655925318e-06,
      "loss": 0.4776,
      "mean_token_accuracy": 0.8410759374499321,
      "num_tokens": 47217084.0,
      "step": 39300
    },
    {
      "entropy": 1.9088293090462685,
      "epoch": 0.12185756057035291,
      "grad_norm": 8.316096305847168,
      "learning_rate": 7.247228651843902e-06,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8348750114440918,
      "num_tokens": 47229268.0,
      "step": 39310
    },
    {
      "entropy": 1.9051673859357834,
      "epoch": 0.1218885596954026,
      "grad_norm": 8.587472915649414,
      "learning_rate": 7.246306999524752e-06,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8288885727524757,
      "num_tokens": 47241225.0,
      "step": 39320
    },
    {
      "entropy": 1.970414823293686,
      "epoch": 0.12191955882045229,
      "grad_norm": 8.662064552307129,
      "learning_rate": 7.2453856987442604e-06,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8233033016324043,
      "num_tokens": 47251991.0,
      "step": 39330
    },
    {
      "entropy": 1.8904896154999733,
      "epoch": 0.12195055794550198,
      "grad_norm": 8.939061164855957,
      "learning_rate": 7.244464749279004e-06,
      "loss": 0.547,
      "mean_token_accuracy": 0.8356327429413796,
      "num_tokens": 47264187.0,
      "step": 39340
    },
    {
      "entropy": 2.0048918604850767,
      "epoch": 0.12198155707055168,
      "grad_norm": 7.971733570098877,
      "learning_rate": 7.243544150905766e-06,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8271526664495468,
      "num_tokens": 47275694.0,
      "step": 39350
    },
    {
      "entropy": 2.0341754078865053,
      "epoch": 0.12201255619560138,
      "grad_norm": 8.931192398071289,
      "learning_rate": 7.242623903401524e-06,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8209575936198235,
      "num_tokens": 47286270.0,
      "step": 39360
    },
    {
      "entropy": 1.960363219678402,
      "epoch": 0.12204355532065107,
      "grad_norm": 8.31447696685791,
      "learning_rate": 7.241704006543459e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8278711885213852,
      "num_tokens": 47297957.0,
      "step": 39370
    },
    {
      "entropy": 1.9565995991230012,
      "epoch": 0.12207455444570077,
      "grad_norm": 10.019770622253418,
      "learning_rate": 7.240784460108944e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.830487422645092,
      "num_tokens": 47310043.0,
      "step": 39380
    },
    {
      "entropy": 1.8893634766340255,
      "epoch": 0.12210555357075047,
      "grad_norm": 5.0957865715026855,
      "learning_rate": 7.239865263875553e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8399478271603584,
      "num_tokens": 47322846.0,
      "step": 39390
    },
    {
      "entropy": 1.9584024906158448,
      "epoch": 0.12213655269580016,
      "grad_norm": 9.545947074890137,
      "learning_rate": 7.2389464176210585e-06,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8273420795798302,
      "num_tokens": 47333837.0,
      "step": 39400
    },
    {
      "entropy": 1.9519995272159576,
      "epoch": 0.12216755182084986,
      "grad_norm": 6.452544212341309,
      "learning_rate": 7.2380279211234304e-06,
      "loss": 0.5955,
      "mean_token_accuracy": 0.8233728304505348,
      "num_tokens": 47345921.0,
      "step": 39410
    },
    {
      "entropy": 1.9542613357305527,
      "epoch": 0.12219855094589956,
      "grad_norm": 9.183480262756348,
      "learning_rate": 7.237109774160834e-06,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8244384691119194,
      "num_tokens": 47357133.0,
      "step": 39420
    },
    {
      "entropy": 2.000173199176788,
      "epoch": 0.12222955007094925,
      "grad_norm": 10.723604202270508,
      "learning_rate": 7.236191976511631e-06,
      "loss": 0.5979,
      "mean_token_accuracy": 0.823377488553524,
      "num_tokens": 47367446.0,
      "step": 39430
    },
    {
      "entropy": 2.0126789212226868,
      "epoch": 0.12226054919599895,
      "grad_norm": 9.821727752685547,
      "learning_rate": 7.235274527954382e-06,
      "loss": 0.636,
      "mean_token_accuracy": 0.8074411496520042,
      "num_tokens": 47378541.0,
      "step": 39440
    },
    {
      "entropy": 1.9093073785305024,
      "epoch": 0.12229154832104863,
      "grad_norm": 9.164787292480469,
      "learning_rate": 7.234357428267842e-06,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8361682832241059,
      "num_tokens": 47390114.0,
      "step": 39450
    },
    {
      "entropy": 1.8918178245425223,
      "epoch": 0.12232254744609833,
      "grad_norm": 9.560025215148926,
      "learning_rate": 7.233440677230964e-06,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8367931827902794,
      "num_tokens": 47402972.0,
      "step": 39460
    },
    {
      "entropy": 1.9258590131998061,
      "epoch": 0.12235354657114803,
      "grad_norm": 8.87900161743164,
      "learning_rate": 7.232524274622897e-06,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8345812693238258,
      "num_tokens": 47414317.0,
      "step": 39470
    },
    {
      "entropy": 1.8653698325157166,
      "epoch": 0.12238454569619772,
      "grad_norm": 9.456033706665039,
      "learning_rate": 7.231608220222983e-06,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8317572355270386,
      "num_tokens": 47426321.0,
      "step": 39480
    },
    {
      "entropy": 1.9535607546567917,
      "epoch": 0.12241554482124742,
      "grad_norm": 4.313974857330322,
      "learning_rate": 7.230692513810767e-06,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8326726511120797,
      "num_tokens": 47437625.0,
      "step": 39490
    },
    {
      "entropy": 1.8951896831393242,
      "epoch": 0.12244654394629712,
      "grad_norm": 8.987655639648438,
      "learning_rate": 7.229777155165975e-06,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8377861216664314,
      "num_tokens": 47450365.0,
      "step": 39500
    },
    {
      "entropy": 1.9469685688614846,
      "epoch": 0.12247754307134681,
      "grad_norm": 8.353129386901855,
      "learning_rate": 7.228862144068547e-06,
      "loss": 0.5508,
      "mean_token_accuracy": 0.828496278822422,
      "num_tokens": 47462011.0,
      "step": 39510
    },
    {
      "entropy": 1.8816884592175485,
      "epoch": 0.12250854219639651,
      "grad_norm": 8.542236328125,
      "learning_rate": 7.227947480298603e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8291819423437119,
      "num_tokens": 47474823.0,
      "step": 39520
    },
    {
      "entropy": 1.8621670812368394,
      "epoch": 0.1225395413214462,
      "grad_norm": 8.68166446685791,
      "learning_rate": 7.227033163636462e-06,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8337508887052536,
      "num_tokens": 47487106.0,
      "step": 39530
    },
    {
      "entropy": 1.9243188560009004,
      "epoch": 0.1225705404464959,
      "grad_norm": 9.343770027160645,
      "learning_rate": 7.226119193862641e-06,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8281685680150985,
      "num_tokens": 47498743.0,
      "step": 39540
    },
    {
      "entropy": 1.9087723910808563,
      "epoch": 0.1226015395715456,
      "grad_norm": 10.595560073852539,
      "learning_rate": 7.225205570757848e-06,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8282433733344078,
      "num_tokens": 47511087.0,
      "step": 39550
    },
    {
      "entropy": 1.9597302109003067,
      "epoch": 0.1226325386965953,
      "grad_norm": 9.186569213867188,
      "learning_rate": 7.224292294102985e-06,
      "loss": 0.6176,
      "mean_token_accuracy": 0.8227703988552093,
      "num_tokens": 47522853.0,
      "step": 39560
    },
    {
      "entropy": 1.8456875741481782,
      "epoch": 0.12266353782164498,
      "grad_norm": 4.475244998931885,
      "learning_rate": 7.223379363679148e-06,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8218344271183013,
      "num_tokens": 47535932.0,
      "step": 39570
    },
    {
      "entropy": 1.8878474682569504,
      "epoch": 0.12269453694669467,
      "grad_norm": 8.224106788635254,
      "learning_rate": 7.222466779267628e-06,
      "loss": 0.6166,
      "mean_token_accuracy": 0.8177327990531922,
      "num_tokens": 47547596.0,
      "step": 39580
    },
    {
      "entropy": 1.9054535388946534,
      "epoch": 0.12272553607174437,
      "grad_norm": 9.366242408752441,
      "learning_rate": 7.221554540649909e-06,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8298028498888016,
      "num_tokens": 47560679.0,
      "step": 39590
    },
    {
      "entropy": 1.9364402756094932,
      "epoch": 0.12275653519679407,
      "grad_norm": 9.393874168395996,
      "learning_rate": 7.220642647607665e-06,
      "loss": 0.5802,
      "mean_token_accuracy": 0.8262747645378112,
      "num_tokens": 47572344.0,
      "step": 39600
    },
    {
      "entropy": 1.904061770439148,
      "epoch": 0.12278753432184376,
      "grad_norm": 11.156982421875,
      "learning_rate": 7.219731099922768e-06,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8359477326273919,
      "num_tokens": 47583272.0,
      "step": 39610
    },
    {
      "entropy": 1.8150178447365761,
      "epoch": 0.12281853344689346,
      "grad_norm": 8.800067901611328,
      "learning_rate": 7.218819897377277e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.84474868029356,
      "num_tokens": 47596939.0,
      "step": 39620
    },
    {
      "entropy": 1.7370637387037278,
      "epoch": 0.12284953257194316,
      "grad_norm": 8.489538192749023,
      "learning_rate": 7.217909039753451e-06,
      "loss": 0.4292,
      "mean_token_accuracy": 0.8503855854272843,
      "num_tokens": 47610478.0,
      "step": 39630
    },
    {
      "entropy": 1.8361937582492829,
      "epoch": 0.12288053169699285,
      "grad_norm": 10.117167472839355,
      "learning_rate": 7.216998526833735e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8379788339138031,
      "num_tokens": 47623032.0,
      "step": 39640
    },
    {
      "entropy": 1.880786618590355,
      "epoch": 0.12291153082204255,
      "grad_norm": 8.611593246459961,
      "learning_rate": 7.216088358400767e-06,
      "loss": 0.5068,
      "mean_token_accuracy": 0.841580656170845,
      "num_tokens": 47634664.0,
      "step": 39650
    },
    {
      "entropy": 1.8434013739228248,
      "epoch": 0.12294252994709225,
      "grad_norm": 11.259652137756348,
      "learning_rate": 7.2151785342373795e-06,
      "loss": 0.544,
      "mean_token_accuracy": 0.8298772439360619,
      "num_tokens": 47646858.0,
      "step": 39660
    },
    {
      "entropy": 1.882999302446842,
      "epoch": 0.12297352907214194,
      "grad_norm": 9.449716567993164,
      "learning_rate": 7.214269054126593e-06,
      "loss": 0.6034,
      "mean_token_accuracy": 0.8209249958395958,
      "num_tokens": 47658958.0,
      "step": 39670
    },
    {
      "entropy": 1.9060813054442405,
      "epoch": 0.12300452819719164,
      "grad_norm": 4.450017929077148,
      "learning_rate": 7.2133599178516235e-06,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8366493061184883,
      "num_tokens": 47670981.0,
      "step": 39680
    },
    {
      "entropy": 2.007936453819275,
      "epoch": 0.12303552732224134,
      "grad_norm": 10.02921199798584,
      "learning_rate": 7.212451125195874e-06,
      "loss": 0.6157,
      "mean_token_accuracy": 0.8246709734201432,
      "num_tokens": 47681702.0,
      "step": 39690
    },
    {
      "entropy": 2.0430832996964456,
      "epoch": 0.12306652644729102,
      "grad_norm": 9.025290489196777,
      "learning_rate": 7.211542675942941e-06,
      "loss": 0.6219,
      "mean_token_accuracy": 0.816537082195282,
      "num_tokens": 47692748.0,
      "step": 39700
    },
    {
      "entropy": 1.9663964182138443,
      "epoch": 0.12309752557234072,
      "grad_norm": 10.015572547912598,
      "learning_rate": 7.2106345698766134e-06,
      "loss": 0.5715,
      "mean_token_accuracy": 0.814094303548336,
      "num_tokens": 47703817.0,
      "step": 39710
    },
    {
      "entropy": 1.793771331012249,
      "epoch": 0.12312852469739041,
      "grad_norm": 3.0519356727600098,
      "learning_rate": 7.209726806780866e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8429444998502731,
      "num_tokens": 47717762.0,
      "step": 39720
    },
    {
      "entropy": 1.7859131276607514,
      "epoch": 0.12315952382244011,
      "grad_norm": 3.6266133785247803,
      "learning_rate": 7.208819386439868e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8404285296797752,
      "num_tokens": 47731173.0,
      "step": 39730
    },
    {
      "entropy": 1.8794568166136743,
      "epoch": 0.1231905229474898,
      "grad_norm": 4.375041484832764,
      "learning_rate": 7.207912308637975e-06,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8297057136893272,
      "num_tokens": 47743900.0,
      "step": 39740
    },
    {
      "entropy": 1.9013297393918038,
      "epoch": 0.1232215220725395,
      "grad_norm": 9.238868713378906,
      "learning_rate": 7.207005573159738e-06,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8304916322231293,
      "num_tokens": 47755784.0,
      "step": 39750
    },
    {
      "entropy": 1.9574954420328141,
      "epoch": 0.1232525211975892,
      "grad_norm": 10.237153053283691,
      "learning_rate": 7.2060991797898904e-06,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8244188159704209,
      "num_tokens": 47766731.0,
      "step": 39760
    },
    {
      "entropy": 1.8875176712870598,
      "epoch": 0.1232835203226389,
      "grad_norm": 10.606158256530762,
      "learning_rate": 7.205193128313362e-06,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8495481833815575,
      "num_tokens": 47778152.0,
      "step": 39770
    },
    {
      "entropy": 1.9175768822431565,
      "epoch": 0.12331451944768859,
      "grad_norm": 8.000387191772461,
      "learning_rate": 7.204287418515269e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8478017941117286,
      "num_tokens": 47788937.0,
      "step": 39780
    },
    {
      "entropy": 1.8297215834259988,
      "epoch": 0.12334551857273829,
      "grad_norm": 10.035210609436035,
      "learning_rate": 7.203382050180914e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8538377195596695,
      "num_tokens": 47801439.0,
      "step": 39790
    },
    {
      "entropy": 1.9494767993688584,
      "epoch": 0.12337651769778799,
      "grad_norm": 9.916494369506836,
      "learning_rate": 7.202477023095793e-06,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8265805870294571,
      "num_tokens": 47811893.0,
      "step": 39800
    },
    {
      "entropy": 1.8632332697510718,
      "epoch": 0.12340751682283768,
      "grad_norm": 8.924822807312012,
      "learning_rate": 7.201572337045587e-06,
      "loss": 0.5439,
      "mean_token_accuracy": 0.835116508603096,
      "num_tokens": 47823892.0,
      "step": 39810
    },
    {
      "entropy": 1.8363080993294716,
      "epoch": 0.12343851594788736,
      "grad_norm": 4.758782863616943,
      "learning_rate": 7.200667991816167e-06,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8292215019464493,
      "num_tokens": 47835997.0,
      "step": 39820
    },
    {
      "entropy": 1.9205212816596031,
      "epoch": 0.12346951507293706,
      "grad_norm": 8.591593742370605,
      "learning_rate": 7.199763987193592e-06,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8230644062161445,
      "num_tokens": 47847772.0,
      "step": 39830
    },
    {
      "entropy": 1.9247150912880897,
      "epoch": 0.12350051419798676,
      "grad_norm": 10.074462890625,
      "learning_rate": 7.19886032296411e-06,
      "loss": 0.57,
      "mean_token_accuracy": 0.8206574454903602,
      "num_tokens": 47859033.0,
      "step": 39840
    },
    {
      "entropy": 1.9125849097967147,
      "epoch": 0.12353151332303645,
      "grad_norm": 9.70457649230957,
      "learning_rate": 7.197956998914155e-06,
      "loss": 0.574,
      "mean_token_accuracy": 0.82592853307724,
      "num_tokens": 47870577.0,
      "step": 39850
    },
    {
      "entropy": 1.92913771122694,
      "epoch": 0.12356251244808615,
      "grad_norm": 8.795923233032227,
      "learning_rate": 7.197054014830351e-06,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8308982938528061,
      "num_tokens": 47883093.0,
      "step": 39860
    },
    {
      "entropy": 1.8687468573451043,
      "epoch": 0.12359351157313585,
      "grad_norm": 9.503414154052734,
      "learning_rate": 7.196151370499505e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8317637592554092,
      "num_tokens": 47895551.0,
      "step": 39870
    },
    {
      "entropy": 1.969892618060112,
      "epoch": 0.12362451069818554,
      "grad_norm": 9.0645112991333,
      "learning_rate": 7.195249065708615e-06,
      "loss": 0.6059,
      "mean_token_accuracy": 0.818208034336567,
      "num_tokens": 47906405.0,
      "step": 39880
    },
    {
      "entropy": 1.8821232318878174,
      "epoch": 0.12365550982323524,
      "grad_norm": 8.767228126525879,
      "learning_rate": 7.194347100244863e-06,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8397012487053871,
      "num_tokens": 47918131.0,
      "step": 39890
    },
    {
      "entropy": 1.8662964954972268,
      "epoch": 0.12368650894828494,
      "grad_norm": 9.380695343017578,
      "learning_rate": 7.1934454738956235e-06,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8346205174922943,
      "num_tokens": 47929900.0,
      "step": 39900
    },
    {
      "entropy": 1.9048209875822066,
      "epoch": 0.12371750807333463,
      "grad_norm": 7.6859354972839355,
      "learning_rate": 7.19254418644845e-06,
      "loss": 0.5487,
      "mean_token_accuracy": 0.8317638859152794,
      "num_tokens": 47941788.0,
      "step": 39910
    },
    {
      "entropy": 1.9060806199908256,
      "epoch": 0.12374850719838433,
      "grad_norm": 14.849568367004395,
      "learning_rate": 7.1916432376910865e-06,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8300574287772179,
      "num_tokens": 47953554.0,
      "step": 39920
    },
    {
      "entropy": 1.8868515014648437,
      "epoch": 0.12377950632343403,
      "grad_norm": 8.418484687805176,
      "learning_rate": 7.19074262741146e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8337209805846214,
      "num_tokens": 47965620.0,
      "step": 39930
    },
    {
      "entropy": 1.9150093987584114,
      "epoch": 0.12381050544848371,
      "grad_norm": 8.23653793334961,
      "learning_rate": 7.18984235539769e-06,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8326784566044807,
      "num_tokens": 47977186.0,
      "step": 39940
    },
    {
      "entropy": 1.8675342485308648,
      "epoch": 0.1238415045735334,
      "grad_norm": 8.780365943908691,
      "learning_rate": 7.188942421438074e-06,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8340345472097397,
      "num_tokens": 47988962.0,
      "step": 39950
    },
    {
      "entropy": 1.8479683220386505,
      "epoch": 0.1238725036985831,
      "grad_norm": 9.722196578979492,
      "learning_rate": 7.188042825321099e-06,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8346738666296005,
      "num_tokens": 48001542.0,
      "step": 39960
    },
    {
      "entropy": 1.9192697882652283,
      "epoch": 0.1239035028236328,
      "grad_norm": 9.703927993774414,
      "learning_rate": 7.187143566835436e-06,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8207048952579499,
      "num_tokens": 48013716.0,
      "step": 39970
    },
    {
      "entropy": 1.9120296582579612,
      "epoch": 0.1239345019486825,
      "grad_norm": 8.731795310974121,
      "learning_rate": 7.186244645769942e-06,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8310200199484825,
      "num_tokens": 48026045.0,
      "step": 39980
    },
    {
      "entropy": 1.9499492287635802,
      "epoch": 0.12396550107373219,
      "grad_norm": 7.662987232208252,
      "learning_rate": 7.185346061913657e-06,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8196876659989357,
      "num_tokens": 48037456.0,
      "step": 39990
    },
    {
      "entropy": 1.832308356463909,
      "epoch": 0.12399650019878189,
      "grad_norm": 3.518549919128418,
      "learning_rate": 7.18444781505581e-06,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8477145195007324,
      "num_tokens": 48049861.0,
      "step": 40000
    },
    {
      "entropy": 1.86460652500391,
      "epoch": 0.12402749932383159,
      "grad_norm": 8.867399215698242,
      "learning_rate": 7.183549904985806e-06,
      "loss": 0.5642,
      "mean_token_accuracy": 0.82719586789608,
      "num_tokens": 48062011.0,
      "step": 40010
    },
    {
      "entropy": 1.880849677324295,
      "epoch": 0.12405849844888128,
      "grad_norm": 4.325946807861328,
      "learning_rate": 7.182652331493244e-06,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8259584322571755,
      "num_tokens": 48074386.0,
      "step": 40020
    },
    {
      "entropy": 1.9078103736042977,
      "epoch": 0.12408949757393098,
      "grad_norm": 10.76919937133789,
      "learning_rate": 7.181755094367901e-06,
      "loss": 0.507,
      "mean_token_accuracy": 0.8391536310315132,
      "num_tokens": 48085967.0,
      "step": 40030
    },
    {
      "entropy": 1.8335755363106727,
      "epoch": 0.12412049669898068,
      "grad_norm": 12.115399360656738,
      "learning_rate": 7.1808581933997365e-06,
      "loss": 0.4988,
      "mean_token_accuracy": 0.833656270802021,
      "num_tokens": 48098343.0,
      "step": 40040
    },
    {
      "entropy": 1.8345079183578492,
      "epoch": 0.12415149582403037,
      "grad_norm": 9.696667671203613,
      "learning_rate": 7.1799616283789e-06,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8341936364769935,
      "num_tokens": 48111160.0,
      "step": 40050
    },
    {
      "entropy": 1.7580282986164093,
      "epoch": 0.12418249494908007,
      "grad_norm": 8.78324031829834,
      "learning_rate": 7.179065399095719e-06,
      "loss": 0.5413,
      "mean_token_accuracy": 0.8408230692148209,
      "num_tokens": 48125547.0,
      "step": 40060
    },
    {
      "entropy": 1.8306955844163895,
      "epoch": 0.12421349407412975,
      "grad_norm": 9.833135604858398,
      "learning_rate": 7.178169505340706e-06,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8345167860388756,
      "num_tokens": 48137633.0,
      "step": 40070
    },
    {
      "entropy": 1.819261023402214,
      "epoch": 0.12424449319917945,
      "grad_norm": 13.903244972229004,
      "learning_rate": 7.177273946904556e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8336721554398536,
      "num_tokens": 48151380.0,
      "step": 40080
    },
    {
      "entropy": 1.8060119107365609,
      "epoch": 0.12427549232422914,
      "grad_norm": 3.684872627258301,
      "learning_rate": 7.176378723578145e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8292297214269638,
      "num_tokens": 48164473.0,
      "step": 40090
    },
    {
      "entropy": 1.8856921419501305,
      "epoch": 0.12430649144927884,
      "grad_norm": 10.007135391235352,
      "learning_rate": 7.175483835152539e-06,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8323625057935715,
      "num_tokens": 48176696.0,
      "step": 40100
    },
    {
      "entropy": 1.7466407373547554,
      "epoch": 0.12433749057432854,
      "grad_norm": 4.418228626251221,
      "learning_rate": 7.174589281418974e-06,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8500324800610543,
      "num_tokens": 48190217.0,
      "step": 40110
    },
    {
      "entropy": 1.9157506242394446,
      "epoch": 0.12436848969937823,
      "grad_norm": 8.385701179504395,
      "learning_rate": 7.17369506216888e-06,
      "loss": 0.5272,
      "mean_token_accuracy": 0.825266569852829,
      "num_tokens": 48201921.0,
      "step": 40120
    },
    {
      "entropy": 1.858496817946434,
      "epoch": 0.12439948882442793,
      "grad_norm": 4.637474060058594,
      "learning_rate": 7.172801177193862e-06,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8193719312548637,
      "num_tokens": 48214518.0,
      "step": 40130
    },
    {
      "entropy": 1.8998822063207625,
      "epoch": 0.12443048794947763,
      "grad_norm": 9.965042114257812,
      "learning_rate": 7.171907626285708e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.838045471906662,
      "num_tokens": 48225714.0,
      "step": 40140
    },
    {
      "entropy": 1.8577251955866814,
      "epoch": 0.12446148707452732,
      "grad_norm": 9.38553237915039,
      "learning_rate": 7.171014409236389e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8318695619702339,
      "num_tokens": 48238441.0,
      "step": 40150
    },
    {
      "entropy": 1.9177394479513168,
      "epoch": 0.12449248619957702,
      "grad_norm": 8.650186538696289,
      "learning_rate": 7.1701215258380555e-06,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8154436364769936,
      "num_tokens": 48250481.0,
      "step": 40160
    },
    {
      "entropy": 1.9266600027680396,
      "epoch": 0.12452348532462672,
      "grad_norm": 8.298869132995605,
      "learning_rate": 7.169228975883042e-06,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8180478677153588,
      "num_tokens": 48262220.0,
      "step": 40170
    },
    {
      "entropy": 1.9378298938274383,
      "epoch": 0.12455448444967641,
      "grad_norm": 4.73800802230835,
      "learning_rate": 7.16833675916386e-06,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8223996505141258,
      "num_tokens": 48274072.0,
      "step": 40180
    },
    {
      "entropy": 1.8640633895993233,
      "epoch": 0.1245854835747261,
      "grad_norm": 10.636198043823242,
      "learning_rate": 7.167444875473203e-06,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8277820602059365,
      "num_tokens": 48286360.0,
      "step": 40190
    },
    {
      "entropy": 1.9393206879496574,
      "epoch": 0.1246164826997758,
      "grad_norm": 4.20045804977417,
      "learning_rate": 7.166553324603949e-06,
      "loss": 0.578,
      "mean_token_accuracy": 0.8166181564331054,
      "num_tokens": 48297250.0,
      "step": 40200
    },
    {
      "entropy": 1.8666782602667809,
      "epoch": 0.12464748182482549,
      "grad_norm": 9.612086296081543,
      "learning_rate": 7.165662106349151e-06,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8358882114291191,
      "num_tokens": 48309504.0,
      "step": 40210
    },
    {
      "entropy": 1.8223158940672874,
      "epoch": 0.12467848094987519,
      "grad_norm": 8.702556610107422,
      "learning_rate": 7.164771220502042e-06,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8401806324720382,
      "num_tokens": 48321758.0,
      "step": 40220
    },
    {
      "entropy": 1.8826921790838242,
      "epoch": 0.12470948007492488,
      "grad_norm": 8.89100456237793,
      "learning_rate": 7.16388066685604e-06,
      "loss": 0.5317,
      "mean_token_accuracy": 0.823623314499855,
      "num_tokens": 48334423.0,
      "step": 40230
    },
    {
      "entropy": 1.935545490682125,
      "epoch": 0.12474047919997458,
      "grad_norm": 8.564839363098145,
      "learning_rate": 7.16299044520474e-06,
      "loss": 0.552,
      "mean_token_accuracy": 0.8268688634037972,
      "num_tokens": 48345619.0,
      "step": 40240
    },
    {
      "entropy": 1.835379946231842,
      "epoch": 0.12477147832502428,
      "grad_norm": 8.765735626220703,
      "learning_rate": 7.162100555341913e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8396931126713753,
      "num_tokens": 48358114.0,
      "step": 40250
    },
    {
      "entropy": 1.8680355235934258,
      "epoch": 0.12480247745007397,
      "grad_norm": 10.267287254333496,
      "learning_rate": 7.161210997061516e-06,
      "loss": 0.6458,
      "mean_token_accuracy": 0.8112869113683701,
      "num_tokens": 48370421.0,
      "step": 40260
    },
    {
      "entropy": 1.8904762849211694,
      "epoch": 0.12483347657512367,
      "grad_norm": 10.460432052612305,
      "learning_rate": 7.1603217701576784e-06,
      "loss": 0.5603,
      "mean_token_accuracy": 0.8190462201833725,
      "num_tokens": 48382227.0,
      "step": 40270
    },
    {
      "entropy": 1.8971378430724144,
      "epoch": 0.12486447570017337,
      "grad_norm": 7.277894020080566,
      "learning_rate": 7.159432874424715e-06,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8331350639462471,
      "num_tokens": 48393629.0,
      "step": 40280
    },
    {
      "entropy": 1.941260239481926,
      "epoch": 0.12489547482522306,
      "grad_norm": 9.197176933288574,
      "learning_rate": 7.158544309657114e-06,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8198348104953765,
      "num_tokens": 48404836.0,
      "step": 40290
    },
    {
      "entropy": 1.834352783858776,
      "epoch": 0.12492647395027276,
      "grad_norm": 4.567273139953613,
      "learning_rate": 7.157656075649543e-06,
      "loss": 0.556,
      "mean_token_accuracy": 0.8348585799336433,
      "num_tokens": 48416706.0,
      "step": 40300
    },
    {
      "entropy": 1.7823708355426788,
      "epoch": 0.12495747307532244,
      "grad_norm": 4.625446796417236,
      "learning_rate": 7.1567681721968504e-06,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8396563857793808,
      "num_tokens": 48430018.0,
      "step": 40310
    },
    {
      "entropy": 1.8015142977237701,
      "epoch": 0.12498847220037214,
      "grad_norm": 10.068672180175781,
      "learning_rate": 7.155880599094063e-06,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8157304137945175,
      "num_tokens": 48442115.0,
      "step": 40320
    },
    {
      "entropy": 1.8101561158895492,
      "epoch": 0.12501947132542185,
      "grad_norm": 9.83664321899414,
      "learning_rate": 7.154993356136379e-06,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8348063215613365,
      "num_tokens": 48454395.0,
      "step": 40330
    },
    {
      "entropy": 1.8592476963996887,
      "epoch": 0.12505047045047155,
      "grad_norm": 8.112977027893066,
      "learning_rate": 7.154106443119184e-06,
      "loss": 0.525,
      "mean_token_accuracy": 0.8234350681304932,
      "num_tokens": 48466396.0,
      "step": 40340
    },
    {
      "entropy": 1.8576691791415214,
      "epoch": 0.12508146957552124,
      "grad_norm": 4.848905086517334,
      "learning_rate": 7.153219859838033e-06,
      "loss": 0.5681,
      "mean_token_accuracy": 0.8281613394618035,
      "num_tokens": 48479521.0,
      "step": 40350
    },
    {
      "entropy": 1.831693847477436,
      "epoch": 0.12511246870057094,
      "grad_norm": 7.906863689422607,
      "learning_rate": 7.152333606088664e-06,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8364334732294083,
      "num_tokens": 48492267.0,
      "step": 40360
    },
    {
      "entropy": 1.805116631090641,
      "epoch": 0.12514346782562064,
      "grad_norm": 7.8519415855407715,
      "learning_rate": 7.151447681666986e-06,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8286394074559211,
      "num_tokens": 48505894.0,
      "step": 40370
    },
    {
      "entropy": 1.8447122775018214,
      "epoch": 0.1251744669506703,
      "grad_norm": 7.955524444580078,
      "learning_rate": 7.150562086369092e-06,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8388951927423477,
      "num_tokens": 48518153.0,
      "step": 40380
    },
    {
      "entropy": 1.8237828686833382,
      "epoch": 0.12520546607572,
      "grad_norm": 10.909595489501953,
      "learning_rate": 7.149676819991247e-06,
      "loss": 0.5013,
      "mean_token_accuracy": 0.835619455575943,
      "num_tokens": 48530887.0,
      "step": 40390
    },
    {
      "entropy": 1.9085283294320106,
      "epoch": 0.1252364652007697,
      "grad_norm": 9.184427261352539,
      "learning_rate": 7.148791882329893e-06,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8193691223859787,
      "num_tokens": 48542762.0,
      "step": 40400
    },
    {
      "entropy": 1.8631555259227752,
      "epoch": 0.1252674643258194,
      "grad_norm": 9.436893463134766,
      "learning_rate": 7.147907273181649e-06,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8169412419199944,
      "num_tokens": 48555091.0,
      "step": 40410
    },
    {
      "entropy": 1.836050059646368,
      "epoch": 0.1252984634508691,
      "grad_norm": 10.38294792175293,
      "learning_rate": 7.1470229923433125e-06,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8353980749845504,
      "num_tokens": 48567717.0,
      "step": 40420
    },
    {
      "entropy": 1.9533065795898437,
      "epoch": 0.1253294625759188,
      "grad_norm": 3.965822696685791,
      "learning_rate": 7.146139039611852e-06,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8248067319393158,
      "num_tokens": 48579059.0,
      "step": 40430
    },
    {
      "entropy": 1.8541604191064835,
      "epoch": 0.12536046170096848,
      "grad_norm": 8.920310020446777,
      "learning_rate": 7.1452554147844155e-06,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8245360970497131,
      "num_tokens": 48591120.0,
      "step": 40440
    },
    {
      "entropy": 1.8779646888375283,
      "epoch": 0.12539146082601818,
      "grad_norm": 8.486520767211914,
      "learning_rate": 7.144372117658325e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8279622659087181,
      "num_tokens": 48603019.0,
      "step": 40450
    },
    {
      "entropy": 1.9320873647928238,
      "epoch": 0.12542245995106788,
      "grad_norm": 11.511739730834961,
      "learning_rate": 7.143489148031079e-06,
      "loss": 0.5977,
      "mean_token_accuracy": 0.8194258600473404,
      "num_tokens": 48615247.0,
      "step": 40460
    },
    {
      "entropy": 1.9659615710377694,
      "epoch": 0.12545345907611757,
      "grad_norm": 9.763450622558594,
      "learning_rate": 7.142606505700348e-06,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8184546142816543,
      "num_tokens": 48626590.0,
      "step": 40470
    },
    {
      "entropy": 1.9306480765342713,
      "epoch": 0.12548445820116727,
      "grad_norm": 8.713022232055664,
      "learning_rate": 7.141724190463982e-06,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8337675094604492,
      "num_tokens": 48637473.0,
      "step": 40480
    },
    {
      "entropy": 1.798674686253071,
      "epoch": 0.12551545732621697,
      "grad_norm": 11.227383613586426,
      "learning_rate": 7.140842202120004e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8351314097642899,
      "num_tokens": 48651671.0,
      "step": 40490
    },
    {
      "entropy": 1.9205691695213318,
      "epoch": 0.12554645645126666,
      "grad_norm": 8.593749046325684,
      "learning_rate": 7.139960540466611e-06,
      "loss": 0.5708,
      "mean_token_accuracy": 0.8320924565196037,
      "num_tokens": 48662844.0,
      "step": 40500
    },
    {
      "entropy": 1.9018804863095284,
      "epoch": 0.12557745557631636,
      "grad_norm": 9.196910858154297,
      "learning_rate": 7.13907920530217e-06,
      "loss": 0.5599,
      "mean_token_accuracy": 0.8320263445377349,
      "num_tokens": 48674513.0,
      "step": 40510
    },
    {
      "entropy": 1.8978378668427467,
      "epoch": 0.12560845470136606,
      "grad_norm": 7.797028064727783,
      "learning_rate": 7.138198196425235e-06,
      "loss": 0.541,
      "mean_token_accuracy": 0.826276271045208,
      "num_tokens": 48686033.0,
      "step": 40520
    },
    {
      "entropy": 1.8717727780342102,
      "epoch": 0.12563945382641575,
      "grad_norm": 8.808572769165039,
      "learning_rate": 7.137317513634519e-06,
      "loss": 0.5139,
      "mean_token_accuracy": 0.827865032851696,
      "num_tokens": 48698508.0,
      "step": 40530
    },
    {
      "entropy": 1.8457770988345146,
      "epoch": 0.12567045295146545,
      "grad_norm": 8.519002914428711,
      "learning_rate": 7.136437156728917e-06,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8214613869786263,
      "num_tokens": 48711134.0,
      "step": 40540
    },
    {
      "entropy": 1.935419850051403,
      "epoch": 0.12570145207651515,
      "grad_norm": 8.7819185256958,
      "learning_rate": 7.135557125507497e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8435436561703682,
      "num_tokens": 48722163.0,
      "step": 40550
    },
    {
      "entropy": 1.8241950839757919,
      "epoch": 0.12573245120156484,
      "grad_norm": 3.9725289344787598,
      "learning_rate": 7.134677419769499e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.8425084307789803,
      "num_tokens": 48735211.0,
      "step": 40560
    },
    {
      "entropy": 1.8317349657416344,
      "epoch": 0.12576345032661454,
      "grad_norm": 3.390087842941284,
      "learning_rate": 7.133798039314337e-06,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8422825932502747,
      "num_tokens": 48748014.0,
      "step": 40570
    },
    {
      "entropy": 1.7657338783144951,
      "epoch": 0.12579444945166424,
      "grad_norm": 8.908466339111328,
      "learning_rate": 7.1329189839415956e-06,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8356280282139779,
      "num_tokens": 48761505.0,
      "step": 40580
    },
    {
      "entropy": 1.8681147009134293,
      "epoch": 0.12582544857671393,
      "grad_norm": 9.203214645385742,
      "learning_rate": 7.132040253451038e-06,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8317851752042771,
      "num_tokens": 48773868.0,
      "step": 40590
    },
    {
      "entropy": 1.8376054048538208,
      "epoch": 0.12585644770176363,
      "grad_norm": 4.301458835601807,
      "learning_rate": 7.131161847642594e-06,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8354909658432007,
      "num_tokens": 48786722.0,
      "step": 40600
    },
    {
      "entropy": 1.968809324502945,
      "epoch": 0.12588744682681333,
      "grad_norm": 9.997700691223145,
      "learning_rate": 7.130283766316368e-06,
      "loss": 0.6149,
      "mean_token_accuracy": 0.8223230019211769,
      "num_tokens": 48797626.0,
      "step": 40610
    },
    {
      "entropy": 1.9517834931612015,
      "epoch": 0.125918445951863,
      "grad_norm": 8.660726547241211,
      "learning_rate": 7.1294060092726395e-06,
      "loss": 0.6263,
      "mean_token_accuracy": 0.8167603313922882,
      "num_tokens": 48809941.0,
      "step": 40620
    },
    {
      "entropy": 1.854266294836998,
      "epoch": 0.1259494450769127,
      "grad_norm": 8.519408226013184,
      "learning_rate": 7.128528576311854e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8381375819444656,
      "num_tokens": 48822363.0,
      "step": 40630
    },
    {
      "entropy": 1.8118067890405656,
      "epoch": 0.1259804442019624,
      "grad_norm": 10.48330020904541,
      "learning_rate": 7.127651467234633e-06,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8347623705863952,
      "num_tokens": 48835835.0,
      "step": 40640
    },
    {
      "entropy": 1.8399090513586998,
      "epoch": 0.12601144332701208,
      "grad_norm": 9.834306716918945,
      "learning_rate": 7.12677468184177e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8451863750815392,
      "num_tokens": 48848639.0,
      "step": 40650
    },
    {
      "entropy": 1.8340865150094032,
      "epoch": 0.12604244245206178,
      "grad_norm": 10.710108757019043,
      "learning_rate": 7.125898219934229e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8310162261128425,
      "num_tokens": 48861498.0,
      "step": 40660
    },
    {
      "entropy": 1.9059862732887267,
      "epoch": 0.12607344157711148,
      "grad_norm": 8.74216365814209,
      "learning_rate": 7.125022081313144e-06,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8258930623531342,
      "num_tokens": 48873676.0,
      "step": 40670
    },
    {
      "entropy": 1.8104983791708946,
      "epoch": 0.12610444070216117,
      "grad_norm": 3.50832462310791,
      "learning_rate": 7.124146265779823e-06,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8388922065496445,
      "num_tokens": 48887359.0,
      "step": 40680
    },
    {
      "entropy": 1.886466035246849,
      "epoch": 0.12613543982721087,
      "grad_norm": 9.473098754882812,
      "learning_rate": 7.123270773135742e-06,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8378918841481209,
      "num_tokens": 48898642.0,
      "step": 40690
    },
    {
      "entropy": 1.9465019404888153,
      "epoch": 0.12616643895226057,
      "grad_norm": 8.107137680053711,
      "learning_rate": 7.12239560318255e-06,
      "loss": 0.5828,
      "mean_token_accuracy": 0.8316583067178727,
      "num_tokens": 48909355.0,
      "step": 40700
    },
    {
      "entropy": 1.864406055212021,
      "epoch": 0.12619743807731026,
      "grad_norm": 7.923045635223389,
      "learning_rate": 7.121520755722065e-06,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8234068945050239,
      "num_tokens": 48921713.0,
      "step": 40710
    },
    {
      "entropy": 1.917890764772892,
      "epoch": 0.12622843720235996,
      "grad_norm": 8.965694427490234,
      "learning_rate": 7.120646230556275e-06,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8340031638741493,
      "num_tokens": 48933362.0,
      "step": 40720
    },
    {
      "entropy": 1.9002999886870384,
      "epoch": 0.12625943632740966,
      "grad_norm": 9.010135650634766,
      "learning_rate": 7.119772027487341e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8289919942617416,
      "num_tokens": 48945208.0,
      "step": 40730
    },
    {
      "entropy": 1.862901757657528,
      "epoch": 0.12629043545245935,
      "grad_norm": 9.43112564086914,
      "learning_rate": 7.118898146317591e-06,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8338583365082741,
      "num_tokens": 48957105.0,
      "step": 40740
    },
    {
      "entropy": 1.8847228810191154,
      "epoch": 0.12632143457750905,
      "grad_norm": 4.666851043701172,
      "learning_rate": 7.118024586849524e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8274060249328613,
      "num_tokens": 48970207.0,
      "step": 40750
    },
    {
      "entropy": 1.8614213794469834,
      "epoch": 0.12635243370255875,
      "grad_norm": 8.402881622314453,
      "learning_rate": 7.117151348885809e-06,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8313090309500695,
      "num_tokens": 48983272.0,
      "step": 40760
    },
    {
      "entropy": 1.8629606261849403,
      "epoch": 0.12638343282760844,
      "grad_norm": 3.4327898025512695,
      "learning_rate": 7.116278432229283e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8382740557193756,
      "num_tokens": 48995280.0,
      "step": 40770
    },
    {
      "entropy": 1.866362100839615,
      "epoch": 0.12641443195265814,
      "grad_norm": 6.627174377441406,
      "learning_rate": 7.1154058366829534e-06,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8405808374285698,
      "num_tokens": 49007496.0,
      "step": 40780
    },
    {
      "entropy": 1.9819986671209335,
      "epoch": 0.12644543107770784,
      "grad_norm": 9.322339057922363,
      "learning_rate": 7.114533562049997e-06,
      "loss": 0.6276,
      "mean_token_accuracy": 0.8162664338946343,
      "num_tokens": 49018388.0,
      "step": 40790
    },
    {
      "entropy": 1.849839760363102,
      "epoch": 0.12647643020275753,
      "grad_norm": 3.9357082843780518,
      "learning_rate": 7.113661608133757e-06,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8274443417787551,
      "num_tokens": 49030285.0,
      "step": 40800
    },
    {
      "entropy": 1.8829088985919953,
      "epoch": 0.12650742932780723,
      "grad_norm": 7.758213996887207,
      "learning_rate": 7.112789974737751e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.831216461956501,
      "num_tokens": 49042179.0,
      "step": 40810
    },
    {
      "entropy": 1.9304823964834212,
      "epoch": 0.12653842845285693,
      "grad_norm": 8.798954963684082,
      "learning_rate": 7.1119186616656555e-06,
      "loss": 0.5907,
      "mean_token_accuracy": 0.8195964187383652,
      "num_tokens": 49054008.0,
      "step": 40820
    },
    {
      "entropy": 1.9207462221384048,
      "epoch": 0.12656942757790662,
      "grad_norm": 9.050512313842773,
      "learning_rate": 7.111047668721327e-06,
      "loss": 0.575,
      "mean_token_accuracy": 0.8204113721847535,
      "num_tokens": 49066034.0,
      "step": 40830
    },
    {
      "entropy": 1.908941000699997,
      "epoch": 0.12660042670295632,
      "grad_norm": 4.945900917053223,
      "learning_rate": 7.11017699570878e-06,
      "loss": 0.559,
      "mean_token_accuracy": 0.8230469167232514,
      "num_tokens": 49077639.0,
      "step": 40840
    },
    {
      "entropy": 1.8663305729627608,
      "epoch": 0.12663142582800602,
      "grad_norm": 8.779427528381348,
      "learning_rate": 7.109306642432202e-06,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8439178988337517,
      "num_tokens": 49090125.0,
      "step": 40850
    },
    {
      "entropy": 1.9162701606750487,
      "epoch": 0.1266624249530557,
      "grad_norm": 9.195842742919922,
      "learning_rate": 7.108436608695949e-06,
      "loss": 0.5807,
      "mean_token_accuracy": 0.8183003440499306,
      "num_tokens": 49101421.0,
      "step": 40860
    },
    {
      "entropy": 1.9350932016968727,
      "epoch": 0.12669342407810538,
      "grad_norm": 9.8731689453125,
      "learning_rate": 7.10756689430454e-06,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8179444178938866,
      "num_tokens": 49113375.0,
      "step": 40870
    },
    {
      "entropy": 1.9218543514609336,
      "epoch": 0.12672442320315508,
      "grad_norm": 8.230029106140137,
      "learning_rate": 7.106697499062666e-06,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8303100064396858,
      "num_tokens": 49124773.0,
      "step": 40880
    },
    {
      "entropy": 1.8881905749440193,
      "epoch": 0.12675542232820478,
      "grad_norm": 5.785170555114746,
      "learning_rate": 7.105828422775184e-06,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8263862758874894,
      "num_tokens": 49137292.0,
      "step": 40890
    },
    {
      "entropy": 1.8283898428082466,
      "epoch": 0.12678642145325447,
      "grad_norm": 10.07806396484375,
      "learning_rate": 7.1049596652471145e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8381018027663231,
      "num_tokens": 49149805.0,
      "step": 40900
    },
    {
      "entropy": 1.8706416577100753,
      "epoch": 0.12681742057830417,
      "grad_norm": 8.058819770812988,
      "learning_rate": 7.104091226283651e-06,
      "loss": 0.5054,
      "mean_token_accuracy": 0.84037394374609,
      "num_tokens": 49160775.0,
      "step": 40910
    },
    {
      "entropy": 1.8797523587942124,
      "epoch": 0.12684841970335387,
      "grad_norm": 8.583271026611328,
      "learning_rate": 7.103223105690148e-06,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8319562718272209,
      "num_tokens": 49172562.0,
      "step": 40920
    },
    {
      "entropy": 1.81065763682127,
      "epoch": 0.12687941882840356,
      "grad_norm": 9.554974555969238,
      "learning_rate": 7.1023553032721315e-06,
      "loss": 0.5111,
      "mean_token_accuracy": 0.839790141582489,
      "num_tokens": 49185408.0,
      "step": 40930
    },
    {
      "entropy": 1.9385173588991165,
      "epoch": 0.12691041795345326,
      "grad_norm": 9.629557609558105,
      "learning_rate": 7.101487818835289e-06,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8281932666897773,
      "num_tokens": 49197278.0,
      "step": 40940
    },
    {
      "entropy": 1.8193309232592583,
      "epoch": 0.12694141707850295,
      "grad_norm": 8.458683013916016,
      "learning_rate": 7.100620652185476e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.8422279581427574,
      "num_tokens": 49210376.0,
      "step": 40950
    },
    {
      "entropy": 1.8436155632138251,
      "epoch": 0.12697241620355265,
      "grad_norm": 10.37833023071289,
      "learning_rate": 7.099753803128716e-06,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8349693238735199,
      "num_tokens": 49222925.0,
      "step": 40960
    },
    {
      "entropy": 1.836123764514923,
      "epoch": 0.12700341532860235,
      "grad_norm": 9.300790786743164,
      "learning_rate": 7.0988872714711934e-06,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8327738150954247,
      "num_tokens": 49235480.0,
      "step": 40970
    },
    {
      "entropy": 1.8801463842391968,
      "epoch": 0.12703441445365204,
      "grad_norm": 5.394232749938965,
      "learning_rate": 7.098021057019264e-06,
      "loss": 0.541,
      "mean_token_accuracy": 0.8303487420082092,
      "num_tokens": 49247573.0,
      "step": 40980
    },
    {
      "entropy": 1.8794305130839348,
      "epoch": 0.12706541357870174,
      "grad_norm": 9.10645580291748,
      "learning_rate": 7.097155159579446e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8420027121901512,
      "num_tokens": 49258928.0,
      "step": 40990
    },
    {
      "entropy": 1.9318907380104064,
      "epoch": 0.12709641270375144,
      "grad_norm": 9.83584976196289,
      "learning_rate": 7.09628957895842e-06,
      "loss": 0.5834,
      "mean_token_accuracy": 0.8214644491672516,
      "num_tokens": 49270094.0,
      "step": 41000
    },
    {
      "entropy": 1.8540407776832581,
      "epoch": 0.12712741182880113,
      "grad_norm": 9.077452659606934,
      "learning_rate": 7.095424314963037e-06,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8308973863720894,
      "num_tokens": 49282965.0,
      "step": 41010
    },
    {
      "entropy": 1.8202777698636055,
      "epoch": 0.12715841095385083,
      "grad_norm": 4.704195976257324,
      "learning_rate": 7.094559367400309e-06,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8255807936191559,
      "num_tokens": 49296075.0,
      "step": 41020
    },
    {
      "entropy": 1.9383866339921951,
      "epoch": 0.12718941007890053,
      "grad_norm": 9.143877983093262,
      "learning_rate": 7.093694736077415e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8385669678449631,
      "num_tokens": 49306680.0,
      "step": 41030
    },
    {
      "entropy": 1.742047442495823,
      "epoch": 0.12722040920395022,
      "grad_norm": 8.927555084228516,
      "learning_rate": 7.092830420801696e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8529659613966942,
      "num_tokens": 49320860.0,
      "step": 41040
    },
    {
      "entropy": 1.8793509498238563,
      "epoch": 0.12725140832899992,
      "grad_norm": 8.80550765991211,
      "learning_rate": 7.09196642138066e-06,
      "loss": 0.5804,
      "mean_token_accuracy": 0.819065073132515,
      "num_tokens": 49332538.0,
      "step": 41050
    },
    {
      "entropy": 1.8639131098985673,
      "epoch": 0.12728240745404962,
      "grad_norm": 8.924798965454102,
      "learning_rate": 7.091102737621975e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.829890587925911,
      "num_tokens": 49345210.0,
      "step": 41060
    },
    {
      "entropy": 1.8677813604474067,
      "epoch": 0.12731340657909931,
      "grad_norm": 8.423454284667969,
      "learning_rate": 7.0902393693334806e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8379834398627282,
      "num_tokens": 49357016.0,
      "step": 41070
    },
    {
      "entropy": 1.9531373485922814,
      "epoch": 0.127344405704149,
      "grad_norm": 10.435996055603027,
      "learning_rate": 7.089376316323171e-06,
      "loss": 0.5963,
      "mean_token_accuracy": 0.8160718679428101,
      "num_tokens": 49368452.0,
      "step": 41080
    },
    {
      "entropy": 1.9246175095438958,
      "epoch": 0.1273754048291987,
      "grad_norm": 8.71939468383789,
      "learning_rate": 7.088513578399207e-06,
      "loss": 0.562,
      "mean_token_accuracy": 0.8220181420445443,
      "num_tokens": 49380204.0,
      "step": 41090
    },
    {
      "entropy": 1.8995565429329873,
      "epoch": 0.1274064039542484,
      "grad_norm": 9.864094734191895,
      "learning_rate": 7.08765115536992e-06,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8301453128457069,
      "num_tokens": 49392226.0,
      "step": 41100
    },
    {
      "entropy": 1.900149242579937,
      "epoch": 0.1274374030792981,
      "grad_norm": 8.183351516723633,
      "learning_rate": 7.086789047043793e-06,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8305352702736855,
      "num_tokens": 49404110.0,
      "step": 41110
    },
    {
      "entropy": 1.9237231731414794,
      "epoch": 0.12746840220434777,
      "grad_norm": 8.36330509185791,
      "learning_rate": 7.08592725322948e-06,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8335365414619446,
      "num_tokens": 49416490.0,
      "step": 41120
    },
    {
      "entropy": 1.9189850777387618,
      "epoch": 0.12749940132939747,
      "grad_norm": 9.289202690124512,
      "learning_rate": 7.085065773735793e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8426647335290909,
      "num_tokens": 49427534.0,
      "step": 41130
    },
    {
      "entropy": 1.8757270485162736,
      "epoch": 0.12753040045444716,
      "grad_norm": 9.52793025970459,
      "learning_rate": 7.084204608371712e-06,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8288914114236832,
      "num_tokens": 49439317.0,
      "step": 41140
    },
    {
      "entropy": 1.8509258836507798,
      "epoch": 0.12756139957949686,
      "grad_norm": 5.029160976409912,
      "learning_rate": 7.083343756946375e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8297637715935707,
      "num_tokens": 49451446.0,
      "step": 41150
    },
    {
      "entropy": 1.9340157762169838,
      "epoch": 0.12759239870454656,
      "grad_norm": 11.50483226776123,
      "learning_rate": 7.082483219269084e-06,
      "loss": 0.5791,
      "mean_token_accuracy": 0.8197537034749984,
      "num_tokens": 49463109.0,
      "step": 41160
    },
    {
      "entropy": 1.8594611391425133,
      "epoch": 0.12762339782959625,
      "grad_norm": 7.6252899169921875,
      "learning_rate": 7.081622995149303e-06,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8369016021490097,
      "num_tokens": 49474727.0,
      "step": 41170
    },
    {
      "entropy": 1.800793182849884,
      "epoch": 0.12765439695464595,
      "grad_norm": 1.9972714185714722,
      "learning_rate": 7.080763084396659e-06,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8380374610424042,
      "num_tokens": 49488788.0,
      "step": 41180
    },
    {
      "entropy": 1.9101984828710556,
      "epoch": 0.12768539607969565,
      "grad_norm": 10.091278076171875,
      "learning_rate": 7.0799034868209375e-06,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8259019926190376,
      "num_tokens": 49501586.0,
      "step": 41190
    },
    {
      "entropy": 1.8499779477715492,
      "epoch": 0.12771639520474534,
      "grad_norm": 2.6960954666137695,
      "learning_rate": 7.079044202232089e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.833436419069767,
      "num_tokens": 49514195.0,
      "step": 41200
    },
    {
      "entropy": 1.825632943212986,
      "epoch": 0.12774739432979504,
      "grad_norm": 9.283682823181152,
      "learning_rate": 7.078185230440225e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8395808920264244,
      "num_tokens": 49525989.0,
      "step": 41210
    },
    {
      "entropy": 1.9196676731109619,
      "epoch": 0.12777839345484474,
      "grad_norm": 8.81406021118164,
      "learning_rate": 7.0773265712556175e-06,
      "loss": 0.6053,
      "mean_token_accuracy": 0.8251327946782112,
      "num_tokens": 49536614.0,
      "step": 41220
    },
    {
      "entropy": 1.8985184118151666,
      "epoch": 0.12780939257989443,
      "grad_norm": 9.597168922424316,
      "learning_rate": 7.076468224488697e-06,
      "loss": 0.5692,
      "mean_token_accuracy": 0.821132518351078,
      "num_tokens": 49548013.0,
      "step": 41230
    },
    {
      "entropy": 1.8932278618216514,
      "epoch": 0.12784039170494413,
      "grad_norm": 8.59653091430664,
      "learning_rate": 7.075610189950059e-06,
      "loss": 0.5927,
      "mean_token_accuracy": 0.817491953074932,
      "num_tokens": 49559982.0,
      "step": 41240
    },
    {
      "entropy": 1.9004715710878373,
      "epoch": 0.12787139082999383,
      "grad_norm": 9.525247573852539,
      "learning_rate": 7.074752467450462e-06,
      "loss": 0.5614,
      "mean_token_accuracy": 0.8346248850226402,
      "num_tokens": 49571257.0,
      "step": 41250
    },
    {
      "entropy": 1.9199392378330231,
      "epoch": 0.12790238995504352,
      "grad_norm": 9.400187492370605,
      "learning_rate": 7.073895056800815e-06,
      "loss": 0.5788,
      "mean_token_accuracy": 0.8198289066553116,
      "num_tokens": 49582933.0,
      "step": 41260
    },
    {
      "entropy": 1.9200360536575318,
      "epoch": 0.12793338908009322,
      "grad_norm": 8.804792404174805,
      "learning_rate": 7.0730379578121956e-06,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8220379948616028,
      "num_tokens": 49594138.0,
      "step": 41270
    },
    {
      "entropy": 1.9512885123491288,
      "epoch": 0.12796438820514291,
      "grad_norm": 10.201251983642578,
      "learning_rate": 7.0721811702958406e-06,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8327943772077561,
      "num_tokens": 49604773.0,
      "step": 41280
    },
    {
      "entropy": 1.9171435490250588,
      "epoch": 0.1279953873301926,
      "grad_norm": 10.569693565368652,
      "learning_rate": 7.071324694063147e-06,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8239866450428963,
      "num_tokens": 49615737.0,
      "step": 41290
    },
    {
      "entropy": 1.8590099811553955,
      "epoch": 0.1280263864552423,
      "grad_norm": 11.221964836120605,
      "learning_rate": 7.070468528925668e-06,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8409821361303329,
      "num_tokens": 49627714.0,
      "step": 41300
    },
    {
      "entropy": 1.842246988415718,
      "epoch": 0.128057385580292,
      "grad_norm": 3.050896406173706,
      "learning_rate": 7.06961267469512e-06,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8346633806824684,
      "num_tokens": 49640153.0,
      "step": 41310
    },
    {
      "entropy": 1.8636904895305633,
      "epoch": 0.1280883847053417,
      "grad_norm": 8.947188377380371,
      "learning_rate": 7.068757131183378e-06,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8275300249457359,
      "num_tokens": 49652387.0,
      "step": 41320
    },
    {
      "entropy": 1.8928270608186721,
      "epoch": 0.1281193838303914,
      "grad_norm": 9.10622501373291,
      "learning_rate": 7.067901898202475e-06,
      "loss": 0.5765,
      "mean_token_accuracy": 0.8179589688777924,
      "num_tokens": 49664065.0,
      "step": 41330
    },
    {
      "entropy": 1.9435312688350677,
      "epoch": 0.1281503829554411,
      "grad_norm": 8.725778579711914,
      "learning_rate": 7.067046975564605e-06,
      "loss": 0.6324,
      "mean_token_accuracy": 0.8186397060751915,
      "num_tokens": 49674905.0,
      "step": 41340
    },
    {
      "entropy": 1.9112671226263047,
      "epoch": 0.1281813820804908,
      "grad_norm": 8.562727928161621,
      "learning_rate": 7.066192363082123e-06,
      "loss": 0.5768,
      "mean_token_accuracy": 0.8293170660734177,
      "num_tokens": 49686743.0,
      "step": 41350
    },
    {
      "entropy": 1.8668039426207543,
      "epoch": 0.12821238120554046,
      "grad_norm": 8.329222679138184,
      "learning_rate": 7.0653380605675344e-06,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8435063764452935,
      "num_tokens": 49697557.0,
      "step": 41360
    },
    {
      "entropy": 1.8395963311195374,
      "epoch": 0.12824338033059016,
      "grad_norm": 2.403559684753418,
      "learning_rate": 7.064484067833515e-06,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8410920441150666,
      "num_tokens": 49710159.0,
      "step": 41370
    },
    {
      "entropy": 1.9411360412836074,
      "epoch": 0.12827437945563985,
      "grad_norm": 8.784658432006836,
      "learning_rate": 7.063630384692888e-06,
      "loss": 0.6002,
      "mean_token_accuracy": 0.824271696805954,
      "num_tokens": 49720451.0,
      "step": 41380
    },
    {
      "entropy": 1.8381685689091682,
      "epoch": 0.12830537858068955,
      "grad_norm": 7.8702497482299805,
      "learning_rate": 7.062777010958642e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8503654897212982,
      "num_tokens": 49732913.0,
      "step": 41390
    },
    {
      "entropy": 1.9066902339458465,
      "epoch": 0.12833637770573925,
      "grad_norm": 8.476966857910156,
      "learning_rate": 7.061923946443923e-06,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8252906069159508,
      "num_tokens": 49744489.0,
      "step": 41400
    },
    {
      "entropy": 1.8539624221622943,
      "epoch": 0.12836737683078894,
      "grad_norm": 9.746798515319824,
      "learning_rate": 7.061071190962031e-06,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8306064233183861,
      "num_tokens": 49757061.0,
      "step": 41410
    },
    {
      "entropy": 1.8506083533167839,
      "epoch": 0.12839837595583864,
      "grad_norm": 9.114572525024414,
      "learning_rate": 7.060218744326428e-06,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8507398188114166,
      "num_tokens": 49769261.0,
      "step": 41420
    },
    {
      "entropy": 1.8710011199116707,
      "epoch": 0.12842937508088834,
      "grad_norm": 4.284439563751221,
      "learning_rate": 7.059366606350731e-06,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8323528245091438,
      "num_tokens": 49781442.0,
      "step": 41430
    },
    {
      "entropy": 1.9052658289670945,
      "epoch": 0.12846037420593803,
      "grad_norm": 8.909741401672363,
      "learning_rate": 7.0585147768487165e-06,
      "loss": 0.6048,
      "mean_token_accuracy": 0.825349123775959,
      "num_tokens": 49794445.0,
      "step": 41440
    },
    {
      "entropy": 1.911386439204216,
      "epoch": 0.12849137333098773,
      "grad_norm": 9.72156810760498,
      "learning_rate": 7.057663255634316e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8314669832587243,
      "num_tokens": 49806414.0,
      "step": 41450
    },
    {
      "entropy": 1.8788209095597268,
      "epoch": 0.12852237245603743,
      "grad_norm": 4.481322288513184,
      "learning_rate": 7.056812042521619e-06,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8367927253246308,
      "num_tokens": 49818629.0,
      "step": 41460
    },
    {
      "entropy": 1.958992251753807,
      "epoch": 0.12855337158108712,
      "grad_norm": 9.640412330627441,
      "learning_rate": 7.0559611373248725e-06,
      "loss": 0.607,
      "mean_token_accuracy": 0.8203074246644974,
      "num_tokens": 49829898.0,
      "step": 41470
    },
    {
      "entropy": 1.7837151035666465,
      "epoch": 0.12858437070613682,
      "grad_norm": 3.7444610595703125,
      "learning_rate": 7.05511053985848e-06,
      "loss": 0.3599,
      "mean_token_accuracy": 0.8547704428434372,
      "num_tokens": 49844096.0,
      "step": 41480
    },
    {
      "entropy": 1.8662296697497367,
      "epoch": 0.12861536983118652,
      "grad_norm": 4.482416152954102,
      "learning_rate": 7.054260249937003e-06,
      "loss": 0.5229,
      "mean_token_accuracy": 0.8283546611666679,
      "num_tokens": 49856112.0,
      "step": 41490
    },
    {
      "entropy": 1.8327308684587478,
      "epoch": 0.1286463689562362,
      "grad_norm": 4.7443342208862305,
      "learning_rate": 7.053410267375156e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8486336767673492,
      "num_tokens": 49869424.0,
      "step": 41500
    },
    {
      "entropy": 1.7724727407097816,
      "epoch": 0.1286773680812859,
      "grad_norm": 9.06193733215332,
      "learning_rate": 7.052560591987811e-06,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8514471918344497,
      "num_tokens": 49882790.0,
      "step": 41510
    },
    {
      "entropy": 1.8597662687301635,
      "epoch": 0.1287083672063356,
      "grad_norm": 10.311935424804688,
      "learning_rate": 7.051711223589997e-06,
      "loss": 0.5304,
      "mean_token_accuracy": 0.83001659065485,
      "num_tokens": 49894927.0,
      "step": 41520
    },
    {
      "entropy": 1.8753248527646065,
      "epoch": 0.1287393663313853,
      "grad_norm": 9.605613708496094,
      "learning_rate": 7.050862161996901e-06,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8333491206169128,
      "num_tokens": 49906710.0,
      "step": 41530
    },
    {
      "entropy": 1.904360829293728,
      "epoch": 0.128770365456435,
      "grad_norm": 7.814454078674316,
      "learning_rate": 7.050013407023859e-06,
      "loss": 0.5386,
      "mean_token_accuracy": 0.8288786053657532,
      "num_tokens": 49918562.0,
      "step": 41540
    },
    {
      "entropy": 1.9048177361488343,
      "epoch": 0.1288013645814847,
      "grad_norm": 9.190512657165527,
      "learning_rate": 7.049164958486372e-06,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8335204407572746,
      "num_tokens": 49929488.0,
      "step": 41550
    },
    {
      "entropy": 1.8486354887485503,
      "epoch": 0.1288323637065344,
      "grad_norm": 9.387413024902344,
      "learning_rate": 7.048316816200086e-06,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8399450510740281,
      "num_tokens": 49941628.0,
      "step": 41560
    },
    {
      "entropy": 1.8960034802556038,
      "epoch": 0.1288633628315841,
      "grad_norm": 4.520693302154541,
      "learning_rate": 7.047468979980812e-06,
      "loss": 0.5923,
      "mean_token_accuracy": 0.8215907469391823,
      "num_tokens": 49953693.0,
      "step": 41570
    },
    {
      "entropy": 1.838585540652275,
      "epoch": 0.12889436195663379,
      "grad_norm": 9.408111572265625,
      "learning_rate": 7.046621449644507e-06,
      "loss": 0.5268,
      "mean_token_accuracy": 0.825011870265007,
      "num_tokens": 49964922.0,
      "step": 41580
    },
    {
      "entropy": 1.8944636061787605,
      "epoch": 0.12892536108168348,
      "grad_norm": 8.773103713989258,
      "learning_rate": 7.045774225007293e-06,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8297211557626725,
      "num_tokens": 49977343.0,
      "step": 41590
    },
    {
      "entropy": 1.8642171308398248,
      "epoch": 0.12895636020673318,
      "grad_norm": 9.104546546936035,
      "learning_rate": 7.044927305885436e-06,
      "loss": 0.5694,
      "mean_token_accuracy": 0.824115814268589,
      "num_tokens": 49990496.0,
      "step": 41600
    },
    {
      "entropy": 1.832181690633297,
      "epoch": 0.12898735933178285,
      "grad_norm": 4.260911464691162,
      "learning_rate": 7.044080692095364e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8464431285858154,
      "num_tokens": 50002745.0,
      "step": 41610
    },
    {
      "entropy": 1.878463228046894,
      "epoch": 0.12901835845683254,
      "grad_norm": 8.97909927368164,
      "learning_rate": 7.043234383453658e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8333162799477577,
      "num_tokens": 50014652.0,
      "step": 41620
    },
    {
      "entropy": 1.8109195098280906,
      "epoch": 0.12904935758188224,
      "grad_norm": 10.204434394836426,
      "learning_rate": 7.04238837977705e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8232007443904876,
      "num_tokens": 50026970.0,
      "step": 41630
    },
    {
      "entropy": 1.8154733762145043,
      "epoch": 0.12908035670693194,
      "grad_norm": 5.0115132331848145,
      "learning_rate": 7.041542680882431e-06,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8351962670683861,
      "num_tokens": 50039078.0,
      "step": 41640
    },
    {
      "entropy": 1.854199130833149,
      "epoch": 0.12911135583198163,
      "grad_norm": 10.00779914855957,
      "learning_rate": 7.04069728658684e-06,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8258049175143242,
      "num_tokens": 50050935.0,
      "step": 41650
    },
    {
      "entropy": 1.9253348022699357,
      "epoch": 0.12914235495703133,
      "grad_norm": 8.788981437683105,
      "learning_rate": 7.039852196707477e-06,
      "loss": 0.5403,
      "mean_token_accuracy": 0.8358196586370468,
      "num_tokens": 50062097.0,
      "step": 41660
    },
    {
      "entropy": 1.8359443858265876,
      "epoch": 0.12917335408208103,
      "grad_norm": 8.613965034484863,
      "learning_rate": 7.039007411061688e-06,
      "loss": 0.47,
      "mean_token_accuracy": 0.8417585626244545,
      "num_tokens": 50074289.0,
      "step": 41670
    },
    {
      "entropy": 1.7834791973233224,
      "epoch": 0.12920435320713072,
      "grad_norm": 5.170468807220459,
      "learning_rate": 7.038162929466977e-06,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8395596250891686,
      "num_tokens": 50087113.0,
      "step": 41680
    },
    {
      "entropy": 1.8834406480193138,
      "epoch": 0.12923535233218042,
      "grad_norm": 9.960625648498535,
      "learning_rate": 7.037318751741002e-06,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8236372783780098,
      "num_tokens": 50098957.0,
      "step": 41690
    },
    {
      "entropy": 1.8330654799938202,
      "epoch": 0.12926635145723012,
      "grad_norm": 8.917972564697266,
      "learning_rate": 7.036474877701568e-06,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8239554926753044,
      "num_tokens": 50111327.0,
      "step": 41700
    },
    {
      "entropy": 1.9040823325514793,
      "epoch": 0.1292973505822798,
      "grad_norm": 10.261785507202148,
      "learning_rate": 7.03563130716664e-06,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8225290685892105,
      "num_tokens": 50123086.0,
      "step": 41710
    },
    {
      "entropy": 1.91173397898674,
      "epoch": 0.1293283497073295,
      "grad_norm": 9.091620445251465,
      "learning_rate": 7.0347880399543345e-06,
      "loss": 0.5809,
      "mean_token_accuracy": 0.8207633405923843,
      "num_tokens": 50134801.0,
      "step": 41720
    },
    {
      "entropy": 1.8690212473273278,
      "epoch": 0.1293593488323792,
      "grad_norm": 8.664191246032715,
      "learning_rate": 7.0339450758829165e-06,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8263664215803146,
      "num_tokens": 50146507.0,
      "step": 41730
    },
    {
      "entropy": 1.8848260268568993,
      "epoch": 0.1293903479574289,
      "grad_norm": 10.184765815734863,
      "learning_rate": 7.033102414770806e-06,
      "loss": 0.5267,
      "mean_token_accuracy": 0.8280350670218468,
      "num_tokens": 50158176.0,
      "step": 41740
    },
    {
      "entropy": 1.78608690649271,
      "epoch": 0.1294213470824786,
      "grad_norm": 9.575325012207031,
      "learning_rate": 7.032260056436574e-06,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8456524163484573,
      "num_tokens": 50171381.0,
      "step": 41750
    },
    {
      "entropy": 1.9386123955249785,
      "epoch": 0.1294523462075283,
      "grad_norm": 10.201217651367188,
      "learning_rate": 7.031418000698947e-06,
      "loss": 0.5914,
      "mean_token_accuracy": 0.8301969483494759,
      "num_tokens": 50182781.0,
      "step": 41760
    },
    {
      "entropy": 1.847770369052887,
      "epoch": 0.129483345332578,
      "grad_norm": 4.556051254272461,
      "learning_rate": 7.0305762473768e-06,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8307551980018616,
      "num_tokens": 50195460.0,
      "step": 41770
    },
    {
      "entropy": 1.816901859641075,
      "epoch": 0.1295143444576277,
      "grad_norm": 9.289469718933105,
      "learning_rate": 7.0297347962891595e-06,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8283630818128586,
      "num_tokens": 50208808.0,
      "step": 41780
    },
    {
      "entropy": 1.929781760275364,
      "epoch": 0.12954534358267739,
      "grad_norm": 8.88871955871582,
      "learning_rate": 7.028893647255209e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8377395987510681,
      "num_tokens": 50220192.0,
      "step": 41790
    },
    {
      "entropy": 1.8346890568733216,
      "epoch": 0.12957634270772708,
      "grad_norm": 11.108233451843262,
      "learning_rate": 7.028052800094273e-06,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8360419407486915,
      "num_tokens": 50232642.0,
      "step": 41800
    },
    {
      "entropy": 1.762622408568859,
      "epoch": 0.12960734183277678,
      "grad_norm": 3.646545886993408,
      "learning_rate": 7.027212254625838e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8363216891884804,
      "num_tokens": 50245870.0,
      "step": 41810
    },
    {
      "entropy": 1.7347578413784503,
      "epoch": 0.12963834095782648,
      "grad_norm": 10.43207836151123,
      "learning_rate": 7.026372010669536e-06,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8484481438994408,
      "num_tokens": 50259093.0,
      "step": 41820
    },
    {
      "entropy": 1.8771490514278413,
      "epoch": 0.12966934008287617,
      "grad_norm": 10.095263481140137,
      "learning_rate": 7.025532068045149e-06,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8258682683110237,
      "num_tokens": 50270396.0,
      "step": 41830
    },
    {
      "entropy": 1.8018823832273483,
      "epoch": 0.12970033920792587,
      "grad_norm": 10.791528701782227,
      "learning_rate": 7.024692426572615e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8406801581382751,
      "num_tokens": 50283135.0,
      "step": 41840
    },
    {
      "entropy": 1.8664998412132263,
      "epoch": 0.12973133833297557,
      "grad_norm": 11.595536231994629,
      "learning_rate": 7.023853086072019e-06,
      "loss": 0.554,
      "mean_token_accuracy": 0.820669724047184,
      "num_tokens": 50294875.0,
      "step": 41850
    },
    {
      "entropy": 1.8681924358010291,
      "epoch": 0.12976233745802523,
      "grad_norm": 9.37474536895752,
      "learning_rate": 7.0230140463635955e-06,
      "loss": 0.5866,
      "mean_token_accuracy": 0.8231520146131516,
      "num_tokens": 50307197.0,
      "step": 41860
    },
    {
      "entropy": 1.8410033360123634,
      "epoch": 0.12979333658307493,
      "grad_norm": 4.718626022338867,
      "learning_rate": 7.022175307267729e-06,
      "loss": 0.467,
      "mean_token_accuracy": 0.8391537711024284,
      "num_tokens": 50319844.0,
      "step": 41870
    },
    {
      "entropy": 1.9163996964693069,
      "epoch": 0.12982433570812463,
      "grad_norm": 13.432290077209473,
      "learning_rate": 7.021336868604959e-06,
      "loss": 0.6464,
      "mean_token_accuracy": 0.8215458050370217,
      "num_tokens": 50331051.0,
      "step": 41880
    },
    {
      "entropy": 1.8775553047657012,
      "epoch": 0.12985533483317432,
      "grad_norm": 9.11666488647461,
      "learning_rate": 7.0204987301959715e-06,
      "loss": 0.5546,
      "mean_token_accuracy": 0.8308037981390953,
      "num_tokens": 50342313.0,
      "step": 41890
    },
    {
      "entropy": 1.7834023706614972,
      "epoch": 0.12988633395822402,
      "grad_norm": 5.124344825744629,
      "learning_rate": 7.019660891861601e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8429507181048393,
      "num_tokens": 50356360.0,
      "step": 41900
    },
    {
      "entropy": 1.9084928244352342,
      "epoch": 0.12991733308327372,
      "grad_norm": 8.772656440734863,
      "learning_rate": 7.018823353422832e-06,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8257664754986763,
      "num_tokens": 50367277.0,
      "step": 41910
    },
    {
      "entropy": 1.8719779431819916,
      "epoch": 0.1299483322083234,
      "grad_norm": 8.429474830627441,
      "learning_rate": 7.017986114700802e-06,
      "loss": 0.501,
      "mean_token_accuracy": 0.8469484955072403,
      "num_tokens": 50378849.0,
      "step": 41920
    },
    {
      "entropy": 1.8775894209742545,
      "epoch": 0.1299793313333731,
      "grad_norm": 8.218339920043945,
      "learning_rate": 7.0171491755167954e-06,
      "loss": 0.5791,
      "mean_token_accuracy": 0.813252392411232,
      "num_tokens": 50390771.0,
      "step": 41930
    },
    {
      "entropy": 1.8514499336481094,
      "epoch": 0.1300103304584228,
      "grad_norm": 8.702057838439941,
      "learning_rate": 7.016312535692245e-06,
      "loss": 0.5392,
      "mean_token_accuracy": 0.837669375538826,
      "num_tokens": 50402598.0,
      "step": 41940
    },
    {
      "entropy": 1.7951854154467584,
      "epoch": 0.1300413295834725,
      "grad_norm": 9.224007606506348,
      "learning_rate": 7.0154761950487325e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8424011334776879,
      "num_tokens": 50415365.0,
      "step": 41950
    },
    {
      "entropy": 1.941623505949974,
      "epoch": 0.1300723287085222,
      "grad_norm": 10.595728874206543,
      "learning_rate": 7.01464015340799e-06,
      "loss": 0.6463,
      "mean_token_accuracy": 0.8008893147110939,
      "num_tokens": 50426866.0,
      "step": 41960
    },
    {
      "entropy": 1.8391387566924096,
      "epoch": 0.1301033278335719,
      "grad_norm": 9.028034210205078,
      "learning_rate": 7.0138044105918975e-06,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8306760504841805,
      "num_tokens": 50440268.0,
      "step": 41970
    },
    {
      "entropy": 1.7658006258308887,
      "epoch": 0.1301343269586216,
      "grad_norm": 2.6869993209838867,
      "learning_rate": 7.0129689664224855e-06,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8479226931929589,
      "num_tokens": 50455088.0,
      "step": 41980
    },
    {
      "entropy": 1.8967247605323792,
      "epoch": 0.1301653260836713,
      "grad_norm": 7.826026916503906,
      "learning_rate": 7.012133820721929e-06,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8311007678508758,
      "num_tokens": 50466652.0,
      "step": 41990
    },
    {
      "entropy": 1.958921417593956,
      "epoch": 0.130196325208721,
      "grad_norm": 8.257652282714844,
      "learning_rate": 7.011298973312554e-06,
      "loss": 0.5903,
      "mean_token_accuracy": 0.8199125394225121,
      "num_tokens": 50477973.0,
      "step": 42000
    },
    {
      "entropy": 1.8253989905118941,
      "epoch": 0.13022732433377068,
      "grad_norm": 3.0014355182647705,
      "learning_rate": 7.0104644240168294e-06,
      "loss": 0.5041,
      "mean_token_accuracy": 0.835718783736229,
      "num_tokens": 50490889.0,
      "step": 42010
    },
    {
      "entropy": 1.9271552190184593,
      "epoch": 0.13025832345882038,
      "grad_norm": 10.399243354797363,
      "learning_rate": 7.0096301726573835e-06,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8131484746932983,
      "num_tokens": 50502026.0,
      "step": 42020
    },
    {
      "entropy": 1.9419305875897408,
      "epoch": 0.13028932258387008,
      "grad_norm": 7.492198467254639,
      "learning_rate": 7.008796219056981e-06,
      "loss": 0.6331,
      "mean_token_accuracy": 0.82049780189991,
      "num_tokens": 50513287.0,
      "step": 42030
    },
    {
      "entropy": 1.9706764385104178,
      "epoch": 0.13032032170891977,
      "grad_norm": 10.456650733947754,
      "learning_rate": 7.00796256303854e-06,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8240112096071244,
      "num_tokens": 50524654.0,
      "step": 42040
    },
    {
      "entropy": 1.8961515158414841,
      "epoch": 0.13035132083396947,
      "grad_norm": 10.818836212158203,
      "learning_rate": 7.0071292044251215e-06,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8393558323383331,
      "num_tokens": 50537697.0,
      "step": 42050
    },
    {
      "entropy": 1.9397668451070786,
      "epoch": 0.13038231995901917,
      "grad_norm": 7.472477912902832,
      "learning_rate": 7.006296143039939e-06,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8343762636184693,
      "num_tokens": 50549295.0,
      "step": 42060
    },
    {
      "entropy": 1.8642112627625465,
      "epoch": 0.13041331908406886,
      "grad_norm": 2.947805643081665,
      "learning_rate": 7.00546337870635e-06,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8320413127541542,
      "num_tokens": 50561911.0,
      "step": 42070
    },
    {
      "entropy": 1.8363241106271744,
      "epoch": 0.13044431820911856,
      "grad_norm": 4.179972171783447,
      "learning_rate": 7.0046309112478594e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8355718359351159,
      "num_tokens": 50574636.0,
      "step": 42080
    },
    {
      "entropy": 1.8695838272571563,
      "epoch": 0.13047531733416826,
      "grad_norm": 10.452752113342285,
      "learning_rate": 7.003798740488118e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8281556889414787,
      "num_tokens": 50586479.0,
      "step": 42090
    },
    {
      "entropy": 1.8435761332511902,
      "epoch": 0.13050631645921792,
      "grad_norm": 9.886906623840332,
      "learning_rate": 7.0029668662509255e-06,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8295658230781555,
      "num_tokens": 50599307.0,
      "step": 42100
    },
    {
      "entropy": 1.8621219590306282,
      "epoch": 0.13053731558426762,
      "grad_norm": 7.598883628845215,
      "learning_rate": 7.002135288360228e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8365325689315796,
      "num_tokens": 50611631.0,
      "step": 42110
    },
    {
      "entropy": 1.926572097837925,
      "epoch": 0.13056831470931732,
      "grad_norm": 7.883036136627197,
      "learning_rate": 7.0013040066401135e-06,
      "loss": 0.5725,
      "mean_token_accuracy": 0.8214625731110573,
      "num_tokens": 50622981.0,
      "step": 42120
    },
    {
      "entropy": 1.9384382754564284,
      "epoch": 0.13059931383436701,
      "grad_norm": 8.586524963378906,
      "learning_rate": 7.000473020914823e-06,
      "loss": 0.585,
      "mean_token_accuracy": 0.8287076443433762,
      "num_tokens": 50633779.0,
      "step": 42130
    },
    {
      "entropy": 1.9370570868253707,
      "epoch": 0.1306303129594167,
      "grad_norm": 10.588152885437012,
      "learning_rate": 6.999642331008736e-06,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8158860892057419,
      "num_tokens": 50645735.0,
      "step": 42140
    },
    {
      "entropy": 1.8363999262452126,
      "epoch": 0.1306613120844664,
      "grad_norm": 4.617247581481934,
      "learning_rate": 6.998811936746385e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8244207426905632,
      "num_tokens": 50659623.0,
      "step": 42150
    },
    {
      "entropy": 1.9262761980295182,
      "epoch": 0.1306923112095161,
      "grad_norm": 11.821525573730469,
      "learning_rate": 6.9979818379524435e-06,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8174602270126343,
      "num_tokens": 50672505.0,
      "step": 42160
    },
    {
      "entropy": 1.8478658609092236,
      "epoch": 0.1307233103345658,
      "grad_norm": 5.417337894439697,
      "learning_rate": 6.997152034451732e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8429075211286545,
      "num_tokens": 50686049.0,
      "step": 42170
    },
    {
      "entropy": 1.897514969110489,
      "epoch": 0.1307543094596155,
      "grad_norm": 7.976996421813965,
      "learning_rate": 6.9963225260692145e-06,
      "loss": 0.5191,
      "mean_token_accuracy": 0.832906337082386,
      "num_tokens": 50698264.0,
      "step": 42180
    },
    {
      "entropy": 1.8931497901678085,
      "epoch": 0.1307853085846652,
      "grad_norm": 4.187560081481934,
      "learning_rate": 6.995493312630006e-06,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8311450779438019,
      "num_tokens": 50710916.0,
      "step": 42190
    },
    {
      "entropy": 1.8941505983471871,
      "epoch": 0.1308163077097149,
      "grad_norm": 8.752896308898926,
      "learning_rate": 6.9946643939593606e-06,
      "loss": 0.5434,
      "mean_token_accuracy": 0.8309932291507721,
      "num_tokens": 50722894.0,
      "step": 42200
    },
    {
      "entropy": 1.897823777794838,
      "epoch": 0.1308473068347646,
      "grad_norm": 8.950542449951172,
      "learning_rate": 6.993835769882677e-06,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8307448908686638,
      "num_tokens": 50735048.0,
      "step": 42210
    },
    {
      "entropy": 2.0057034313678743,
      "epoch": 0.13087830595981428,
      "grad_norm": 8.577478408813477,
      "learning_rate": 6.993007440225504e-06,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8248661920428276,
      "num_tokens": 50745880.0,
      "step": 42220
    },
    {
      "entropy": 1.933904617279768,
      "epoch": 0.13090930508486398,
      "grad_norm": 7.418628692626953,
      "learning_rate": 6.99217940481353e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8272743076086044,
      "num_tokens": 50757689.0,
      "step": 42230
    },
    {
      "entropy": 1.8607298329472541,
      "epoch": 0.13094030420991368,
      "grad_norm": 3.3645823001861572,
      "learning_rate": 6.991351663472591e-06,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8338559105992317,
      "num_tokens": 50770252.0,
      "step": 42240
    },
    {
      "entropy": 1.7977669216692447,
      "epoch": 0.13097130333496337,
      "grad_norm": 8.085532188415527,
      "learning_rate": 6.990524216028667e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8425278753042221,
      "num_tokens": 50783859.0,
      "step": 42250
    },
    {
      "entropy": 1.9316506549715995,
      "epoch": 0.13100230246001307,
      "grad_norm": 9.128771781921387,
      "learning_rate": 6.989697062307879e-06,
      "loss": 0.553,
      "mean_token_accuracy": 0.8342802032828331,
      "num_tokens": 50795907.0,
      "step": 42260
    },
    {
      "entropy": 2.010502940416336,
      "epoch": 0.13103330158506277,
      "grad_norm": 9.819306373596191,
      "learning_rate": 6.988870202136493e-06,
      "loss": 0.6157,
      "mean_token_accuracy": 0.8241182819008828,
      "num_tokens": 50806819.0,
      "step": 42270
    },
    {
      "entropy": 1.879169872403145,
      "epoch": 0.13106430071011246,
      "grad_norm": 8.068976402282715,
      "learning_rate": 6.988043635340924e-06,
      "loss": 0.474,
      "mean_token_accuracy": 0.8331845104694366,
      "num_tokens": 50820181.0,
      "step": 42280
    },
    {
      "entropy": 1.9254466131329537,
      "epoch": 0.13109529983516216,
      "grad_norm": 8.121038436889648,
      "learning_rate": 6.987217361747725e-06,
      "loss": 0.5318,
      "mean_token_accuracy": 0.830107967555523,
      "num_tokens": 50831403.0,
      "step": 42290
    },
    {
      "entropy": 1.9769433185458183,
      "epoch": 0.13112629896021186,
      "grad_norm": 10.498136520385742,
      "learning_rate": 6.986391381183594e-06,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8088495776057243,
      "num_tokens": 50843327.0,
      "step": 42300
    },
    {
      "entropy": 1.8907544031739234,
      "epoch": 0.13115729808526155,
      "grad_norm": 10.089003562927246,
      "learning_rate": 6.98556569347537e-06,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8322427123785019,
      "num_tokens": 50856211.0,
      "step": 42310
    },
    {
      "entropy": 1.9495109453797341,
      "epoch": 0.13118829721031125,
      "grad_norm": 9.126409530639648,
      "learning_rate": 6.984740298450043e-06,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8336078703403473,
      "num_tokens": 50867480.0,
      "step": 42320
    },
    {
      "entropy": 1.8435195326805114,
      "epoch": 0.13121929633536095,
      "grad_norm": 6.529050827026367,
      "learning_rate": 6.983915195934738e-06,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8262422546744347,
      "num_tokens": 50880593.0,
      "step": 42330
    },
    {
      "entropy": 1.9224420145154,
      "epoch": 0.13125029546041064,
      "grad_norm": 10.954397201538086,
      "learning_rate": 6.983090385756723e-06,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8218831300735474,
      "num_tokens": 50891387.0,
      "step": 42340
    },
    {
      "entropy": 1.8449978575110435,
      "epoch": 0.1312812945854603,
      "grad_norm": 8.803427696228027,
      "learning_rate": 6.982265867743417e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8435740604996681,
      "num_tokens": 50903726.0,
      "step": 42350
    },
    {
      "entropy": 1.8841196730732919,
      "epoch": 0.13131229371051,
      "grad_norm": 9.251840591430664,
      "learning_rate": 6.981441641722373e-06,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8290949910879135,
      "num_tokens": 50916111.0,
      "step": 42360
    },
    {
      "entropy": 1.927403984963894,
      "epoch": 0.1313432928355597,
      "grad_norm": 8.574084281921387,
      "learning_rate": 6.98061770752129e-06,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8280455321073532,
      "num_tokens": 50928284.0,
      "step": 42370
    },
    {
      "entropy": 1.9771150171756744,
      "epoch": 0.1313742919606094,
      "grad_norm": 8.762701988220215,
      "learning_rate": 6.9797940649680106e-06,
      "loss": 0.6193,
      "mean_token_accuracy": 0.8151744335889817,
      "num_tokens": 50939233.0,
      "step": 42380
    },
    {
      "entropy": 1.8265860572457313,
      "epoch": 0.1314052910856591,
      "grad_norm": 7.463681697845459,
      "learning_rate": 6.978970713890515e-06,
      "loss": 0.555,
      "mean_token_accuracy": 0.8328731596469879,
      "num_tokens": 50953036.0,
      "step": 42390
    },
    {
      "entropy": 1.8653569996356965,
      "epoch": 0.1314362902107088,
      "grad_norm": 8.80778694152832,
      "learning_rate": 6.978147654116929e-06,
      "loss": 0.47,
      "mean_token_accuracy": 0.8439178183674813,
      "num_tokens": 50964848.0,
      "step": 42400
    },
    {
      "entropy": 1.939113649725914,
      "epoch": 0.1314672893357585,
      "grad_norm": 10.408567428588867,
      "learning_rate": 6.977324885475521e-06,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8219144955277443,
      "num_tokens": 50976553.0,
      "step": 42410
    },
    {
      "entropy": 1.862153060734272,
      "epoch": 0.1314982884608082,
      "grad_norm": 8.101511001586914,
      "learning_rate": 6.976502407794701e-06,
      "loss": 0.485,
      "mean_token_accuracy": 0.835495936870575,
      "num_tokens": 50988487.0,
      "step": 42420
    },
    {
      "entropy": 1.9604466244578362,
      "epoch": 0.13152928758585788,
      "grad_norm": 9.159782409667969,
      "learning_rate": 6.975680220903015e-06,
      "loss": 0.5845,
      "mean_token_accuracy": 0.8208717614412308,
      "num_tokens": 51000283.0,
      "step": 42430
    },
    {
      "entropy": 1.9139268666505813,
      "epoch": 0.13156028671090758,
      "grad_norm": 8.362812042236328,
      "learning_rate": 6.974858324629158e-06,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8323205769062042,
      "num_tokens": 51012209.0,
      "step": 42440
    },
    {
      "entropy": 1.9230769395828247,
      "epoch": 0.13159128583595728,
      "grad_norm": 8.45923900604248,
      "learning_rate": 6.97403671880196e-06,
      "loss": 0.572,
      "mean_token_accuracy": 0.8329043105244637,
      "num_tokens": 51023308.0,
      "step": 42450
    },
    {
      "entropy": 1.9842663645744323,
      "epoch": 0.13162228496100697,
      "grad_norm": 8.042434692382812,
      "learning_rate": 6.973215403250397e-06,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8260912299156189,
      "num_tokens": 51033767.0,
      "step": 42460
    },
    {
      "entropy": 1.9136120170354842,
      "epoch": 0.13165328408605667,
      "grad_norm": 8.325801849365234,
      "learning_rate": 6.972394377803584e-06,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8208369135856628,
      "num_tokens": 51045947.0,
      "step": 42470
    },
    {
      "entropy": 1.8511994555592537,
      "epoch": 0.13168428321110637,
      "grad_norm": 3.737189531326294,
      "learning_rate": 6.9715736422907764e-06,
      "loss": 0.4705,
      "mean_token_accuracy": 0.8399424910545349,
      "num_tokens": 51057715.0,
      "step": 42480
    },
    {
      "entropy": 1.8442467346787452,
      "epoch": 0.13171528233615606,
      "grad_norm": 4.051930904388428,
      "learning_rate": 6.9707531965413695e-06,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8524442002177238,
      "num_tokens": 51068742.0,
      "step": 42490
    },
    {
      "entropy": 1.8845135629177094,
      "epoch": 0.13174628146120576,
      "grad_norm": 9.697474479675293,
      "learning_rate": 6.969933040384902e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8385147228837013,
      "num_tokens": 51080521.0,
      "step": 42500
    },
    {
      "entropy": 1.7795793518424035,
      "epoch": 0.13177728058625546,
      "grad_norm": 9.526708602905273,
      "learning_rate": 6.96911317365105e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8431052088737487,
      "num_tokens": 51093716.0,
      "step": 42510
    },
    {
      "entropy": 1.9105603873729706,
      "epoch": 0.13180827971130515,
      "grad_norm": 8.357869148254395,
      "learning_rate": 6.968293596169631e-06,
      "loss": 0.5976,
      "mean_token_accuracy": 0.8246324315667153,
      "num_tokens": 51105239.0,
      "step": 42520
    },
    {
      "entropy": 1.9591112911701203,
      "epoch": 0.13183927883635485,
      "grad_norm": 9.68211841583252,
      "learning_rate": 6.967474307770603e-06,
      "loss": 0.6515,
      "mean_token_accuracy": 0.8138914480805397,
      "num_tokens": 51116502.0,
      "step": 42530
    },
    {
      "entropy": 1.9038333266973495,
      "epoch": 0.13187027796140455,
      "grad_norm": 8.462784767150879,
      "learning_rate": 6.966655308284064e-06,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8396757513284683,
      "num_tokens": 51128381.0,
      "step": 42540
    },
    {
      "entropy": 1.8404983699321746,
      "epoch": 0.13190127708645424,
      "grad_norm": 8.014533042907715,
      "learning_rate": 6.965836597540249e-06,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8473035827279091,
      "num_tokens": 51140787.0,
      "step": 42550
    },
    {
      "entropy": 1.9219214513897895,
      "epoch": 0.13193227621150394,
      "grad_norm": 8.607987403869629,
      "learning_rate": 6.965018175369538e-06,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8373000919818878,
      "num_tokens": 51152611.0,
      "step": 42560
    },
    {
      "entropy": 1.8740851491689683,
      "epoch": 0.13196327533655364,
      "grad_norm": 8.044096946716309,
      "learning_rate": 6.9642000416024435e-06,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8397703349590302,
      "num_tokens": 51164281.0,
      "step": 42570
    },
    {
      "entropy": 1.8905795052647592,
      "epoch": 0.13199427446160333,
      "grad_norm": 9.585006713867188,
      "learning_rate": 6.963382196069625e-06,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8185367554426193,
      "num_tokens": 51175917.0,
      "step": 42580
    },
    {
      "entropy": 1.814333714544773,
      "epoch": 0.13202527358665303,
      "grad_norm": 10.914127349853516,
      "learning_rate": 6.962564638601874e-06,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8299461394548416,
      "num_tokens": 51188596.0,
      "step": 42590
    },
    {
      "entropy": 1.8761808335781098,
      "epoch": 0.1320562727117027,
      "grad_norm": 8.772909164428711,
      "learning_rate": 6.961747369030127e-06,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8223336800932884,
      "num_tokens": 51200294.0,
      "step": 42600
    },
    {
      "entropy": 1.8656433135271073,
      "epoch": 0.1320872718367524,
      "grad_norm": 4.012499809265137,
      "learning_rate": 6.960930387185456e-06,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8211655914783478,
      "num_tokens": 51212299.0,
      "step": 42610
    },
    {
      "entropy": 1.8340272575616836,
      "epoch": 0.1321182709618021,
      "grad_norm": 5.148853302001953,
      "learning_rate": 6.960113692899071e-06,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8375645697116851,
      "num_tokens": 51224694.0,
      "step": 42620
    },
    {
      "entropy": 1.8271689996123315,
      "epoch": 0.1321492700868518,
      "grad_norm": 8.511850357055664,
      "learning_rate": 6.9592972860023235e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8452630266547203,
      "num_tokens": 51237562.0,
      "step": 42630
    },
    {
      "entropy": 1.9014787808060647,
      "epoch": 0.13218026921190149,
      "grad_norm": 4.459075927734375,
      "learning_rate": 6.9584811663267015e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8347061321139335,
      "num_tokens": 51248656.0,
      "step": 42640
    },
    {
      "entropy": 1.9535549938678742,
      "epoch": 0.13221126833695118,
      "grad_norm": 4.374874114990234,
      "learning_rate": 6.9576653337038325e-06,
      "loss": 0.6071,
      "mean_token_accuracy": 0.8113219693303109,
      "num_tokens": 51259712.0,
      "step": 42650
    },
    {
      "entropy": 1.8858298167586327,
      "epoch": 0.13224226746200088,
      "grad_norm": 11.858930587768555,
      "learning_rate": 6.956849787965481e-06,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8303458750247955,
      "num_tokens": 51272446.0,
      "step": 42660
    },
    {
      "entropy": 1.8564813017845154,
      "epoch": 0.13227326658705058,
      "grad_norm": 3.350623846054077,
      "learning_rate": 6.956034528943548e-06,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8342030212283135,
      "num_tokens": 51285301.0,
      "step": 42670
    },
    {
      "entropy": 1.849312388896942,
      "epoch": 0.13230426571210027,
      "grad_norm": 9.046114921569824,
      "learning_rate": 6.955219556470077e-06,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8443035989999771,
      "num_tokens": 51297696.0,
      "step": 42680
    },
    {
      "entropy": 1.9661724478006364,
      "epoch": 0.13233526483714997,
      "grad_norm": 9.500844955444336,
      "learning_rate": 6.954404870377246e-06,
      "loss": 0.6121,
      "mean_token_accuracy": 0.8136408120393753,
      "num_tokens": 51307961.0,
      "step": 42690
    },
    {
      "entropy": 1.939977452158928,
      "epoch": 0.13236626396219967,
      "grad_norm": 8.7796049118042,
      "learning_rate": 6.953590470497371e-06,
      "loss": 0.6126,
      "mean_token_accuracy": 0.8261479705572128,
      "num_tokens": 51318495.0,
      "step": 42700
    },
    {
      "entropy": 1.840474684536457,
      "epoch": 0.13239726308724936,
      "grad_norm": 11.025080680847168,
      "learning_rate": 6.952776356662905e-06,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8339684426784515,
      "num_tokens": 51330208.0,
      "step": 42710
    },
    {
      "entropy": 1.947993564605713,
      "epoch": 0.13242826221229906,
      "grad_norm": 9.111898422241211,
      "learning_rate": 6.9519625287064375e-06,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8302145257592202,
      "num_tokens": 51340861.0,
      "step": 42720
    },
    {
      "entropy": 1.8799726784229278,
      "epoch": 0.13245926133734875,
      "grad_norm": 9.264261245727539,
      "learning_rate": 6.951148986460699e-06,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8337329775094986,
      "num_tokens": 51353102.0,
      "step": 42730
    },
    {
      "entropy": 1.959149533510208,
      "epoch": 0.13249026046239845,
      "grad_norm": 10.266985893249512,
      "learning_rate": 6.950335729758554e-06,
      "loss": 0.6076,
      "mean_token_accuracy": 0.821322014927864,
      "num_tokens": 51364079.0,
      "step": 42740
    },
    {
      "entropy": 1.937872663140297,
      "epoch": 0.13252125958744815,
      "grad_norm": 9.663759231567383,
      "learning_rate": 6.949522758433003e-06,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8321762755513191,
      "num_tokens": 51375560.0,
      "step": 42750
    },
    {
      "entropy": 1.8674385949969292,
      "epoch": 0.13255225871249784,
      "grad_norm": 7.8949432373046875,
      "learning_rate": 6.948710072317184e-06,
      "loss": 0.4933,
      "mean_token_accuracy": 0.837440250813961,
      "num_tokens": 51388270.0,
      "step": 42760
    },
    {
      "entropy": 1.9320979446172715,
      "epoch": 0.13258325783754754,
      "grad_norm": 8.221677780151367,
      "learning_rate": 6.9478976712443755e-06,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8377137213945389,
      "num_tokens": 51399663.0,
      "step": 42770
    },
    {
      "entropy": 1.8442893743515014,
      "epoch": 0.13261425696259724,
      "grad_norm": 7.525363445281982,
      "learning_rate": 6.947085555047985e-06,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8382806748151779,
      "num_tokens": 51411959.0,
      "step": 42780
    },
    {
      "entropy": 1.8832678958773612,
      "epoch": 0.13264525608764693,
      "grad_norm": 8.396281242370605,
      "learning_rate": 6.946273723561562e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8351016476750374,
      "num_tokens": 51423692.0,
      "step": 42790
    },
    {
      "entropy": 1.8216581985354423,
      "epoch": 0.13267625521269663,
      "grad_norm": 8.608697891235352,
      "learning_rate": 6.9454621766187904e-06,
      "loss": 0.481,
      "mean_token_accuracy": 0.8447302252054214,
      "num_tokens": 51436099.0,
      "step": 42800
    },
    {
      "entropy": 1.9112284123897552,
      "epoch": 0.13270725433774633,
      "grad_norm": 5.335425853729248,
      "learning_rate": 6.944650914053489e-06,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8203225836157799,
      "num_tokens": 51448674.0,
      "step": 42810
    },
    {
      "entropy": 1.9009429544210434,
      "epoch": 0.13273825346279602,
      "grad_norm": 8.94002914428711,
      "learning_rate": 6.943839935699615e-06,
      "loss": 0.5229,
      "mean_token_accuracy": 0.8352271884679794,
      "num_tokens": 51460644.0,
      "step": 42820
    },
    {
      "entropy": 1.8937845051288604,
      "epoch": 0.13276925258784572,
      "grad_norm": 4.647216320037842,
      "learning_rate": 6.943029241391259e-06,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8236138209700584,
      "num_tokens": 51473179.0,
      "step": 42830
    },
    {
      "entropy": 1.8776622116565704,
      "epoch": 0.1328002517128954,
      "grad_norm": 9.4772367477417,
      "learning_rate": 6.942218830962648e-06,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8267105832695961,
      "num_tokens": 51486098.0,
      "step": 42840
    },
    {
      "entropy": 1.7324663981795312,
      "epoch": 0.1328312508379451,
      "grad_norm": 2.6686737537384033,
      "learning_rate": 6.941408704248144e-06,
      "loss": 0.3974,
      "mean_token_accuracy": 0.8572231993079186,
      "num_tokens": 51499848.0,
      "step": 42850
    },
    {
      "entropy": 1.8544006049633026,
      "epoch": 0.13286224996299478,
      "grad_norm": 9.000062942504883,
      "learning_rate": 6.940598861082245e-06,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8476372644305229,
      "num_tokens": 51512048.0,
      "step": 42860
    },
    {
      "entropy": 1.9546643912792205,
      "epoch": 0.13289324908804448,
      "grad_norm": 11.137999534606934,
      "learning_rate": 6.939789301299585e-06,
      "loss": 0.6104,
      "mean_token_accuracy": 0.8146396011114121,
      "num_tokens": 51523110.0,
      "step": 42870
    },
    {
      "entropy": 1.9276799812912941,
      "epoch": 0.13292424821309418,
      "grad_norm": 10.72694206237793,
      "learning_rate": 6.938980024734927e-06,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8259019777178764,
      "num_tokens": 51534581.0,
      "step": 42880
    },
    {
      "entropy": 1.7917723521590232,
      "epoch": 0.13295524733814387,
      "grad_norm": 9.843888282775879,
      "learning_rate": 6.938171031223178e-06,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8381556645035744,
      "num_tokens": 51548074.0,
      "step": 42890
    },
    {
      "entropy": 1.9410393938422204,
      "epoch": 0.13298624646319357,
      "grad_norm": 10.020930290222168,
      "learning_rate": 6.937362320599377e-06,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8282965987920761,
      "num_tokens": 51559680.0,
      "step": 42900
    },
    {
      "entropy": 1.92443605363369,
      "epoch": 0.13301724558824327,
      "grad_norm": 10.336004257202148,
      "learning_rate": 6.936553892698692e-06,
      "loss": 0.557,
      "mean_token_accuracy": 0.8238545969128609,
      "num_tokens": 51571461.0,
      "step": 42910
    },
    {
      "entropy": 1.833638320863247,
      "epoch": 0.13304824471329296,
      "grad_norm": 10.432703971862793,
      "learning_rate": 6.935745747356429e-06,
      "loss": 0.4945,
      "mean_token_accuracy": 0.836958509683609,
      "num_tokens": 51583715.0,
      "step": 42920
    },
    {
      "entropy": 1.9126337066292762,
      "epoch": 0.13307924383834266,
      "grad_norm": 4.075279712677002,
      "learning_rate": 6.934937884408032e-06,
      "loss": 0.511,
      "mean_token_accuracy": 0.8253563806414604,
      "num_tokens": 51595798.0,
      "step": 42930
    },
    {
      "entropy": 1.9492529153823852,
      "epoch": 0.13311024296339236,
      "grad_norm": 8.318151473999023,
      "learning_rate": 6.934130303689072e-06,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8317864894866943,
      "num_tokens": 51606958.0,
      "step": 42940
    },
    {
      "entropy": 1.9043660476803779,
      "epoch": 0.13314124208844205,
      "grad_norm": 4.388843059539795,
      "learning_rate": 6.93332300503526e-06,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8340587362647056,
      "num_tokens": 51619061.0,
      "step": 42950
    },
    {
      "entropy": 1.874220597743988,
      "epoch": 0.13317224121349175,
      "grad_norm": 3.5497045516967773,
      "learning_rate": 6.932515988282438e-06,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8257475927472114,
      "num_tokens": 51630509.0,
      "step": 42960
    },
    {
      "entropy": 1.8349529922008514,
      "epoch": 0.13320324033854145,
      "grad_norm": 8.229619979858398,
      "learning_rate": 6.931709253266582e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.836476719379425,
      "num_tokens": 51642250.0,
      "step": 42970
    },
    {
      "entropy": 1.9145320609211922,
      "epoch": 0.13323423946359114,
      "grad_norm": 8.655306816101074,
      "learning_rate": 6.930902799823801e-06,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8278056040406228,
      "num_tokens": 51654224.0,
      "step": 42980
    },
    {
      "entropy": 1.9441554173827171,
      "epoch": 0.13326523858864084,
      "grad_norm": 10.367897987365723,
      "learning_rate": 6.9300966277903415e-06,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8317381590604782,
      "num_tokens": 51665436.0,
      "step": 42990
    },
    {
      "entropy": 1.8716817542910575,
      "epoch": 0.13329623771369054,
      "grad_norm": 8.810930252075195,
      "learning_rate": 6.929290737002579e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8309359416365624,
      "num_tokens": 51678512.0,
      "step": 43000
    },
    {
      "entropy": 1.9239093586802483,
      "epoch": 0.13332723683874023,
      "grad_norm": 10.611305236816406,
      "learning_rate": 6.928485127297019e-06,
      "loss": 0.5381,
      "mean_token_accuracy": 0.828672143816948,
      "num_tokens": 51689903.0,
      "step": 43010
    },
    {
      "entropy": 1.9985890626907348,
      "epoch": 0.13335823596378993,
      "grad_norm": 9.104059219360352,
      "learning_rate": 6.92767979851031e-06,
      "loss": 0.6427,
      "mean_token_accuracy": 0.8104574084281921,
      "num_tokens": 51700881.0,
      "step": 43020
    },
    {
      "entropy": 1.8614750012755394,
      "epoch": 0.13338923508883962,
      "grad_norm": 3.7970998287200928,
      "learning_rate": 6.926874750479225e-06,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8261910125613212,
      "num_tokens": 51713424.0,
      "step": 43030
    },
    {
      "entropy": 1.9317557483911514,
      "epoch": 0.13342023421388932,
      "grad_norm": 4.564321041107178,
      "learning_rate": 6.926069983040672e-06,
      "loss": 0.514,
      "mean_token_accuracy": 0.8412420928478241,
      "num_tokens": 51724984.0,
      "step": 43040
    },
    {
      "entropy": 1.8474729374051093,
      "epoch": 0.13345123333893902,
      "grad_norm": 4.81390905380249,
      "learning_rate": 6.925265496031692e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8341810539364815,
      "num_tokens": 51738001.0,
      "step": 43050
    },
    {
      "entropy": 1.938997507095337,
      "epoch": 0.13348223246398871,
      "grad_norm": 10.528800010681152,
      "learning_rate": 6.92446128928946e-06,
      "loss": 0.611,
      "mean_token_accuracy": 0.8227848649024964,
      "num_tokens": 51749495.0,
      "step": 43060
    },
    {
      "entropy": 1.724546130001545,
      "epoch": 0.1335132315890384,
      "grad_norm": 2.564570188522339,
      "learning_rate": 6.9236573626512815e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8508135721087455,
      "num_tokens": 51763526.0,
      "step": 43070
    },
    {
      "entropy": 1.9087968587875366,
      "epoch": 0.1335442307140881,
      "grad_norm": 10.67532730102539,
      "learning_rate": 6.922853715954594e-06,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8410237103700637,
      "num_tokens": 51775239.0,
      "step": 43080
    },
    {
      "entropy": 1.9154332160949707,
      "epoch": 0.13357522983913778,
      "grad_norm": 8.595878601074219,
      "learning_rate": 6.922050349036968e-06,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8399403557181359,
      "num_tokens": 51786721.0,
      "step": 43090
    },
    {
      "entropy": 1.9540342479944228,
      "epoch": 0.13360622896418747,
      "grad_norm": 8.74820327758789,
      "learning_rate": 6.921247261736105e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8402211844921113,
      "num_tokens": 51797894.0,
      "step": 43100
    },
    {
      "entropy": 1.9339706212282182,
      "epoch": 0.13363722808923717,
      "grad_norm": 10.516135215759277,
      "learning_rate": 6.920444453889838e-06,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8250453129410744,
      "num_tokens": 51808969.0,
      "step": 43110
    },
    {
      "entropy": 1.9804834306240082,
      "epoch": 0.13366822721428687,
      "grad_norm": 9.053743362426758,
      "learning_rate": 6.919641925336133e-06,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8214120969176293,
      "num_tokens": 51820364.0,
      "step": 43120
    },
    {
      "entropy": 1.856301885843277,
      "epoch": 0.13369922633933656,
      "grad_norm": 9.301741600036621,
      "learning_rate": 6.9188396759130886e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8374891221523285,
      "num_tokens": 51832827.0,
      "step": 43130
    },
    {
      "entropy": 1.817978872358799,
      "epoch": 0.13373022546438626,
      "grad_norm": 9.00844955444336,
      "learning_rate": 6.918037705458932e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8360246539115905,
      "num_tokens": 51845614.0,
      "step": 43140
    },
    {
      "entropy": 1.9156497776508332,
      "epoch": 0.13376122458943596,
      "grad_norm": 9.389613151550293,
      "learning_rate": 6.9172360138120205e-06,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8285941392183304,
      "num_tokens": 51857344.0,
      "step": 43150
    },
    {
      "entropy": 1.9007001653313638,
      "epoch": 0.13379222371448565,
      "grad_norm": 9.915302276611328,
      "learning_rate": 6.9164346008108465e-06,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8260676816105843,
      "num_tokens": 51869381.0,
      "step": 43160
    },
    {
      "entropy": 1.8302084445953368,
      "epoch": 0.13382322283953535,
      "grad_norm": 9.13771915435791,
      "learning_rate": 6.915633466294033e-06,
      "loss": 0.502,
      "mean_token_accuracy": 0.841213583946228,
      "num_tokens": 51882099.0,
      "step": 43170
    },
    {
      "entropy": 1.9292520344257356,
      "epoch": 0.13385422196458505,
      "grad_norm": 9.99196720123291,
      "learning_rate": 6.914832610100331e-06,
      "loss": 0.562,
      "mean_token_accuracy": 0.82747802734375,
      "num_tokens": 51893348.0,
      "step": 43180
    },
    {
      "entropy": 1.8885715886950494,
      "epoch": 0.13388522108963474,
      "grad_norm": 9.511621475219727,
      "learning_rate": 6.914032032068623e-06,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8361998051404953,
      "num_tokens": 51906070.0,
      "step": 43190
    },
    {
      "entropy": 1.8899712771177293,
      "epoch": 0.13391622021468444,
      "grad_norm": 12.438838005065918,
      "learning_rate": 6.913231732037921e-06,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8116952329874039,
      "num_tokens": 51917513.0,
      "step": 43200
    },
    {
      "entropy": 1.8960667297244072,
      "epoch": 0.13394721933973414,
      "grad_norm": 10.735957145690918,
      "learning_rate": 6.912431709847373e-06,
      "loss": 0.545,
      "mean_token_accuracy": 0.8333559215068818,
      "num_tokens": 51929508.0,
      "step": 43210
    },
    {
      "entropy": 1.8930713683366776,
      "epoch": 0.13397821846478383,
      "grad_norm": 8.979076385498047,
      "learning_rate": 6.911631965336252e-06,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8422817379236222,
      "num_tokens": 51941545.0,
      "step": 43220
    },
    {
      "entropy": 1.8466507881879806,
      "epoch": 0.13400921758983353,
      "grad_norm": 8.20052719116211,
      "learning_rate": 6.9108324983439605e-06,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8367902502417565,
      "num_tokens": 51953888.0,
      "step": 43230
    },
    {
      "entropy": 1.9022724777460098,
      "epoch": 0.13404021671488323,
      "grad_norm": 8.77592945098877,
      "learning_rate": 6.910033308710034e-06,
      "loss": 0.5512,
      "mean_token_accuracy": 0.819185683131218,
      "num_tokens": 51966016.0,
      "step": 43240
    },
    {
      "entropy": 1.9349356770515442,
      "epoch": 0.13407121583993292,
      "grad_norm": 6.25929594039917,
      "learning_rate": 6.909234396274137e-06,
      "loss": 0.549,
      "mean_token_accuracy": 0.833576287329197,
      "num_tokens": 51978893.0,
      "step": 43250
    },
    {
      "entropy": 1.9031900018453598,
      "epoch": 0.13410221496498262,
      "grad_norm": 8.607972145080566,
      "learning_rate": 6.908435760876063e-06,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8291514754295349,
      "num_tokens": 51990673.0,
      "step": 43260
    },
    {
      "entropy": 1.9223877504467963,
      "epoch": 0.13413321409003232,
      "grad_norm": 8.238653182983398,
      "learning_rate": 6.9076374023557366e-06,
      "loss": 0.5656,
      "mean_token_accuracy": 0.8217443853616715,
      "num_tokens": 52002484.0,
      "step": 43270
    },
    {
      "entropy": 1.9485286980867387,
      "epoch": 0.134164213215082,
      "grad_norm": 13.640851974487305,
      "learning_rate": 6.90683932055321e-06,
      "loss": 0.6573,
      "mean_token_accuracy": 0.8062111675739289,
      "num_tokens": 52013371.0,
      "step": 43280
    },
    {
      "entropy": 1.8719741210341454,
      "epoch": 0.1341952123401317,
      "grad_norm": 7.591920852661133,
      "learning_rate": 6.906041515308666e-06,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8311554431915283,
      "num_tokens": 52025599.0,
      "step": 43290
    },
    {
      "entropy": 1.867848064005375,
      "epoch": 0.1342262114651814,
      "grad_norm": 6.9264068603515625,
      "learning_rate": 6.905243986462417e-06,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8396237522363663,
      "num_tokens": 52038600.0,
      "step": 43300
    },
    {
      "entropy": 1.9002704456448556,
      "epoch": 0.1342572105902311,
      "grad_norm": 8.932539939880371,
      "learning_rate": 6.9044467338549005e-06,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8397117152810096,
      "num_tokens": 52050301.0,
      "step": 43310
    },
    {
      "entropy": 1.839927391707897,
      "epoch": 0.1342882097152808,
      "grad_norm": 3.8596723079681396,
      "learning_rate": 6.903649757326689e-06,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8357837229967118,
      "num_tokens": 52062178.0,
      "step": 43320
    },
    {
      "entropy": 1.9066562354564667,
      "epoch": 0.1343192088403305,
      "grad_norm": 9.047643661499023,
      "learning_rate": 6.902853056718479e-06,
      "loss": 0.5685,
      "mean_token_accuracy": 0.8244990050792694,
      "num_tokens": 52073513.0,
      "step": 43330
    },
    {
      "entropy": 1.9126151755452157,
      "epoch": 0.13435020796538016,
      "grad_norm": 7.049984455108643,
      "learning_rate": 6.9020566318711e-06,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8374730840325355,
      "num_tokens": 52084870.0,
      "step": 43340
    },
    {
      "entropy": 1.9268995508551598,
      "epoch": 0.13438120709042986,
      "grad_norm": 8.504199981689453,
      "learning_rate": 6.901260482625506e-06,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8227284833788872,
      "num_tokens": 52096377.0,
      "step": 43350
    },
    {
      "entropy": 1.8570437088608742,
      "epoch": 0.13441220621547956,
      "grad_norm": 4.3164873123168945,
      "learning_rate": 6.90046460882278e-06,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8381210267543793,
      "num_tokens": 52109355.0,
      "step": 43360
    },
    {
      "entropy": 1.834708495438099,
      "epoch": 0.13444320534052925,
      "grad_norm": 8.836848258972168,
      "learning_rate": 6.8996690103041376e-06,
      "loss": 0.523,
      "mean_token_accuracy": 0.8306301310658455,
      "num_tokens": 52121829.0,
      "step": 43370
    },
    {
      "entropy": 1.9287886828184129,
      "epoch": 0.13447420446557895,
      "grad_norm": 10.67462158203125,
      "learning_rate": 6.898873686910913e-06,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8287528812885284,
      "num_tokens": 52133362.0,
      "step": 43380
    },
    {
      "entropy": 1.855184331536293,
      "epoch": 0.13450520359062865,
      "grad_norm": 3.361020803451538,
      "learning_rate": 6.898078638484581e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8325510248541832,
      "num_tokens": 52145855.0,
      "step": 43390
    },
    {
      "entropy": 1.9872781455516815,
      "epoch": 0.13453620271567834,
      "grad_norm": 4.882389545440674,
      "learning_rate": 6.897283864866734e-06,
      "loss": 0.6029,
      "mean_token_accuracy": 0.8120069310069085,
      "num_tokens": 52157083.0,
      "step": 43400
    },
    {
      "entropy": 1.9066018536686897,
      "epoch": 0.13456720184072804,
      "grad_norm": 8.167949676513672,
      "learning_rate": 6.896489365899096e-06,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8314293026924133,
      "num_tokens": 52169100.0,
      "step": 43410
    },
    {
      "entropy": 1.8998596712946891,
      "epoch": 0.13459820096577774,
      "grad_norm": 8.666936874389648,
      "learning_rate": 6.895695141423521e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8273270547389984,
      "num_tokens": 52181307.0,
      "step": 43420
    },
    {
      "entropy": 1.8541322633624078,
      "epoch": 0.13462920009082743,
      "grad_norm": 4.933404922485352,
      "learning_rate": 6.894901191281985e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8414107546210289,
      "num_tokens": 52193857.0,
      "step": 43430
    },
    {
      "entropy": 1.8842070639133452,
      "epoch": 0.13466019921587713,
      "grad_norm": 7.285519599914551,
      "learning_rate": 6.894107515316597e-06,
      "loss": 0.513,
      "mean_token_accuracy": 0.8374163269996643,
      "num_tokens": 52205689.0,
      "step": 43440
    },
    {
      "entropy": 1.8152766466140746,
      "epoch": 0.13469119834092683,
      "grad_norm": 7.589842796325684,
      "learning_rate": 6.893314113369588e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.845301553606987,
      "num_tokens": 52219579.0,
      "step": 43450
    },
    {
      "entropy": 1.756292749941349,
      "epoch": 0.13472219746597652,
      "grad_norm": 6.797213554382324,
      "learning_rate": 6.89252098528332e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8514433890581131,
      "num_tokens": 52232565.0,
      "step": 43460
    },
    {
      "entropy": 1.8172416999936103,
      "epoch": 0.13475319659102622,
      "grad_norm": 7.267201900482178,
      "learning_rate": 6.891728130900279e-06,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8358834356069564,
      "num_tokens": 52246625.0,
      "step": 43470
    },
    {
      "entropy": 1.9737324267625809,
      "epoch": 0.13478419571607592,
      "grad_norm": 7.657471656799316,
      "learning_rate": 6.89093555006308e-06,
      "loss": 0.5858,
      "mean_token_accuracy": 0.827315254509449,
      "num_tokens": 52257290.0,
      "step": 43480
    },
    {
      "entropy": 1.8453372538089752,
      "epoch": 0.1348151948411256,
      "grad_norm": 3.9033000469207764,
      "learning_rate": 6.890143242614467e-06,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8415728509426117,
      "num_tokens": 52269698.0,
      "step": 43490
    },
    {
      "entropy": 1.912654523551464,
      "epoch": 0.1348461939661753,
      "grad_norm": 9.669717788696289,
      "learning_rate": 6.889351208397301e-06,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8255023837089539,
      "num_tokens": 52281310.0,
      "step": 43500
    },
    {
      "entropy": 1.9051228925585746,
      "epoch": 0.134877193091225,
      "grad_norm": 9.635804176330566,
      "learning_rate": 6.888559447254581e-06,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8345887675881386,
      "num_tokens": 52293144.0,
      "step": 43510
    },
    {
      "entropy": 1.8088420122861861,
      "epoch": 0.1349081922162747,
      "grad_norm": 7.230260372161865,
      "learning_rate": 6.887767959029426e-06,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8482095554471016,
      "num_tokens": 52305778.0,
      "step": 43520
    },
    {
      "entropy": 1.7873062670230866,
      "epoch": 0.1349391913413244,
      "grad_norm": 7.785470008850098,
      "learning_rate": 6.886976743565082e-06,
      "loss": 0.4747,
      "mean_token_accuracy": 0.847273476421833,
      "num_tokens": 52318326.0,
      "step": 43530
    },
    {
      "entropy": 1.8959727630019187,
      "epoch": 0.1349701904663741,
      "grad_norm": 4.0128254890441895,
      "learning_rate": 6.886185800704923e-06,
      "loss": 0.5779,
      "mean_token_accuracy": 0.8261519700288773,
      "num_tokens": 52329198.0,
      "step": 43540
    },
    {
      "entropy": 1.8708949625492095,
      "epoch": 0.1350011895914238,
      "grad_norm": 8.658716201782227,
      "learning_rate": 6.8853951302924424e-06,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8351444467902184,
      "num_tokens": 52340797.0,
      "step": 43550
    },
    {
      "entropy": 1.9420608460903168,
      "epoch": 0.1350321887164735,
      "grad_norm": 8.919011116027832,
      "learning_rate": 6.884604732171271e-06,
      "loss": 0.6175,
      "mean_token_accuracy": 0.8187188416719436,
      "num_tokens": 52352295.0,
      "step": 43560
    },
    {
      "entropy": 1.8305167585611344,
      "epoch": 0.13506318784152319,
      "grad_norm": 7.717691898345947,
      "learning_rate": 6.883814606185152e-06,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8335830017924308,
      "num_tokens": 52364889.0,
      "step": 43570
    },
    {
      "entropy": 1.8474395513534545,
      "epoch": 0.13509418696657288,
      "grad_norm": 4.04104471206665,
      "learning_rate": 6.883024752177963e-06,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8278833538293838,
      "num_tokens": 52377375.0,
      "step": 43580
    },
    {
      "entropy": 1.8440303042531014,
      "epoch": 0.13512518609162255,
      "grad_norm": 8.369099617004395,
      "learning_rate": 6.882235169993708e-06,
      "loss": 0.5352,
      "mean_token_accuracy": 0.838912108540535,
      "num_tokens": 52389129.0,
      "step": 43590
    },
    {
      "entropy": 1.7567294076085092,
      "epoch": 0.13515618521667225,
      "grad_norm": 8.187392234802246,
      "learning_rate": 6.881445859476506e-06,
      "loss": 0.4261,
      "mean_token_accuracy": 0.8403543129563331,
      "num_tokens": 52402761.0,
      "step": 43600
    },
    {
      "entropy": 1.854997991025448,
      "epoch": 0.13518718434172194,
      "grad_norm": 3.2255561351776123,
      "learning_rate": 6.880656820470614e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8469349488615989,
      "num_tokens": 52414592.0,
      "step": 43610
    },
    {
      "entropy": 1.9099668189883232,
      "epoch": 0.13521818346677164,
      "grad_norm": 8.775009155273438,
      "learning_rate": 6.8798680528204045e-06,
      "loss": 0.5708,
      "mean_token_accuracy": 0.8262275233864784,
      "num_tokens": 52425981.0,
      "step": 43620
    },
    {
      "entropy": 1.9568244695663453,
      "epoch": 0.13524918259182134,
      "grad_norm": 7.415703773498535,
      "learning_rate": 6.879079556370377e-06,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8358083561062812,
      "num_tokens": 52436580.0,
      "step": 43630
    },
    {
      "entropy": 1.87071183770895,
      "epoch": 0.13528018171687103,
      "grad_norm": 12.355653762817383,
      "learning_rate": 6.87829133096516e-06,
      "loss": 0.5773,
      "mean_token_accuracy": 0.8205010086297989,
      "num_tokens": 52449363.0,
      "step": 43640
    },
    {
      "entropy": 1.7835715875029563,
      "epoch": 0.13531118084192073,
      "grad_norm": 10.02258014678955,
      "learning_rate": 6.877503376449503e-06,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8421624451875687,
      "num_tokens": 52462468.0,
      "step": 43650
    },
    {
      "entropy": 1.834514120221138,
      "epoch": 0.13534217996697043,
      "grad_norm": 9.3626127243042,
      "learning_rate": 6.876715692668278e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8375687584280967,
      "num_tokens": 52475023.0,
      "step": 43660
    },
    {
      "entropy": 1.9362987548112869,
      "epoch": 0.13537317909202012,
      "grad_norm": 8.014119148254395,
      "learning_rate": 6.875928279466486e-06,
      "loss": 0.582,
      "mean_token_accuracy": 0.8320098280906677,
      "num_tokens": 52486718.0,
      "step": 43670
    },
    {
      "entropy": 1.9744535475969314,
      "epoch": 0.13540417821706982,
      "grad_norm": 9.7212495803833,
      "learning_rate": 6.87514113668925e-06,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8281274557113647,
      "num_tokens": 52497664.0,
      "step": 43680
    },
    {
      "entropy": 1.893913634121418,
      "epoch": 0.13543517734211952,
      "grad_norm": 6.767265796661377,
      "learning_rate": 6.874354264181815e-06,
      "loss": 0.5278,
      "mean_token_accuracy": 0.835284897685051,
      "num_tokens": 52509266.0,
      "step": 43690
    },
    {
      "entropy": 1.985574059188366,
      "epoch": 0.1354661764671692,
      "grad_norm": 5.145320892333984,
      "learning_rate": 6.873567661789554e-06,
      "loss": 0.5915,
      "mean_token_accuracy": 0.8212827607989311,
      "num_tokens": 52520661.0,
      "step": 43700
    },
    {
      "entropy": 1.8134831815958024,
      "epoch": 0.1354971755922189,
      "grad_norm": 3.865335702896118,
      "learning_rate": 6.872781329357961e-06,
      "loss": 0.4826,
      "mean_token_accuracy": 0.839052714407444,
      "num_tokens": 52533364.0,
      "step": 43710
    },
    {
      "entropy": 1.9578730061650276,
      "epoch": 0.1355281747172686,
      "grad_norm": 10.674678802490234,
      "learning_rate": 6.871995266732656e-06,
      "loss": 0.6165,
      "mean_token_accuracy": 0.8204101115465164,
      "num_tokens": 52544383.0,
      "step": 43720
    },
    {
      "entropy": 1.7825062423944473,
      "epoch": 0.1355591738423183,
      "grad_norm": 8.773101806640625,
      "learning_rate": 6.871209473759379e-06,
      "loss": 0.4185,
      "mean_token_accuracy": 0.8400455713272095,
      "num_tokens": 52557813.0,
      "step": 43730
    },
    {
      "entropy": 1.9574034690856934,
      "epoch": 0.135590172967368,
      "grad_norm": 9.174714088439941,
      "learning_rate": 6.870423950283998e-06,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8255863264203072,
      "num_tokens": 52569076.0,
      "step": 43740
    },
    {
      "entropy": 1.9655730903148652,
      "epoch": 0.1356211720924177,
      "grad_norm": 10.547797203063965,
      "learning_rate": 6.869638696152497e-06,
      "loss": 0.602,
      "mean_token_accuracy": 0.8277508243918419,
      "num_tokens": 52579434.0,
      "step": 43750
    },
    {
      "entropy": 1.8552222028374672,
      "epoch": 0.1356521712174674,
      "grad_norm": 8.642329216003418,
      "learning_rate": 6.868853711210994e-06,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8307379022240639,
      "num_tokens": 52592462.0,
      "step": 43760
    },
    {
      "entropy": 1.897879421710968,
      "epoch": 0.1356831703425171,
      "grad_norm": 4.2802042961120605,
      "learning_rate": 6.868068995305721e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.830087074637413,
      "num_tokens": 52605129.0,
      "step": 43770
    },
    {
      "entropy": 1.9668879002332686,
      "epoch": 0.1357141694675668,
      "grad_norm": 10.869790077209473,
      "learning_rate": 6.8672845482830375e-06,
      "loss": 0.63,
      "mean_token_accuracy": 0.8210312753915787,
      "num_tokens": 52616024.0,
      "step": 43780
    },
    {
      "entropy": 1.909705390036106,
      "epoch": 0.13574516859261648,
      "grad_norm": 8.73975658416748,
      "learning_rate": 6.866500369989424e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8324807018041611,
      "num_tokens": 52627564.0,
      "step": 43790
    },
    {
      "entropy": 1.8812494575977325,
      "epoch": 0.13577616771766618,
      "grad_norm": 9.084870338439941,
      "learning_rate": 6.865716460271482e-06,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8389278829097748,
      "num_tokens": 52639374.0,
      "step": 43800
    },
    {
      "entropy": 1.9060865387320518,
      "epoch": 0.13580716684271588,
      "grad_norm": 8.113476753234863,
      "learning_rate": 6.86493281897594e-06,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8345111593604088,
      "num_tokens": 52650722.0,
      "step": 43810
    },
    {
      "entropy": 1.8717187106609345,
      "epoch": 0.13583816596776557,
      "grad_norm": 9.743805885314941,
      "learning_rate": 6.864149445949645e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8225096851587296,
      "num_tokens": 52663058.0,
      "step": 43820
    },
    {
      "entropy": 1.8713490128517152,
      "epoch": 0.13586916509281524,
      "grad_norm": 8.773591995239258,
      "learning_rate": 6.86336634103957e-06,
      "loss": 0.5154,
      "mean_token_accuracy": 0.8360903859138489,
      "num_tokens": 52675165.0,
      "step": 43830
    },
    {
      "entropy": 1.8310488358139991,
      "epoch": 0.13590016421786494,
      "grad_norm": 4.114312648773193,
      "learning_rate": 6.862583504092806e-06,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8354193419218063,
      "num_tokens": 52686891.0,
      "step": 43840
    },
    {
      "entropy": 1.8633979707956314,
      "epoch": 0.13593116334291463,
      "grad_norm": 9.394161224365234,
      "learning_rate": 6.861800934956568e-06,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8260962456464768,
      "num_tokens": 52698634.0,
      "step": 43850
    },
    {
      "entropy": 1.8397690102458,
      "epoch": 0.13596216246796433,
      "grad_norm": 8.046875,
      "learning_rate": 6.861018633478194e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8343220546841621,
      "num_tokens": 52711492.0,
      "step": 43860
    },
    {
      "entropy": 1.8941639497876168,
      "epoch": 0.13599316159301403,
      "grad_norm": 10.234258651733398,
      "learning_rate": 6.8602365995051445e-06,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8383172944188118,
      "num_tokens": 52723250.0,
      "step": 43870
    },
    {
      "entropy": 1.8545018136501312,
      "epoch": 0.13602416071806372,
      "grad_norm": 13.808713912963867,
      "learning_rate": 6.8594548328849984e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8330124363303184,
      "num_tokens": 52734990.0,
      "step": 43880
    },
    {
      "entropy": 1.8971347585320473,
      "epoch": 0.13605515984311342,
      "grad_norm": 9.022436141967773,
      "learning_rate": 6.858673333465455e-06,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8370542734861374,
      "num_tokens": 52746095.0,
      "step": 43890
    },
    {
      "entropy": 1.8831921055912972,
      "epoch": 0.13608615896816312,
      "grad_norm": 11.447751998901367,
      "learning_rate": 6.8578921010943434e-06,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8259268268942833,
      "num_tokens": 52758290.0,
      "step": 43900
    },
    {
      "entropy": 1.8786703854799272,
      "epoch": 0.13611715809321281,
      "grad_norm": 7.6989898681640625,
      "learning_rate": 6.857111135619603e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8493966236710548,
      "num_tokens": 52770682.0,
      "step": 43910
    },
    {
      "entropy": 1.8685062855482102,
      "epoch": 0.1361481572182625,
      "grad_norm": 9.15949821472168,
      "learning_rate": 6.856330436889304e-06,
      "loss": 0.555,
      "mean_token_accuracy": 0.8257764980196953,
      "num_tokens": 52782612.0,
      "step": 43920
    },
    {
      "entropy": 1.9584510385990144,
      "epoch": 0.1361791563433122,
      "grad_norm": 8.610404968261719,
      "learning_rate": 6.855550004751631e-06,
      "loss": 0.6142,
      "mean_token_accuracy": 0.8210650518536567,
      "num_tokens": 52793680.0,
      "step": 43930
    },
    {
      "entropy": 1.8771604105830193,
      "epoch": 0.1362101554683619,
      "grad_norm": 9.448259353637695,
      "learning_rate": 6.854769839054892e-06,
      "loss": 0.578,
      "mean_token_accuracy": 0.8262861162424088,
      "num_tokens": 52804679.0,
      "step": 43940
    },
    {
      "entropy": 1.9119238778948784,
      "epoch": 0.1362411545934116,
      "grad_norm": 7.940917015075684,
      "learning_rate": 6.853989939647518e-06,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8279161244630814,
      "num_tokens": 52816435.0,
      "step": 43950
    },
    {
      "entropy": 1.9507309287786483,
      "epoch": 0.1362721537184613,
      "grad_norm": 8.971407890319824,
      "learning_rate": 6.853210306378055e-06,
      "loss": 0.5933,
      "mean_token_accuracy": 0.8229886546730996,
      "num_tokens": 52827096.0,
      "step": 43960
    },
    {
      "entropy": 2.0049089699983598,
      "epoch": 0.136303152843511,
      "grad_norm": 9.031655311584473,
      "learning_rate": 6.852430939095177e-06,
      "loss": 0.6456,
      "mean_token_accuracy": 0.8153550997376442,
      "num_tokens": 52838566.0,
      "step": 43970
    },
    {
      "entropy": 1.9605593144893647,
      "epoch": 0.1363341519685607,
      "grad_norm": 9.522255897521973,
      "learning_rate": 6.851651837647672e-06,
      "loss": 0.6225,
      "mean_token_accuracy": 0.8113553240895272,
      "num_tokens": 52848874.0,
      "step": 43980
    },
    {
      "entropy": 1.9099364891648292,
      "epoch": 0.1363651510936104,
      "grad_norm": 9.196107864379883,
      "learning_rate": 6.85087300188445e-06,
      "loss": 0.5653,
      "mean_token_accuracy": 0.8262931287288666,
      "num_tokens": 52861218.0,
      "step": 43990
    },
    {
      "entropy": 1.911503429710865,
      "epoch": 0.13639615021866008,
      "grad_norm": 9.200566291809082,
      "learning_rate": 6.850094431654544e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8331650167703628,
      "num_tokens": 52872651.0,
      "step": 44000
    },
    {
      "entropy": 1.8342396020889282,
      "epoch": 0.13642714934370978,
      "grad_norm": 4.3369903564453125,
      "learning_rate": 6.849316126807107e-06,
      "loss": 0.507,
      "mean_token_accuracy": 0.8284252122044563,
      "num_tokens": 52885546.0,
      "step": 44010
    },
    {
      "entropy": 1.8490222096443176,
      "epoch": 0.13645814846875948,
      "grad_norm": 7.278261184692383,
      "learning_rate": 6.848538087191405e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8356017455458641,
      "num_tokens": 52897211.0,
      "step": 44020
    },
    {
      "entropy": 1.8598289757966995,
      "epoch": 0.13648914759380917,
      "grad_norm": 8.271986961364746,
      "learning_rate": 6.8477603126568325e-06,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8204525783658028,
      "num_tokens": 52908373.0,
      "step": 44030
    },
    {
      "entropy": 1.8169844523072243,
      "epoch": 0.13652014671885887,
      "grad_norm": 9.633807182312012,
      "learning_rate": 6.846982803052898e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8419378191232681,
      "num_tokens": 52921312.0,
      "step": 44040
    },
    {
      "entropy": 1.8356873735785484,
      "epoch": 0.13655114584390857,
      "grad_norm": 10.545353889465332,
      "learning_rate": 6.846205558229234e-06,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8387926116585731,
      "num_tokens": 52934203.0,
      "step": 44050
    },
    {
      "entropy": 1.9119671627879142,
      "epoch": 0.13658214496895826,
      "grad_norm": 10.298490524291992,
      "learning_rate": 6.845428578035587e-06,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8357212603092193,
      "num_tokens": 52945457.0,
      "step": 44060
    },
    {
      "entropy": 1.9574100762605666,
      "epoch": 0.13661314409400796,
      "grad_norm": 7.986303806304932,
      "learning_rate": 6.8446518623218284e-06,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8229815408587455,
      "num_tokens": 52956675.0,
      "step": 44070
    },
    {
      "entropy": 1.9151659101247787,
      "epoch": 0.13664414321905763,
      "grad_norm": 8.7662935256958,
      "learning_rate": 6.843875410937946e-06,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8375581040978431,
      "num_tokens": 52967730.0,
      "step": 44080
    },
    {
      "entropy": 1.8822844669222831,
      "epoch": 0.13667514234410733,
      "grad_norm": 8.069694519042969,
      "learning_rate": 6.8430992237340455e-06,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8303315103054046,
      "num_tokens": 52979174.0,
      "step": 44090
    },
    {
      "entropy": 1.83747291713953,
      "epoch": 0.13670614146915702,
      "grad_norm": 8.96426773071289,
      "learning_rate": 6.8423233005603554e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8289276763796807,
      "num_tokens": 52991305.0,
      "step": 44100
    },
    {
      "entropy": 1.8574891343712807,
      "epoch": 0.13673714059420672,
      "grad_norm": 8.317970275878906,
      "learning_rate": 6.8415476412672185e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8376831218600274,
      "num_tokens": 53002731.0,
      "step": 44110
    },
    {
      "entropy": 1.929608154296875,
      "epoch": 0.13676813971925642,
      "grad_norm": 8.995396614074707,
      "learning_rate": 6.8407722457051005e-06,
      "loss": 0.5975,
      "mean_token_accuracy": 0.8171190902590751,
      "num_tokens": 53013927.0,
      "step": 44120
    },
    {
      "entropy": 1.777741050720215,
      "epoch": 0.1367991388443061,
      "grad_norm": 9.382816314697266,
      "learning_rate": 6.839997113724582e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8454552739858627,
      "num_tokens": 53026933.0,
      "step": 44130
    },
    {
      "entropy": 1.8432798728346824,
      "epoch": 0.1368301379693558,
      "grad_norm": 7.455301284790039,
      "learning_rate": 6.839222245176366e-06,
      "loss": 0.5518,
      "mean_token_accuracy": 0.837872776389122,
      "num_tokens": 53039377.0,
      "step": 44140
    },
    {
      "entropy": 1.8326244458556176,
      "epoch": 0.1368611370944055,
      "grad_norm": 8.096274375915527,
      "learning_rate": 6.838447639911271e-06,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8261276423931122,
      "num_tokens": 53050778.0,
      "step": 44150
    },
    {
      "entropy": 1.7385433629155158,
      "epoch": 0.1368921362194552,
      "grad_norm": 8.745050430297852,
      "learning_rate": 6.837673297780233e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8421786040067673,
      "num_tokens": 53064304.0,
      "step": 44160
    },
    {
      "entropy": 1.8994769722223281,
      "epoch": 0.1369231353445049,
      "grad_norm": 9.911214828491211,
      "learning_rate": 6.836899218634308e-06,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8285346269607544,
      "num_tokens": 53076071.0,
      "step": 44170
    },
    {
      "entropy": 1.87768052816391,
      "epoch": 0.1369541344695546,
      "grad_norm": 8.06799602508545,
      "learning_rate": 6.836125402324671e-06,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8310927867889404,
      "num_tokens": 53087494.0,
      "step": 44180
    },
    {
      "entropy": 1.9110476791858673,
      "epoch": 0.1369851335946043,
      "grad_norm": 9.325477600097656,
      "learning_rate": 6.835351848702615e-06,
      "loss": 0.6014,
      "mean_token_accuracy": 0.824842332303524,
      "num_tokens": 53098984.0,
      "step": 44190
    },
    {
      "entropy": 1.8674473196268082,
      "epoch": 0.137016132719654,
      "grad_norm": 9.16952896118164,
      "learning_rate": 6.834578557619546e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8323730126023292,
      "num_tokens": 53110328.0,
      "step": 44200
    },
    {
      "entropy": 1.801041378080845,
      "epoch": 0.13704713184470368,
      "grad_norm": 4.72969388961792,
      "learning_rate": 6.8338055289269914e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.840267626941204,
      "num_tokens": 53123078.0,
      "step": 44210
    },
    {
      "entropy": 1.9226965308189392,
      "epoch": 0.13707813096975338,
      "grad_norm": 9.973706245422363,
      "learning_rate": 6.8330327624765955e-06,
      "loss": 0.6247,
      "mean_token_accuracy": 0.8245911300182343,
      "num_tokens": 53133601.0,
      "step": 44220
    },
    {
      "entropy": 1.9381461694836617,
      "epoch": 0.13710913009480308,
      "grad_norm": 8.738226890563965,
      "learning_rate": 6.832260258120124e-06,
      "loss": 0.5691,
      "mean_token_accuracy": 0.8299503594636917,
      "num_tokens": 53145004.0,
      "step": 44230
    },
    {
      "entropy": 1.7989469826221467,
      "epoch": 0.13714012921985277,
      "grad_norm": 9.312309265136719,
      "learning_rate": 6.831488015709451e-06,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8428971692919731,
      "num_tokens": 53158057.0,
      "step": 44240
    },
    {
      "entropy": 1.9178980767726899,
      "epoch": 0.13717112834490247,
      "grad_norm": 8.284287452697754,
      "learning_rate": 6.830716035096575e-06,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8296015933156013,
      "num_tokens": 53169211.0,
      "step": 44250
    },
    {
      "entropy": 1.8435182958841323,
      "epoch": 0.13720212746995217,
      "grad_norm": 9.238348007202148,
      "learning_rate": 6.82994431613361e-06,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8327393636107445,
      "num_tokens": 53180990.0,
      "step": 44260
    },
    {
      "entropy": 1.8937449261546135,
      "epoch": 0.13723312659500186,
      "grad_norm": 4.894327163696289,
      "learning_rate": 6.829172858672786e-06,
      "loss": 0.5545,
      "mean_token_accuracy": 0.827644082903862,
      "num_tokens": 53192974.0,
      "step": 44270
    },
    {
      "entropy": 1.910013110935688,
      "epoch": 0.13726412572005156,
      "grad_norm": 7.417638301849365,
      "learning_rate": 6.828401662566448e-06,
      "loss": 0.5548,
      "mean_token_accuracy": 0.8346083343029023,
      "num_tokens": 53204669.0,
      "step": 44280
    },
    {
      "entropy": 1.878174401819706,
      "epoch": 0.13729512484510126,
      "grad_norm": 8.256863594055176,
      "learning_rate": 6.827630727667063e-06,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8261244371533394,
      "num_tokens": 53216346.0,
      "step": 44290
    },
    {
      "entropy": 1.8674972161650658,
      "epoch": 0.13732612397015095,
      "grad_norm": 8.606951713562012,
      "learning_rate": 6.826860053827209e-06,
      "loss": 0.5128,
      "mean_token_accuracy": 0.832400643825531,
      "num_tokens": 53228130.0,
      "step": 44300
    },
    {
      "entropy": 1.8570816084742545,
      "epoch": 0.13735712309520065,
      "grad_norm": 8.84512710571289,
      "learning_rate": 6.826089640899584e-06,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8290506705641747,
      "num_tokens": 53240500.0,
      "step": 44310
    },
    {
      "entropy": 1.9010854482650756,
      "epoch": 0.13738812222025035,
      "grad_norm": 9.149563789367676,
      "learning_rate": 6.825319488737001e-06,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8214688524603844,
      "num_tokens": 53252154.0,
      "step": 44320
    },
    {
      "entropy": 1.8761808782815934,
      "epoch": 0.13741912134530002,
      "grad_norm": 8.693964958190918,
      "learning_rate": 6.824549597192389e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8360074177384377,
      "num_tokens": 53264128.0,
      "step": 44330
    },
    {
      "entropy": 1.8591908350586892,
      "epoch": 0.1374501204703497,
      "grad_norm": 6.832935810089111,
      "learning_rate": 6.823779966118794e-06,
      "loss": 0.5326,
      "mean_token_accuracy": 0.831575682759285,
      "num_tokens": 53275450.0,
      "step": 44340
    },
    {
      "entropy": 1.840942220389843,
      "epoch": 0.1374811195953994,
      "grad_norm": 8.024560928344727,
      "learning_rate": 6.823010595369376e-06,
      "loss": 0.481,
      "mean_token_accuracy": 0.836443580687046,
      "num_tokens": 53287524.0,
      "step": 44350
    },
    {
      "entropy": 1.9482527136802674,
      "epoch": 0.1375121187204491,
      "grad_norm": 9.615019798278809,
      "learning_rate": 6.8222414847974136e-06,
      "loss": 0.565,
      "mean_token_accuracy": 0.8262650355696678,
      "num_tokens": 53298379.0,
      "step": 44360
    },
    {
      "entropy": 1.8953028246760368,
      "epoch": 0.1375431178454988,
      "grad_norm": 5.4591169357299805,
      "learning_rate": 6.821472634256301e-06,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8326457425951957,
      "num_tokens": 53310130.0,
      "step": 44370
    },
    {
      "entropy": 1.9527557328343392,
      "epoch": 0.1375741169705485,
      "grad_norm": 9.788069725036621,
      "learning_rate": 6.820704043599545e-06,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8213037893176078,
      "num_tokens": 53321390.0,
      "step": 44380
    },
    {
      "entropy": 1.8823087111115455,
      "epoch": 0.1376051160955982,
      "grad_norm": 9.298927307128906,
      "learning_rate": 6.819935712680769e-06,
      "loss": 0.5455,
      "mean_token_accuracy": 0.8245194494724274,
      "num_tokens": 53333069.0,
      "step": 44390
    },
    {
      "entropy": 1.920275841653347,
      "epoch": 0.1376361152206479,
      "grad_norm": 8.774785041809082,
      "learning_rate": 6.819167641353716e-06,
      "loss": 0.5892,
      "mean_token_accuracy": 0.8146108031272888,
      "num_tokens": 53344813.0,
      "step": 44400
    },
    {
      "entropy": 1.87754784822464,
      "epoch": 0.1376671143456976,
      "grad_norm": 4.192923545837402,
      "learning_rate": 6.818399829472239e-06,
      "loss": 0.5525,
      "mean_token_accuracy": 0.8271669283509254,
      "num_tokens": 53357822.0,
      "step": 44410
    },
    {
      "entropy": 1.7767975226044654,
      "epoch": 0.13769811347074729,
      "grad_norm": 9.163516998291016,
      "learning_rate": 6.8176322768903065e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8373310551047325,
      "num_tokens": 53371230.0,
      "step": 44420
    },
    {
      "entropy": 1.8346783101558686,
      "epoch": 0.13772911259579698,
      "grad_norm": 8.969266891479492,
      "learning_rate": 6.816864983462007e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8468596264719963,
      "num_tokens": 53383491.0,
      "step": 44430
    },
    {
      "entropy": 1.8975114315748214,
      "epoch": 0.13776011172084668,
      "grad_norm": 8.912147521972656,
      "learning_rate": 6.816097949041537e-06,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8271856248378754,
      "num_tokens": 53395233.0,
      "step": 44440
    },
    {
      "entropy": 1.7993320614099502,
      "epoch": 0.13779111084589638,
      "grad_norm": 4.392065048217773,
      "learning_rate": 6.815331173483213e-06,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8502204686403274,
      "num_tokens": 53407103.0,
      "step": 44450
    },
    {
      "entropy": 1.7951432079076768,
      "epoch": 0.13782210997094607,
      "grad_norm": 8.865302085876465,
      "learning_rate": 6.8145646566414645e-06,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8381197184324265,
      "num_tokens": 53418949.0,
      "step": 44460
    },
    {
      "entropy": 1.787670373916626,
      "epoch": 0.13785310909599577,
      "grad_norm": 9.359115600585938,
      "learning_rate": 6.813798398370836e-06,
      "loss": 0.4616,
      "mean_token_accuracy": 0.8410790026187897,
      "num_tokens": 53431653.0,
      "step": 44470
    },
    {
      "entropy": 1.8918751433491707,
      "epoch": 0.13788410822104546,
      "grad_norm": 8.384889602661133,
      "learning_rate": 6.813032398525985e-06,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8350237265229226,
      "num_tokens": 53443023.0,
      "step": 44480
    },
    {
      "entropy": 1.8514575868844987,
      "epoch": 0.13791510734609516,
      "grad_norm": 9.768966674804688,
      "learning_rate": 6.812266656961686e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8352868661284447,
      "num_tokens": 53454879.0,
      "step": 44490
    },
    {
      "entropy": 1.7858368843793868,
      "epoch": 0.13794610647114486,
      "grad_norm": 10.17784309387207,
      "learning_rate": 6.811501173532825e-06,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8304376199841499,
      "num_tokens": 53467387.0,
      "step": 44500
    },
    {
      "entropy": 1.8139680430293084,
      "epoch": 0.13797710559619455,
      "grad_norm": 3.656538724899292,
      "learning_rate": 6.810735948094402e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8408811286091804,
      "num_tokens": 53479197.0,
      "step": 44510
    },
    {
      "entropy": 1.8897014811635018,
      "epoch": 0.13800810472124425,
      "grad_norm": 8.641545295715332,
      "learning_rate": 6.809970980501534e-06,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8215809658169746,
      "num_tokens": 53490882.0,
      "step": 44520
    },
    {
      "entropy": 1.9230589419603348,
      "epoch": 0.13803910384629395,
      "grad_norm": 10.131928443908691,
      "learning_rate": 6.80920627060945e-06,
      "loss": 0.5744,
      "mean_token_accuracy": 0.8328511983156204,
      "num_tokens": 53502073.0,
      "step": 44530
    },
    {
      "entropy": 1.924891072511673,
      "epoch": 0.13807010297134364,
      "grad_norm": 8.855517387390137,
      "learning_rate": 6.808441818273496e-06,
      "loss": 0.5667,
      "mean_token_accuracy": 0.8318179234862327,
      "num_tokens": 53513609.0,
      "step": 44540
    },
    {
      "entropy": 1.9333911418914795,
      "epoch": 0.13810110209639334,
      "grad_norm": 8.190418243408203,
      "learning_rate": 6.807677623349122e-06,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8251995086669922,
      "num_tokens": 53524500.0,
      "step": 44550
    },
    {
      "entropy": 1.8267549514770507,
      "epoch": 0.13813210122144304,
      "grad_norm": 9.940940856933594,
      "learning_rate": 6.806913685691902e-06,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8373084679245949,
      "num_tokens": 53537133.0,
      "step": 44560
    },
    {
      "entropy": 1.8623538598418237,
      "epoch": 0.1381631003464927,
      "grad_norm": 9.177823066711426,
      "learning_rate": 6.806150005157519e-06,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8380098685622215,
      "num_tokens": 53548906.0,
      "step": 44570
    },
    {
      "entropy": 1.8576925709843635,
      "epoch": 0.1381940994715424,
      "grad_norm": 4.811609268188477,
      "learning_rate": 6.805386581601771e-06,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8375003471970558,
      "num_tokens": 53561573.0,
      "step": 44580
    },
    {
      "entropy": 1.8996227890253068,
      "epoch": 0.1382250985965921,
      "grad_norm": 9.413639068603516,
      "learning_rate": 6.804623414880566e-06,
      "loss": 0.5631,
      "mean_token_accuracy": 0.8216565132141114,
      "num_tokens": 53572897.0,
      "step": 44590
    },
    {
      "entropy": 1.9721861988306046,
      "epoch": 0.1382560977216418,
      "grad_norm": 8.800419807434082,
      "learning_rate": 6.803860504849928e-06,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8295346736907959,
      "num_tokens": 53583631.0,
      "step": 44600
    },
    {
      "entropy": 1.848584523051977,
      "epoch": 0.1382870968466915,
      "grad_norm": 10.385255813598633,
      "learning_rate": 6.803097851365994e-06,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8350134581327439,
      "num_tokens": 53596466.0,
      "step": 44610
    },
    {
      "entropy": 1.854781810939312,
      "epoch": 0.1383180959717412,
      "grad_norm": 9.363265991210938,
      "learning_rate": 6.8023354542850115e-06,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8366737559437751,
      "num_tokens": 53608422.0,
      "step": 44620
    },
    {
      "entropy": 1.9337295114994049,
      "epoch": 0.1383490950967909,
      "grad_norm": 10.87320613861084,
      "learning_rate": 6.8015733134633434e-06,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8331938117742539,
      "num_tokens": 53619179.0,
      "step": 44630
    },
    {
      "entropy": 1.864050853252411,
      "epoch": 0.13838009422184058,
      "grad_norm": 9.210199356079102,
      "learning_rate": 6.800811428757463e-06,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8297410145401954,
      "num_tokens": 53631637.0,
      "step": 44640
    },
    {
      "entropy": 1.8213204242289067,
      "epoch": 0.13841109334689028,
      "grad_norm": 8.101444244384766,
      "learning_rate": 6.80004980002396e-06,
      "loss": 0.448,
      "mean_token_accuracy": 0.833535049855709,
      "num_tokens": 53644709.0,
      "step": 44650
    },
    {
      "entropy": 1.8359388574957847,
      "epoch": 0.13844209247193998,
      "grad_norm": 7.781729221343994,
      "learning_rate": 6.799288427119529e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8498460426926613,
      "num_tokens": 53656295.0,
      "step": 44660
    },
    {
      "entropy": 1.7105117201805116,
      "epoch": 0.13847309159698967,
      "grad_norm": 7.457498073577881,
      "learning_rate": 6.798527309900985e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8544366523623467,
      "num_tokens": 53670465.0,
      "step": 44670
    },
    {
      "entropy": 1.913690346479416,
      "epoch": 0.13850409072203937,
      "grad_norm": 9.645509719848633,
      "learning_rate": 6.797766448225251e-06,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8224819481372834,
      "num_tokens": 53681032.0,
      "step": 44680
    },
    {
      "entropy": 1.7610852643847466,
      "epoch": 0.13853508984708907,
      "grad_norm": 9.441847801208496,
      "learning_rate": 6.797005841949362e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8374917894601822,
      "num_tokens": 53693660.0,
      "step": 44690
    },
    {
      "entropy": 1.8001830980181694,
      "epoch": 0.13856608897213876,
      "grad_norm": 7.770538806915283,
      "learning_rate": 6.796245490930466e-06,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8382193520665169,
      "num_tokens": 53706447.0,
      "step": 44700
    },
    {
      "entropy": 1.8964716732501983,
      "epoch": 0.13859708809718846,
      "grad_norm": 10.237454414367676,
      "learning_rate": 6.795485395025823e-06,
      "loss": 0.63,
      "mean_token_accuracy": 0.8149755626916886,
      "num_tokens": 53717276.0,
      "step": 44710
    },
    {
      "entropy": 1.8226286932826041,
      "epoch": 0.13862808722223816,
      "grad_norm": 8.327227592468262,
      "learning_rate": 6.794725554092804e-06,
      "loss": 0.532,
      "mean_token_accuracy": 0.8249738708138465,
      "num_tokens": 53729351.0,
      "step": 44720
    },
    {
      "entropy": 1.8385003119707108,
      "epoch": 0.13865908634728785,
      "grad_norm": 8.358758926391602,
      "learning_rate": 6.793965967988893e-06,
      "loss": 0.5196,
      "mean_token_accuracy": 0.823200698196888,
      "num_tokens": 53741541.0,
      "step": 44730
    },
    {
      "entropy": 1.8498780086636544,
      "epoch": 0.13869008547233755,
      "grad_norm": 9.434165000915527,
      "learning_rate": 6.793206636571682e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8404179245233536,
      "num_tokens": 53753633.0,
      "step": 44740
    },
    {
      "entropy": 1.7737918436527251,
      "epoch": 0.13872108459738725,
      "grad_norm": 9.19772720336914,
      "learning_rate": 6.792447559698879e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8522493004798889,
      "num_tokens": 53766858.0,
      "step": 44750
    },
    {
      "entropy": 1.928390011191368,
      "epoch": 0.13875208372243694,
      "grad_norm": 9.316328048706055,
      "learning_rate": 6.791688737228301e-06,
      "loss": 0.5819,
      "mean_token_accuracy": 0.8271341755986213,
      "num_tokens": 53777506.0,
      "step": 44760
    },
    {
      "entropy": 1.7646124705672264,
      "epoch": 0.13878308284748664,
      "grad_norm": 3.9269862174987793,
      "learning_rate": 6.790930169017873e-06,
      "loss": 0.4169,
      "mean_token_accuracy": 0.8518116131424904,
      "num_tokens": 53790737.0,
      "step": 44770
    },
    {
      "entropy": 1.8543679133057593,
      "epoch": 0.13881408197253634,
      "grad_norm": 4.760611534118652,
      "learning_rate": 6.790171854925639e-06,
      "loss": 0.492,
      "mean_token_accuracy": 0.8325586140155792,
      "num_tokens": 53803127.0,
      "step": 44780
    },
    {
      "entropy": 1.9045397624373437,
      "epoch": 0.13884508109758603,
      "grad_norm": 9.993179321289062,
      "learning_rate": 6.789413794809746e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8328649774193764,
      "num_tokens": 53814755.0,
      "step": 44790
    },
    {
      "entropy": 1.8256755113601684,
      "epoch": 0.13887608022263573,
      "grad_norm": 9.889800071716309,
      "learning_rate": 6.788655988528456e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8507124841213226,
      "num_tokens": 53827255.0,
      "step": 44800
    },
    {
      "entropy": 1.8313057616353035,
      "epoch": 0.13890707934768542,
      "grad_norm": 8.37853717803955,
      "learning_rate": 6.787898435940142e-06,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8321856454014778,
      "num_tokens": 53838659.0,
      "step": 44810
    },
    {
      "entropy": 1.8516384482383728,
      "epoch": 0.1389380784727351,
      "grad_norm": 4.396270751953125,
      "learning_rate": 6.787141136903286e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8430250212550163,
      "num_tokens": 53851127.0,
      "step": 44820
    },
    {
      "entropy": 1.9117999702692032,
      "epoch": 0.1389690775977848,
      "grad_norm": 4.586118698120117,
      "learning_rate": 6.7863840912764766e-06,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8265089154243469,
      "num_tokens": 53862914.0,
      "step": 44830
    },
    {
      "entropy": 1.956819300353527,
      "epoch": 0.1390000767228345,
      "grad_norm": 8.921218872070312,
      "learning_rate": 6.785627298918424e-06,
      "loss": 0.6126,
      "mean_token_accuracy": 0.81765276491642,
      "num_tokens": 53874681.0,
      "step": 44840
    },
    {
      "entropy": 1.8389661602675915,
      "epoch": 0.13903107584788418,
      "grad_norm": 3.82698917388916,
      "learning_rate": 6.784870759687936e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8415364250540733,
      "num_tokens": 53888519.0,
      "step": 44850
    },
    {
      "entropy": 1.8882953882217408,
      "epoch": 0.13906207497293388,
      "grad_norm": 13.667028427124023,
      "learning_rate": 6.78411447344394e-06,
      "loss": 0.5228,
      "mean_token_accuracy": 0.822119127213955,
      "num_tokens": 53900801.0,
      "step": 44860
    },
    {
      "entropy": 1.8657517284154892,
      "epoch": 0.13909307409798358,
      "grad_norm": 8.731566429138184,
      "learning_rate": 6.783358440045469e-06,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8443102672696113,
      "num_tokens": 53912704.0,
      "step": 44870
    },
    {
      "entropy": 1.898390594124794,
      "epoch": 0.13912407322303327,
      "grad_norm": 8.59327507019043,
      "learning_rate": 6.782602659351665e-06,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8349470987915992,
      "num_tokens": 53924654.0,
      "step": 44880
    },
    {
      "entropy": 1.8541928052902221,
      "epoch": 0.13915507234808297,
      "grad_norm": 4.574117660522461,
      "learning_rate": 6.781847131221781e-06,
      "loss": 0.5823,
      "mean_token_accuracy": 0.8248166784644126,
      "num_tokens": 53936606.0,
      "step": 44890
    },
    {
      "entropy": 1.929851384460926,
      "epoch": 0.13918607147313267,
      "grad_norm": 8.305390357971191,
      "learning_rate": 6.781091855515185e-06,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8298721700906754,
      "num_tokens": 53948064.0,
      "step": 44900
    },
    {
      "entropy": 1.9231369107961656,
      "epoch": 0.13921707059818236,
      "grad_norm": 10.014037132263184,
      "learning_rate": 6.780336832091346e-06,
      "loss": 0.5509,
      "mean_token_accuracy": 0.8278407782316208,
      "num_tokens": 53959163.0,
      "step": 44910
    },
    {
      "entropy": 1.8753797337412834,
      "epoch": 0.13924806972323206,
      "grad_norm": 4.650575160980225,
      "learning_rate": 6.779582060809845e-06,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8299560397863388,
      "num_tokens": 53970669.0,
      "step": 44920
    },
    {
      "entropy": 1.8801968157291413,
      "epoch": 0.13927906884828176,
      "grad_norm": 7.466404438018799,
      "learning_rate": 6.778827541530377e-06,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8341574132442474,
      "num_tokens": 53981942.0,
      "step": 44930
    },
    {
      "entropy": 1.9134157732129098,
      "epoch": 0.13931006797333145,
      "grad_norm": 8.176107406616211,
      "learning_rate": 6.7780732741127416e-06,
      "loss": 0.5927,
      "mean_token_accuracy": 0.8287787228822708,
      "num_tokens": 53993567.0,
      "step": 44940
    },
    {
      "entropy": 1.86323581635952,
      "epoch": 0.13934106709838115,
      "grad_norm": 10.628385543823242,
      "learning_rate": 6.77731925841685e-06,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8439363062381744,
      "num_tokens": 54005593.0,
      "step": 44950
    },
    {
      "entropy": 1.9236777603626252,
      "epoch": 0.13937206622343085,
      "grad_norm": 10.189666748046875,
      "learning_rate": 6.77656549430272e-06,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8301165014505386,
      "num_tokens": 54017353.0,
      "step": 44960
    },
    {
      "entropy": 1.9137274518609046,
      "epoch": 0.13940306534848054,
      "grad_norm": 4.35982608795166,
      "learning_rate": 6.77581198163048e-06,
      "loss": 0.576,
      "mean_token_accuracy": 0.8243458449840546,
      "num_tokens": 54028577.0,
      "step": 44970
    },
    {
      "entropy": 1.881200096011162,
      "epoch": 0.13943406447353024,
      "grad_norm": 8.86418342590332,
      "learning_rate": 6.775058720260368e-06,
      "loss": 0.6144,
      "mean_token_accuracy": 0.831314592063427,
      "num_tokens": 54040844.0,
      "step": 44980
    },
    {
      "entropy": 1.8607298702001571,
      "epoch": 0.13946506359857994,
      "grad_norm": 8.576333999633789,
      "learning_rate": 6.77430571005273e-06,
      "loss": 0.5247,
      "mean_token_accuracy": 0.831570616364479,
      "num_tokens": 54052863.0,
      "step": 44990
    },
    {
      "entropy": 1.9090143859386444,
      "epoch": 0.13949606272362963,
      "grad_norm": 4.669756889343262,
      "learning_rate": 6.7735529508680195e-06,
      "loss": 0.5629,
      "mean_token_accuracy": 0.8271263659000396,
      "num_tokens": 54064516.0,
      "step": 45000
    },
    {
      "entropy": 1.8905669406056405,
      "epoch": 0.13952706184867933,
      "grad_norm": 10.381400108337402,
      "learning_rate": 6.772800442566799e-06,
      "loss": 0.5365,
      "mean_token_accuracy": 0.831216785311699,
      "num_tokens": 54076038.0,
      "step": 45010
    },
    {
      "entropy": 1.7761605516076089,
      "epoch": 0.13955806097372903,
      "grad_norm": 2.322737216949463,
      "learning_rate": 6.772048185009742e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8467302456498146,
      "num_tokens": 54089228.0,
      "step": 45020
    },
    {
      "entropy": 1.9183564558625221,
      "epoch": 0.13958906009877872,
      "grad_norm": 3.822122097015381,
      "learning_rate": 6.771296178057627e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8354233950376511,
      "num_tokens": 54100756.0,
      "step": 45030
    },
    {
      "entropy": 1.89002455919981,
      "epoch": 0.13962005922382842,
      "grad_norm": 10.499687194824219,
      "learning_rate": 6.770544421571341e-06,
      "loss": 0.5127,
      "mean_token_accuracy": 0.833263611793518,
      "num_tokens": 54112772.0,
      "step": 45040
    },
    {
      "entropy": 1.8945647366344929,
      "epoch": 0.13965105834887812,
      "grad_norm": 8.45938777923584,
      "learning_rate": 6.7697929154118806e-06,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8283353626728058,
      "num_tokens": 54124878.0,
      "step": 45050
    },
    {
      "entropy": 1.9062055438756942,
      "epoch": 0.1396820574739278,
      "grad_norm": 8.998255729675293,
      "learning_rate": 6.769041659440348e-06,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8295639351010322,
      "num_tokens": 54136958.0,
      "step": 45060
    },
    {
      "entropy": 1.9289535745978355,
      "epoch": 0.13971305659897748,
      "grad_norm": 7.8160505294799805,
      "learning_rate": 6.768290653517961e-06,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8284988284111023,
      "num_tokens": 54148685.0,
      "step": 45070
    },
    {
      "entropy": 1.8776358231902122,
      "epoch": 0.13974405572402718,
      "grad_norm": 10.067404747009277,
      "learning_rate": 6.767539897506031e-06,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8291481390595437,
      "num_tokens": 54161099.0,
      "step": 45080
    },
    {
      "entropy": 1.8345778226852416,
      "epoch": 0.13977505484907687,
      "grad_norm": 9.045985221862793,
      "learning_rate": 6.766789391265992e-06,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8270506203174591,
      "num_tokens": 54173712.0,
      "step": 45090
    },
    {
      "entropy": 1.9202216163277626,
      "epoch": 0.13980605397412657,
      "grad_norm": 8.717995643615723,
      "learning_rate": 6.7660391346593745e-06,
      "loss": 0.6067,
      "mean_token_accuracy": 0.820853091776371,
      "num_tokens": 54186315.0,
      "step": 45100
    },
    {
      "entropy": 1.9440720230340958,
      "epoch": 0.13983705309917627,
      "grad_norm": 7.580258846282959,
      "learning_rate": 6.765289127547821e-06,
      "loss": 0.5868,
      "mean_token_accuracy": 0.8336421042680741,
      "num_tokens": 54197829.0,
      "step": 45110
    },
    {
      "entropy": 1.9380245164036751,
      "epoch": 0.13986805222422596,
      "grad_norm": 9.486980438232422,
      "learning_rate": 6.764539369793085e-06,
      "loss": 0.5743,
      "mean_token_accuracy": 0.818755242228508,
      "num_tokens": 54209626.0,
      "step": 45120
    },
    {
      "entropy": 1.9459102168679236,
      "epoch": 0.13989905134927566,
      "grad_norm": 8.933557510375977,
      "learning_rate": 6.7637898612570185e-06,
      "loss": 0.5483,
      "mean_token_accuracy": 0.8371159300208092,
      "num_tokens": 54221719.0,
      "step": 45130
    },
    {
      "entropy": 1.7940280467271805,
      "epoch": 0.13993005047432536,
      "grad_norm": 10.48863410949707,
      "learning_rate": 6.7630406018015884e-06,
      "loss": 0.4218,
      "mean_token_accuracy": 0.8437355652451515,
      "num_tokens": 54235073.0,
      "step": 45140
    },
    {
      "entropy": 1.7776363700628282,
      "epoch": 0.13996104959937505,
      "grad_norm": 8.434184074401855,
      "learning_rate": 6.762291591288863e-06,
      "loss": 0.434,
      "mean_token_accuracy": 0.8471253350377083,
      "num_tokens": 54248626.0,
      "step": 45150
    },
    {
      "entropy": 1.8552054047584534,
      "epoch": 0.13999204872442475,
      "grad_norm": 7.70085334777832,
      "learning_rate": 6.761542829581025e-06,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8401309624314308,
      "num_tokens": 54260486.0,
      "step": 45160
    },
    {
      "entropy": 1.856016856431961,
      "epoch": 0.14002304784947445,
      "grad_norm": 10.124335289001465,
      "learning_rate": 6.760794316540352e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8326075285673141,
      "num_tokens": 54273686.0,
      "step": 45170
    },
    {
      "entropy": 1.845085544884205,
      "epoch": 0.14005404697452414,
      "grad_norm": 4.437714576721191,
      "learning_rate": 6.760046052029241e-06,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8339051187038422,
      "num_tokens": 54286613.0,
      "step": 45180
    },
    {
      "entropy": 1.8965241134166717,
      "epoch": 0.14008504609957384,
      "grad_norm": 8.724405288696289,
      "learning_rate": 6.7592980359101864e-06,
      "loss": 0.5568,
      "mean_token_accuracy": 0.818605862557888,
      "num_tokens": 54298537.0,
      "step": 45190
    },
    {
      "entropy": 1.8351117700338364,
      "epoch": 0.14011604522462354,
      "grad_norm": 2.729274034500122,
      "learning_rate": 6.758550268045797e-06,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8237488672137261,
      "num_tokens": 54312345.0,
      "step": 45200
    },
    {
      "entropy": 1.8630397498607636,
      "epoch": 0.14014704434967323,
      "grad_norm": 8.679328918457031,
      "learning_rate": 6.757802748298778e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8270406112074852,
      "num_tokens": 54324911.0,
      "step": 45210
    },
    {
      "entropy": 1.8825726188719272,
      "epoch": 0.14017804347472293,
      "grad_norm": 4.007488250732422,
      "learning_rate": 6.757055476531949e-06,
      "loss": 0.519,
      "mean_token_accuracy": 0.8377444759011269,
      "num_tokens": 54337224.0,
      "step": 45220
    },
    {
      "entropy": 1.978707179427147,
      "epoch": 0.14020904259977263,
      "grad_norm": 8.847635269165039,
      "learning_rate": 6.756308452608234e-06,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8280522257089615,
      "num_tokens": 54348233.0,
      "step": 45230
    },
    {
      "entropy": 1.9876850560307502,
      "epoch": 0.14024004172482232,
      "grad_norm": 6.580830097198486,
      "learning_rate": 6.755561676390661e-06,
      "loss": 0.5664,
      "mean_token_accuracy": 0.8235482305288315,
      "num_tokens": 54359366.0,
      "step": 45240
    },
    {
      "entropy": 1.8557663679122924,
      "epoch": 0.14027104084987202,
      "grad_norm": 8.015181541442871,
      "learning_rate": 6.754815147742368e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8315120905637741,
      "num_tokens": 54372683.0,
      "step": 45250
    },
    {
      "entropy": 1.9107207596302032,
      "epoch": 0.14030203997492172,
      "grad_norm": 7.833492279052734,
      "learning_rate": 6.754068866526591e-06,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8162291869521141,
      "num_tokens": 54384271.0,
      "step": 45260
    },
    {
      "entropy": 1.8534480392932893,
      "epoch": 0.1403330390999714,
      "grad_norm": 8.591071128845215,
      "learning_rate": 6.753322832606681e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8405616924166679,
      "num_tokens": 54396228.0,
      "step": 45270
    },
    {
      "entropy": 1.9540248334407806,
      "epoch": 0.1403640382250211,
      "grad_norm": 8.374753952026367,
      "learning_rate": 6.752577045846086e-06,
      "loss": 0.5839,
      "mean_token_accuracy": 0.8295043364167214,
      "num_tokens": 54407440.0,
      "step": 45280
    },
    {
      "entropy": 1.8010128349065782,
      "epoch": 0.1403950373500708,
      "grad_norm": 8.82718276977539,
      "learning_rate": 6.7518315061083694e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8495183497667312,
      "num_tokens": 54420654.0,
      "step": 45290
    },
    {
      "entropy": 1.9086335971951485,
      "epoch": 0.1404260364751205,
      "grad_norm": 9.302217483520508,
      "learning_rate": 6.751086213257192e-06,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8401185050606728,
      "num_tokens": 54432469.0,
      "step": 45300
    },
    {
      "entropy": 1.84428121894598,
      "epoch": 0.14045703560017017,
      "grad_norm": 8.570785522460938,
      "learning_rate": 6.750341167156322e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.840717563033104,
      "num_tokens": 54444860.0,
      "step": 45310
    },
    {
      "entropy": 1.890746709704399,
      "epoch": 0.14048803472521987,
      "grad_norm": 4.786993980407715,
      "learning_rate": 6.749596367669633e-06,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8295010164380073,
      "num_tokens": 54456895.0,
      "step": 45320
    },
    {
      "entropy": 1.9245510414242744,
      "epoch": 0.14051903385026956,
      "grad_norm": 10.287075996398926,
      "learning_rate": 6.748851814661106e-06,
      "loss": 0.5775,
      "mean_token_accuracy": 0.8316981479525566,
      "num_tokens": 54469287.0,
      "step": 45330
    },
    {
      "entropy": 1.918796244263649,
      "epoch": 0.14055003297531926,
      "grad_norm": 7.518104553222656,
      "learning_rate": 6.748107507994823e-06,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8400950893759728,
      "num_tokens": 54481147.0,
      "step": 45340
    },
    {
      "entropy": 1.878979854285717,
      "epoch": 0.14058103210036896,
      "grad_norm": 4.404152870178223,
      "learning_rate": 6.747363447534975e-06,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8336639732122422,
      "num_tokens": 54492785.0,
      "step": 45350
    },
    {
      "entropy": 1.8506067097187042,
      "epoch": 0.14061203122541865,
      "grad_norm": 9.694655418395996,
      "learning_rate": 6.746619633145854e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.831046088039875,
      "num_tokens": 54505486.0,
      "step": 45360
    },
    {
      "entropy": 1.851573745906353,
      "epoch": 0.14064303035046835,
      "grad_norm": 8.404928207397461,
      "learning_rate": 6.745876064691858e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8353975236415863,
      "num_tokens": 54518430.0,
      "step": 45370
    },
    {
      "entropy": 1.8510615780949593,
      "epoch": 0.14067402947551805,
      "grad_norm": 6.233858108520508,
      "learning_rate": 6.745132742037491e-06,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8379595011472702,
      "num_tokens": 54530823.0,
      "step": 45380
    },
    {
      "entropy": 1.8410978406667708,
      "epoch": 0.14070502860056774,
      "grad_norm": 7.378570556640625,
      "learning_rate": 6.744389665047362e-06,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8501470074057579,
      "num_tokens": 54543450.0,
      "step": 45390
    },
    {
      "entropy": 1.811590349674225,
      "epoch": 0.14073602772561744,
      "grad_norm": 9.43700122833252,
      "learning_rate": 6.743646833586182e-06,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8426756337285042,
      "num_tokens": 54556418.0,
      "step": 45400
    },
    {
      "entropy": 1.8802534580230712,
      "epoch": 0.14076702685066714,
      "grad_norm": 8.894246101379395,
      "learning_rate": 6.742904247518765e-06,
      "loss": 0.5336,
      "mean_token_accuracy": 0.831374479830265,
      "num_tokens": 54568516.0,
      "step": 45410
    },
    {
      "entropy": 1.9234187602996826,
      "epoch": 0.14079802597571683,
      "grad_norm": 8.303701400756836,
      "learning_rate": 6.742161906710033e-06,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8301004365086555,
      "num_tokens": 54580154.0,
      "step": 45420
    },
    {
      "entropy": 1.854028156399727,
      "epoch": 0.14082902510076653,
      "grad_norm": 6.356640815734863,
      "learning_rate": 6.741419811025011e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8381151556968689,
      "num_tokens": 54592717.0,
      "step": 45430
    },
    {
      "entropy": 1.9734964907169341,
      "epoch": 0.14086002422581623,
      "grad_norm": 10.328083038330078,
      "learning_rate": 6.740677960328828e-06,
      "loss": 0.5915,
      "mean_token_accuracy": 0.8209626346826553,
      "num_tokens": 54603430.0,
      "step": 45440
    },
    {
      "entropy": 1.973678994178772,
      "epoch": 0.14089102335086592,
      "grad_norm": 9.18193531036377,
      "learning_rate": 6.739936354486713e-06,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8171725884079933,
      "num_tokens": 54614206.0,
      "step": 45450
    },
    {
      "entropy": 1.9304319381713868,
      "epoch": 0.14092202247591562,
      "grad_norm": 9.106557846069336,
      "learning_rate": 6.7391949933640045e-06,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8340640947222709,
      "num_tokens": 54626347.0,
      "step": 45460
    },
    {
      "entropy": 1.931530450284481,
      "epoch": 0.14095302160096532,
      "grad_norm": 10.796926498413086,
      "learning_rate": 6.738453876826143e-06,
      "loss": 0.561,
      "mean_token_accuracy": 0.8309638902544976,
      "num_tokens": 54637794.0,
      "step": 45470
    },
    {
      "entropy": 1.9570015415549278,
      "epoch": 0.140984020726015,
      "grad_norm": 11.423550605773926,
      "learning_rate": 6.7377130047386695e-06,
      "loss": 0.5831,
      "mean_token_accuracy": 0.820400333404541,
      "num_tokens": 54648619.0,
      "step": 45480
    },
    {
      "entropy": 1.9563427597284317,
      "epoch": 0.1410150198510647,
      "grad_norm": 7.9254560470581055,
      "learning_rate": 6.7369723769672335e-06,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8188218504190445,
      "num_tokens": 54659912.0,
      "step": 45490
    },
    {
      "entropy": 1.9148031935095786,
      "epoch": 0.1410460189761144,
      "grad_norm": 7.281384468078613,
      "learning_rate": 6.736231993377581e-06,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8334671080112457,
      "num_tokens": 54671882.0,
      "step": 45500
    },
    {
      "entropy": 1.9391163542866707,
      "epoch": 0.1410770181011641,
      "grad_norm": 9.63330078125,
      "learning_rate": 6.735491853835571e-06,
      "loss": 0.5381,
      "mean_token_accuracy": 0.8372703313827514,
      "num_tokens": 54682803.0,
      "step": 45510
    },
    {
      "entropy": 1.8276369392871856,
      "epoch": 0.1411080172262138,
      "grad_norm": 8.537032127380371,
      "learning_rate": 6.734751958207155e-06,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8435697957873345,
      "num_tokens": 54695679.0,
      "step": 45520
    },
    {
      "entropy": 1.8639580070972444,
      "epoch": 0.1411390163512635,
      "grad_norm": 7.994620323181152,
      "learning_rate": 6.7340123063583955e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8325873509049415,
      "num_tokens": 54708838.0,
      "step": 45530
    },
    {
      "entropy": 1.8180058419704437,
      "epoch": 0.1411700154763132,
      "grad_norm": 4.438790321350098,
      "learning_rate": 6.733272898155452e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8409150257706642,
      "num_tokens": 54721440.0,
      "step": 45540
    },
    {
      "entropy": 1.8847099259495734,
      "epoch": 0.1412010146013629,
      "grad_norm": 10.436013221740723,
      "learning_rate": 6.732533733464593e-06,
      "loss": 0.58,
      "mean_token_accuracy": 0.8249880090355873,
      "num_tokens": 54733585.0,
      "step": 45550
    },
    {
      "entropy": 1.913513371348381,
      "epoch": 0.14123201372641256,
      "grad_norm": 6.460671901702881,
      "learning_rate": 6.731794812152185e-06,
      "loss": 0.5792,
      "mean_token_accuracy": 0.8275411754846573,
      "num_tokens": 54744839.0,
      "step": 45560
    },
    {
      "entropy": 1.9194407686591148,
      "epoch": 0.14126301285146226,
      "grad_norm": 10.074481964111328,
      "learning_rate": 6.7310561340847e-06,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8203522220253945,
      "num_tokens": 54755728.0,
      "step": 45570
    },
    {
      "entropy": 1.975143238902092,
      "epoch": 0.14129401197651195,
      "grad_norm": 7.195339679718018,
      "learning_rate": 6.73031769912871e-06,
      "loss": 0.6573,
      "mean_token_accuracy": 0.8187574937939643,
      "num_tokens": 54767221.0,
      "step": 45580
    },
    {
      "entropy": 1.9089319556951523,
      "epoch": 0.14132501110156165,
      "grad_norm": 9.949457168579102,
      "learning_rate": 6.729579507150891e-06,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8359342351555824,
      "num_tokens": 54778850.0,
      "step": 45590
    },
    {
      "entropy": 1.9323441043496132,
      "epoch": 0.14135601022661134,
      "grad_norm": 4.49478816986084,
      "learning_rate": 6.728841558018021e-06,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8376797735691071,
      "num_tokens": 54790559.0,
      "step": 45600
    },
    {
      "entropy": 1.9415909707546235,
      "epoch": 0.14138700935166104,
      "grad_norm": 8.446100234985352,
      "learning_rate": 6.72810385159698e-06,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8303654089570045,
      "num_tokens": 54802533.0,
      "step": 45610
    },
    {
      "entropy": 1.917179961502552,
      "epoch": 0.14141800847671074,
      "grad_norm": 3.9077000617980957,
      "learning_rate": 6.7273663877547516e-06,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8232975766062737,
      "num_tokens": 54814772.0,
      "step": 45620
    },
    {
      "entropy": 1.8237308233976364,
      "epoch": 0.14144900760176043,
      "grad_norm": 11.04135799407959,
      "learning_rate": 6.726629166358418e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8421555116772652,
      "num_tokens": 54827998.0,
      "step": 45630
    },
    {
      "entropy": 1.876150907576084,
      "epoch": 0.14148000672681013,
      "grad_norm": 6.673070907592773,
      "learning_rate": 6.725892187275168e-06,
      "loss": 0.4424,
      "mean_token_accuracy": 0.845268502831459,
      "num_tokens": 54840719.0,
      "step": 45640
    },
    {
      "entropy": 1.8938727974891663,
      "epoch": 0.14151100585185983,
      "grad_norm": 4.026385307312012,
      "learning_rate": 6.725155450372289e-06,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8300805896520614,
      "num_tokens": 54852189.0,
      "step": 45650
    },
    {
      "entropy": 1.8416188895702361,
      "epoch": 0.14154200497690952,
      "grad_norm": 8.692848205566406,
      "learning_rate": 6.724418955517171e-06,
      "loss": 0.5154,
      "mean_token_accuracy": 0.8287687584757805,
      "num_tokens": 54864719.0,
      "step": 45660
    },
    {
      "entropy": 1.9456066131591796,
      "epoch": 0.14157300410195922,
      "grad_norm": 9.043981552124023,
      "learning_rate": 6.723682702577305e-06,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8262223243713379,
      "num_tokens": 54876153.0,
      "step": 45670
    },
    {
      "entropy": 1.8887315690517426,
      "epoch": 0.14160400322700892,
      "grad_norm": 8.423904418945312,
      "learning_rate": 6.7229466914202864e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.851503835618496,
      "num_tokens": 54888235.0,
      "step": 45680
    },
    {
      "entropy": 1.862038327753544,
      "epoch": 0.14163500235205861,
      "grad_norm": 3.738830804824829,
      "learning_rate": 6.722210921913808e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8433087572455407,
      "num_tokens": 54900600.0,
      "step": 45690
    },
    {
      "entropy": 1.9590093523263932,
      "epoch": 0.1416660014771083,
      "grad_norm": 4.451807498931885,
      "learning_rate": 6.721475393925665e-06,
      "loss": 0.5955,
      "mean_token_accuracy": 0.8190687134861946,
      "num_tokens": 54911971.0,
      "step": 45700
    },
    {
      "entropy": 1.8686208486557008,
      "epoch": 0.141697000602158,
      "grad_norm": 9.367164611816406,
      "learning_rate": 6.720740107323755e-06,
      "loss": 0.511,
      "mean_token_accuracy": 0.8351000919938087,
      "num_tokens": 54924188.0,
      "step": 45710
    },
    {
      "entropy": 1.9188775599002839,
      "epoch": 0.1417279997272077,
      "grad_norm": 10.714550018310547,
      "learning_rate": 6.720005061976077e-06,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8303559988737106,
      "num_tokens": 54935495.0,
      "step": 45720
    },
    {
      "entropy": 1.8970612213015556,
      "epoch": 0.1417589988522574,
      "grad_norm": 4.424524307250977,
      "learning_rate": 6.7192702577507306e-06,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8270143359899521,
      "num_tokens": 54948599.0,
      "step": 45730
    },
    {
      "entropy": 1.9638367608189582,
      "epoch": 0.1417899979773071,
      "grad_norm": 8.639518737792969,
      "learning_rate": 6.718535694515915e-06,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8268617764115334,
      "num_tokens": 54960096.0,
      "step": 45740
    },
    {
      "entropy": 1.920237709581852,
      "epoch": 0.1418209971023568,
      "grad_norm": 8.430821418762207,
      "learning_rate": 6.717801372139931e-06,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8220153465867043,
      "num_tokens": 54972068.0,
      "step": 45750
    },
    {
      "entropy": 1.7880508966743947,
      "epoch": 0.1418519962274065,
      "grad_norm": 4.553556442260742,
      "learning_rate": 6.717067290491183e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8418186247348786,
      "num_tokens": 54985926.0,
      "step": 45760
    },
    {
      "entropy": 1.9695019006729126,
      "epoch": 0.1418829953524562,
      "grad_norm": 8.724945068359375,
      "learning_rate": 6.716333449438172e-06,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8376459792256356,
      "num_tokens": 54996931.0,
      "step": 45770
    },
    {
      "entropy": 1.8776928693056107,
      "epoch": 0.14191399447750588,
      "grad_norm": 4.291085720062256,
      "learning_rate": 6.715599848849499e-06,
      "loss": 0.5726,
      "mean_token_accuracy": 0.822805991768837,
      "num_tokens": 55009212.0,
      "step": 45780
    },
    {
      "entropy": 1.9443933725357057,
      "epoch": 0.14194499360255558,
      "grad_norm": 8.464522361755371,
      "learning_rate": 6.714866488593871e-06,
      "loss": 0.5886,
      "mean_token_accuracy": 0.8179412320256233,
      "num_tokens": 55019971.0,
      "step": 45790
    },
    {
      "entropy": 1.8963981166481971,
      "epoch": 0.14197599272760528,
      "grad_norm": 8.489872932434082,
      "learning_rate": 6.714133368540089e-06,
      "loss": 0.5414,
      "mean_token_accuracy": 0.830699360370636,
      "num_tokens": 55030869.0,
      "step": 45800
    },
    {
      "entropy": 1.8366913974285126,
      "epoch": 0.14200699185265495,
      "grad_norm": 4.7726216316223145,
      "learning_rate": 6.713400488557057e-06,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8389330074191094,
      "num_tokens": 55043876.0,
      "step": 45810
    },
    {
      "entropy": 1.8235629141330718,
      "epoch": 0.14203799097770464,
      "grad_norm": 4.734610080718994,
      "learning_rate": 6.712667848513782e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8486215993762016,
      "num_tokens": 55056283.0,
      "step": 45820
    },
    {
      "entropy": 1.7981175884604454,
      "epoch": 0.14206899010275434,
      "grad_norm": 9.670796394348145,
      "learning_rate": 6.711935448279365e-06,
      "loss": 0.465,
      "mean_token_accuracy": 0.8392187222838402,
      "num_tokens": 55069610.0,
      "step": 45830
    },
    {
      "entropy": 1.8424267813563346,
      "epoch": 0.14209998922780404,
      "grad_norm": 3.722155809402466,
      "learning_rate": 6.711203287723014e-06,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8339389190077782,
      "num_tokens": 55083021.0,
      "step": 45840
    },
    {
      "entropy": 1.877331507205963,
      "epoch": 0.14213098835285373,
      "grad_norm": 9.001885414123535,
      "learning_rate": 6.7104713667140285e-06,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8358711794018745,
      "num_tokens": 55095426.0,
      "step": 45850
    },
    {
      "entropy": 1.9491191014647484,
      "epoch": 0.14216198747790343,
      "grad_norm": 4.518219470977783,
      "learning_rate": 6.709739685121816e-06,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8208028897643089,
      "num_tokens": 55106699.0,
      "step": 45860
    },
    {
      "entropy": 1.8255974546074867,
      "epoch": 0.14219298660295313,
      "grad_norm": 8.47142505645752,
      "learning_rate": 6.70900824281588e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8375585183501244,
      "num_tokens": 55119196.0,
      "step": 45870
    },
    {
      "entropy": 1.844706942141056,
      "epoch": 0.14222398572800282,
      "grad_norm": 3.8983564376831055,
      "learning_rate": 6.70827703966582e-06,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8367140933871269,
      "num_tokens": 55132280.0,
      "step": 45880
    },
    {
      "entropy": 1.8609853073954583,
      "epoch": 0.14225498485305252,
      "grad_norm": 4.357093811035156,
      "learning_rate": 6.707546075541341e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8403833657503128,
      "num_tokens": 55144004.0,
      "step": 45890
    },
    {
      "entropy": 1.9344159916043282,
      "epoch": 0.14228598397810222,
      "grad_norm": 8.425995826721191,
      "learning_rate": 6.706815350312245e-06,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8343609884381294,
      "num_tokens": 55155185.0,
      "step": 45900
    },
    {
      "entropy": 1.9265785560011863,
      "epoch": 0.1423169831031519,
      "grad_norm": 8.693990707397461,
      "learning_rate": 6.706084863848432e-06,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8264458447694778,
      "num_tokens": 55166299.0,
      "step": 45910
    },
    {
      "entropy": 1.9623035162687301,
      "epoch": 0.1423479822282016,
      "grad_norm": 7.973003387451172,
      "learning_rate": 6.705354616019903e-06,
      "loss": 0.589,
      "mean_token_accuracy": 0.8248161807656288,
      "num_tokens": 55177652.0,
      "step": 45920
    },
    {
      "entropy": 1.7614901930093765,
      "epoch": 0.1423789813532513,
      "grad_norm": 10.886859893798828,
      "learning_rate": 6.704624606696758e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8515376761555672,
      "num_tokens": 55191188.0,
      "step": 45930
    },
    {
      "entropy": 1.9051169365644456,
      "epoch": 0.142409980478301,
      "grad_norm": 9.506697654724121,
      "learning_rate": 6.7038948357491925e-06,
      "loss": 0.6262,
      "mean_token_accuracy": 0.8160225167870522,
      "num_tokens": 55202436.0,
      "step": 45940
    },
    {
      "entropy": 1.8652933433651924,
      "epoch": 0.1424409796033507,
      "grad_norm": 8.896991729736328,
      "learning_rate": 6.703165303047507e-06,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8282453447580338,
      "num_tokens": 55213975.0,
      "step": 45950
    },
    {
      "entropy": 1.8330632865428924,
      "epoch": 0.1424719787284004,
      "grad_norm": 10.671113967895508,
      "learning_rate": 6.702436008462098e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8493910744786263,
      "num_tokens": 55226050.0,
      "step": 45960
    },
    {
      "entropy": 1.9106391370296478,
      "epoch": 0.1425029778534501,
      "grad_norm": 10.167787551879883,
      "learning_rate": 6.701706951863456e-06,
      "loss": 0.5872,
      "mean_token_accuracy": 0.8183905243873596,
      "num_tokens": 55237459.0,
      "step": 45970
    },
    {
      "entropy": 1.898222027719021,
      "epoch": 0.1425339769784998,
      "grad_norm": 9.797660827636719,
      "learning_rate": 6.700978133122177e-06,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8333139047026634,
      "num_tokens": 55249387.0,
      "step": 45980
    },
    {
      "entropy": 1.9070947885513305,
      "epoch": 0.14256497610354948,
      "grad_norm": 9.02385139465332,
      "learning_rate": 6.700249552108953e-06,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8196926236152648,
      "num_tokens": 55260722.0,
      "step": 45990
    },
    {
      "entropy": 1.8567352384328841,
      "epoch": 0.14259597522859918,
      "grad_norm": 4.3364458084106445,
      "learning_rate": 6.699521208694573e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8253255099058151,
      "num_tokens": 55273316.0,
      "step": 46000
    },
    {
      "entropy": 1.9003437846899032,
      "epoch": 0.14262697435364888,
      "grad_norm": 10.944622993469238,
      "learning_rate": 6.6987931027499264e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8321839943528175,
      "num_tokens": 55284892.0,
      "step": 46010
    },
    {
      "entropy": 1.9338586196303367,
      "epoch": 0.14265797347869857,
      "grad_norm": 9.371607780456543,
      "learning_rate": 6.698065234146e-06,
      "loss": 0.6125,
      "mean_token_accuracy": 0.8222884982824326,
      "num_tokens": 55296243.0,
      "step": 46020
    },
    {
      "entropy": 1.773510305583477,
      "epoch": 0.14268897260374827,
      "grad_norm": 4.4485602378845215,
      "learning_rate": 6.697337602753876e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8317829132080078,
      "num_tokens": 55310357.0,
      "step": 46030
    },
    {
      "entropy": 1.794416318833828,
      "epoch": 0.14271997172879797,
      "grad_norm": 8.086502075195312,
      "learning_rate": 6.696610208444741e-06,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8388103753328323,
      "num_tokens": 55323254.0,
      "step": 46040
    },
    {
      "entropy": 1.8933477729558945,
      "epoch": 0.14275097085384764,
      "grad_norm": 4.526361465454102,
      "learning_rate": 6.695883051089873e-06,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8254470497369766,
      "num_tokens": 55335499.0,
      "step": 46050
    },
    {
      "entropy": 1.8915483728051186,
      "epoch": 0.14278196997889733,
      "grad_norm": 8.898006439208984,
      "learning_rate": 6.695156130560652e-06,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8290236473083497,
      "num_tokens": 55347737.0,
      "step": 46060
    },
    {
      "entropy": 1.8370354726910592,
      "epoch": 0.14281296910394703,
      "grad_norm": 8.612515449523926,
      "learning_rate": 6.694429446728551e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8437874168157578,
      "num_tokens": 55360168.0,
      "step": 46070
    },
    {
      "entropy": 1.8509749799966813,
      "epoch": 0.14284396822899673,
      "grad_norm": 8.452802658081055,
      "learning_rate": 6.6937029994651485e-06,
      "loss": 0.511,
      "mean_token_accuracy": 0.8318821370601654,
      "num_tokens": 55372897.0,
      "step": 46080
    },
    {
      "entropy": 1.819749604165554,
      "epoch": 0.14287496735404642,
      "grad_norm": 8.721274375915527,
      "learning_rate": 6.692976788642114e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8412142008543014,
      "num_tokens": 55385814.0,
      "step": 46090
    },
    {
      "entropy": 1.9298714756965638,
      "epoch": 0.14290596647909612,
      "grad_norm": 7.0825700759887695,
      "learning_rate": 6.692250814131215e-06,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8247520595788955,
      "num_tokens": 55397111.0,
      "step": 46100
    },
    {
      "entropy": 1.9013961970806121,
      "epoch": 0.14293696560414582,
      "grad_norm": 10.107673645019531,
      "learning_rate": 6.691525075804319e-06,
      "loss": 0.5349,
      "mean_token_accuracy": 0.830139285326004,
      "num_tokens": 55408213.0,
      "step": 46110
    },
    {
      "entropy": 1.9001147076487541,
      "epoch": 0.1429679647291955,
      "grad_norm": 8.269566535949707,
      "learning_rate": 6.690799573533387e-06,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8280939370393753,
      "num_tokens": 55419892.0,
      "step": 46120
    },
    {
      "entropy": 1.9132464528083801,
      "epoch": 0.1429989638542452,
      "grad_norm": 8.161412239074707,
      "learning_rate": 6.690074307190485e-06,
      "loss": 0.5669,
      "mean_token_accuracy": 0.825434684753418,
      "num_tokens": 55430279.0,
      "step": 46130
    },
    {
      "entropy": 1.8425243273377419,
      "epoch": 0.1430299629792949,
      "grad_norm": 4.587552070617676,
      "learning_rate": 6.689349276647765e-06,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8376808211207389,
      "num_tokens": 55441827.0,
      "step": 46140
    },
    {
      "entropy": 1.946098005771637,
      "epoch": 0.1430609621043446,
      "grad_norm": 11.747321128845215,
      "learning_rate": 6.688624481777485e-06,
      "loss": 0.5897,
      "mean_token_accuracy": 0.8218299329280854,
      "num_tokens": 55452623.0,
      "step": 46150
    },
    {
      "entropy": 1.8413902148604393,
      "epoch": 0.1430919612293943,
      "grad_norm": 8.52793025970459,
      "learning_rate": 6.687899922451993e-06,
      "loss": 0.509,
      "mean_token_accuracy": 0.839706726372242,
      "num_tokens": 55464897.0,
      "step": 46160
    },
    {
      "entropy": 1.88809677362442,
      "epoch": 0.143122960354444,
      "grad_norm": 8.712146759033203,
      "learning_rate": 6.6871755985437425e-06,
      "loss": 0.5336,
      "mean_token_accuracy": 0.826761020720005,
      "num_tokens": 55476274.0,
      "step": 46170
    },
    {
      "entropy": 1.8601938232779502,
      "epoch": 0.1431539594794937,
      "grad_norm": 4.302032947540283,
      "learning_rate": 6.686451509925272e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8418595373630524,
      "num_tokens": 55487836.0,
      "step": 46180
    },
    {
      "entropy": 1.7940224602818489,
      "epoch": 0.1431849586045434,
      "grad_norm": 11.997045516967773,
      "learning_rate": 6.685727656469229e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.8457210063934326,
      "num_tokens": 55499979.0,
      "step": 46190
    },
    {
      "entropy": 1.8284522131085397,
      "epoch": 0.14321595772959309,
      "grad_norm": 8.663432121276855,
      "learning_rate": 6.685004038048349e-06,
      "loss": 0.4899,
      "mean_token_accuracy": 0.8391269639134407,
      "num_tokens": 55512352.0,
      "step": 46200
    },
    {
      "entropy": 1.8878327459096909,
      "epoch": 0.14324695685464278,
      "grad_norm": 8.888419151306152,
      "learning_rate": 6.684280654535462e-06,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8362879782915116,
      "num_tokens": 55523451.0,
      "step": 46210
    },
    {
      "entropy": 1.8454574525356293,
      "epoch": 0.14327795597969248,
      "grad_norm": 8.546029090881348,
      "learning_rate": 6.683557505803507e-06,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8269944965839386,
      "num_tokens": 55535264.0,
      "step": 46220
    },
    {
      "entropy": 1.8240334704518317,
      "epoch": 0.14330895510474218,
      "grad_norm": 4.8377885818481445,
      "learning_rate": 6.6828345917255045e-06,
      "loss": 0.498,
      "mean_token_accuracy": 0.8359172835946083,
      "num_tokens": 55547756.0,
      "step": 46230
    },
    {
      "entropy": 1.9085311383008956,
      "epoch": 0.14333995422979187,
      "grad_norm": 9.510826110839844,
      "learning_rate": 6.682111912174579e-06,
      "loss": 0.6108,
      "mean_token_accuracy": 0.828006249666214,
      "num_tokens": 55559998.0,
      "step": 46240
    },
    {
      "entropy": 1.8106452487409115,
      "epoch": 0.14337095335484157,
      "grad_norm": 9.17758846282959,
      "learning_rate": 6.681389467023951e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8463340416550637,
      "num_tokens": 55573054.0,
      "step": 46250
    },
    {
      "entropy": 1.8135211139917373,
      "epoch": 0.14340195247989126,
      "grad_norm": 8.731499671936035,
      "learning_rate": 6.680667256146936e-06,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8303871005773544,
      "num_tokens": 55586890.0,
      "step": 46260
    },
    {
      "entropy": 1.940975472331047,
      "epoch": 0.14343295160494096,
      "grad_norm": 7.670884132385254,
      "learning_rate": 6.679945279416942e-06,
      "loss": 0.5781,
      "mean_token_accuracy": 0.8210515171289444,
      "num_tokens": 55598136.0,
      "step": 46270
    },
    {
      "entropy": 1.8406999617815019,
      "epoch": 0.14346395072999066,
      "grad_norm": 8.801468849182129,
      "learning_rate": 6.679223536707477e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.8413797855377197,
      "num_tokens": 55610122.0,
      "step": 46280
    },
    {
      "entropy": 1.8838975965976714,
      "epoch": 0.14349494985504035,
      "grad_norm": 7.5125908851623535,
      "learning_rate": 6.678502027892142e-06,
      "loss": 0.502,
      "mean_token_accuracy": 0.84541737139225,
      "num_tokens": 55621817.0,
      "step": 46290
    },
    {
      "entropy": 1.8918793559074403,
      "epoch": 0.14352594898009002,
      "grad_norm": 9.769289016723633,
      "learning_rate": 6.677780752844637e-06,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8287898004055023,
      "num_tokens": 55633088.0,
      "step": 46300
    },
    {
      "entropy": 1.82922722697258,
      "epoch": 0.14355694810513972,
      "grad_norm": 10.320469856262207,
      "learning_rate": 6.677059711438752e-06,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8273350238800049,
      "num_tokens": 55644796.0,
      "step": 46310
    },
    {
      "entropy": 1.8213071301579475,
      "epoch": 0.14358794723018942,
      "grad_norm": 8.950617790222168,
      "learning_rate": 6.676338903548379e-06,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8466832935810089,
      "num_tokens": 55656040.0,
      "step": 46320
    },
    {
      "entropy": 1.7985667988657952,
      "epoch": 0.1436189463552391,
      "grad_norm": 11.022417068481445,
      "learning_rate": 6.675618329047501e-06,
      "loss": 0.485,
      "mean_token_accuracy": 0.8403051659464836,
      "num_tokens": 55668863.0,
      "step": 46330
    },
    {
      "entropy": 1.807174201309681,
      "epoch": 0.1436499454802888,
      "grad_norm": 11.001171112060547,
      "learning_rate": 6.674897987810195e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8363948926329613,
      "num_tokens": 55680958.0,
      "step": 46340
    },
    {
      "entropy": 1.8860581666231155,
      "epoch": 0.1436809446053385,
      "grad_norm": 9.26038646697998,
      "learning_rate": 6.674177879710637e-06,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8372591659426689,
      "num_tokens": 55692194.0,
      "step": 46350
    },
    {
      "entropy": 1.8062124118208884,
      "epoch": 0.1437119437303882,
      "grad_norm": 9.093782424926758,
      "learning_rate": 6.6734580046230955e-06,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8431164383888244,
      "num_tokens": 55704710.0,
      "step": 46360
    },
    {
      "entropy": 1.8013732418417931,
      "epoch": 0.1437429428554379,
      "grad_norm": 9.56591796875,
      "learning_rate": 6.672738362421936e-06,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8444247037172318,
      "num_tokens": 55716737.0,
      "step": 46370
    },
    {
      "entropy": 1.8738340884447098,
      "epoch": 0.1437739419804876,
      "grad_norm": 9.894036293029785,
      "learning_rate": 6.672018952981613e-06,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8256061196327209,
      "num_tokens": 55727558.0,
      "step": 46380
    },
    {
      "entropy": 1.9396587938070298,
      "epoch": 0.1438049411055373,
      "grad_norm": 8.471114158630371,
      "learning_rate": 6.671299776176685e-06,
      "loss": 0.6301,
      "mean_token_accuracy": 0.8130400836467743,
      "num_tokens": 55738367.0,
      "step": 46390
    },
    {
      "entropy": 1.8212429881095886,
      "epoch": 0.143835940230587,
      "grad_norm": 6.425068378448486,
      "learning_rate": 6.6705808318817975e-06,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8256319522857666,
      "num_tokens": 55751238.0,
      "step": 46400
    },
    {
      "entropy": 1.8310098990797996,
      "epoch": 0.14386693935563669,
      "grad_norm": 4.085906982421875,
      "learning_rate": 6.669862119971694e-06,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8291790023446083,
      "num_tokens": 55763570.0,
      "step": 46410
    },
    {
      "entropy": 1.808751115947962,
      "epoch": 0.14389793848068638,
      "grad_norm": 10.913744926452637,
      "learning_rate": 6.669143640321213e-06,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8259348452091217,
      "num_tokens": 55776144.0,
      "step": 46420
    },
    {
      "entropy": 1.8655160009860992,
      "epoch": 0.14392893760573608,
      "grad_norm": 11.0983247756958,
      "learning_rate": 6.668425392805282e-06,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8399639427661896,
      "num_tokens": 55787889.0,
      "step": 46430
    },
    {
      "entropy": 1.863385981321335,
      "epoch": 0.14395993673078578,
      "grad_norm": 8.552124977111816,
      "learning_rate": 6.667707377298932e-06,
      "loss": 0.513,
      "mean_token_accuracy": 0.8352120831608772,
      "num_tokens": 55800483.0,
      "step": 46440
    },
    {
      "entropy": 1.9188006550073624,
      "epoch": 0.14399093585583547,
      "grad_norm": 10.365525245666504,
      "learning_rate": 6.66698959367728e-06,
      "loss": 0.6035,
      "mean_token_accuracy": 0.8209626540541649,
      "num_tokens": 55811326.0,
      "step": 46450
    },
    {
      "entropy": 1.9086585596203804,
      "epoch": 0.14402193498088517,
      "grad_norm": 11.863431930541992,
      "learning_rate": 6.666272041815539e-06,
      "loss": 0.5813,
      "mean_token_accuracy": 0.8237880125641823,
      "num_tokens": 55823396.0,
      "step": 46460
    },
    {
      "entropy": 1.8327384784817695,
      "epoch": 0.14405293410593487,
      "grad_norm": 7.275395393371582,
      "learning_rate": 6.66555472158902e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8353081822395325,
      "num_tokens": 55836892.0,
      "step": 46470
    },
    {
      "entropy": 1.7775721468031407,
      "epoch": 0.14408393323098456,
      "grad_norm": 7.417864799499512,
      "learning_rate": 6.664837632873123e-06,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8478566333651543,
      "num_tokens": 55849778.0,
      "step": 46480
    },
    {
      "entropy": 1.8510145902633668,
      "epoch": 0.14411493235603426,
      "grad_norm": 8.72498607635498,
      "learning_rate": 6.664120775543344e-06,
      "loss": 0.5191,
      "mean_token_accuracy": 0.835884952545166,
      "num_tokens": 55861927.0,
      "step": 46490
    },
    {
      "entropy": 1.811974573135376,
      "epoch": 0.14414593148108396,
      "grad_norm": 8.177042007446289,
      "learning_rate": 6.663404149475273e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8483860984444618,
      "num_tokens": 55873403.0,
      "step": 46500
    },
    {
      "entropy": 1.86693923920393,
      "epoch": 0.14417693060613365,
      "grad_norm": 8.267274856567383,
      "learning_rate": 6.662687754544593e-06,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8321180418133736,
      "num_tokens": 55885678.0,
      "step": 46510
    },
    {
      "entropy": 1.8252136752009391,
      "epoch": 0.14420792973118335,
      "grad_norm": 4.183609485626221,
      "learning_rate": 6.661971590627081e-06,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8298415824770927,
      "num_tokens": 55897994.0,
      "step": 46520
    },
    {
      "entropy": 1.8707138195633888,
      "epoch": 0.14423892885623305,
      "grad_norm": 7.805715084075928,
      "learning_rate": 6.661255657598608e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8322044730186462,
      "num_tokens": 55909873.0,
      "step": 46530
    },
    {
      "entropy": 1.8549921035766601,
      "epoch": 0.14426992798128274,
      "grad_norm": 10.361091613769531,
      "learning_rate": 6.660539955335135e-06,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8390928566455841,
      "num_tokens": 55921497.0,
      "step": 46540
    },
    {
      "entropy": 1.8014539882540703,
      "epoch": 0.1443009271063324,
      "grad_norm": 7.748378753662109,
      "learning_rate": 6.659824483712719e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8297727882862092,
      "num_tokens": 55933974.0,
      "step": 46550
    },
    {
      "entropy": 1.912005639076233,
      "epoch": 0.1443319262313821,
      "grad_norm": 8.874786376953125,
      "learning_rate": 6.659109242607511e-06,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8200063824653625,
      "num_tokens": 55945880.0,
      "step": 46560
    },
    {
      "entropy": 1.8200358718633651,
      "epoch": 0.1443629253564318,
      "grad_norm": 4.0915913581848145,
      "learning_rate": 6.658394231895755e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8318188101053238,
      "num_tokens": 55958253.0,
      "step": 46570
    },
    {
      "entropy": 1.8874425664544106,
      "epoch": 0.1443939244814815,
      "grad_norm": 9.732337951660156,
      "learning_rate": 6.657679451453786e-06,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8317007169127464,
      "num_tokens": 55970172.0,
      "step": 46580
    },
    {
      "entropy": 1.8596255108714104,
      "epoch": 0.1444249236065312,
      "grad_norm": 9.834501266479492,
      "learning_rate": 6.656964901158031e-06,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8399197280406951,
      "num_tokens": 55981424.0,
      "step": 46590
    },
    {
      "entropy": 1.8636808514595031,
      "epoch": 0.1444559227315809,
      "grad_norm": 9.933505058288574,
      "learning_rate": 6.656250580885014e-06,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8327499255537987,
      "num_tokens": 55993099.0,
      "step": 46600
    },
    {
      "entropy": 1.8671812251210214,
      "epoch": 0.1444869218566306,
      "grad_norm": 8.399667739868164,
      "learning_rate": 6.6555364905113505e-06,
      "loss": 0.549,
      "mean_token_accuracy": 0.8313580706715584,
      "num_tokens": 56004188.0,
      "step": 46610
    },
    {
      "entropy": 1.8309295520186424,
      "epoch": 0.1445179209816803,
      "grad_norm": 4.422349452972412,
      "learning_rate": 6.654822629913745e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8452798783779144,
      "num_tokens": 56016240.0,
      "step": 46620
    },
    {
      "entropy": 1.746955469250679,
      "epoch": 0.14454892010672998,
      "grad_norm": 5.409885406494141,
      "learning_rate": 6.654108998968999e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8372356191277504,
      "num_tokens": 56029851.0,
      "step": 46630
    },
    {
      "entropy": 1.8878102406859398,
      "epoch": 0.14457991923177968,
      "grad_norm": 9.581794738769531,
      "learning_rate": 6.653395597554003e-06,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8266360089182854,
      "num_tokens": 56041081.0,
      "step": 46640
    },
    {
      "entropy": 1.84484543800354,
      "epoch": 0.14461091835682938,
      "grad_norm": 7.586556434631348,
      "learning_rate": 6.652682425545742e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8356327712535858,
      "num_tokens": 56053083.0,
      "step": 46650
    },
    {
      "entropy": 1.8373423531651496,
      "epoch": 0.14464191748187907,
      "grad_norm": 9.485453605651855,
      "learning_rate": 6.651969482821293e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8305898755788803,
      "num_tokens": 56065247.0,
      "step": 46660
    },
    {
      "entropy": 1.8729679718613625,
      "epoch": 0.14467291660692877,
      "grad_norm": 4.585037708282471,
      "learning_rate": 6.651256769257825e-06,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8245977357029914,
      "num_tokens": 56077249.0,
      "step": 46670
    },
    {
      "entropy": 1.8576960027217866,
      "epoch": 0.14470391573197847,
      "grad_norm": 8.548659324645996,
      "learning_rate": 6.650544284732601e-06,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8239507541060448,
      "num_tokens": 56089496.0,
      "step": 46680
    },
    {
      "entropy": 1.7963466018438339,
      "epoch": 0.14473491485702816,
      "grad_norm": 8.17918872833252,
      "learning_rate": 6.649832029122969e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8414889872074127,
      "num_tokens": 56102439.0,
      "step": 46690
    },
    {
      "entropy": 1.8463149771094323,
      "epoch": 0.14476591398207786,
      "grad_norm": 8.336082458496094,
      "learning_rate": 6.6491200023063785e-06,
      "loss": 0.532,
      "mean_token_accuracy": 0.8314433738589286,
      "num_tokens": 56115196.0,
      "step": 46700
    },
    {
      "entropy": 1.8520590156316756,
      "epoch": 0.14479691310712756,
      "grad_norm": 9.052793502807617,
      "learning_rate": 6.648408204160365e-06,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8344256520271301,
      "num_tokens": 56127775.0,
      "step": 46710
    },
    {
      "entropy": 1.8827403590083123,
      "epoch": 0.14482791223217725,
      "grad_norm": 9.495018005371094,
      "learning_rate": 6.647696634562557e-06,
      "loss": 0.5524,
      "mean_token_accuracy": 0.825685128569603,
      "num_tokens": 56140171.0,
      "step": 46720
    },
    {
      "entropy": 1.9633108615875243,
      "epoch": 0.14485891135722695,
      "grad_norm": 8.340209007263184,
      "learning_rate": 6.646985293390675e-06,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8286648213863372,
      "num_tokens": 56151174.0,
      "step": 46730
    },
    {
      "entropy": 1.912299408018589,
      "epoch": 0.14488991048227665,
      "grad_norm": 9.881667137145996,
      "learning_rate": 6.64627418052253e-06,
      "loss": 0.5312,
      "mean_token_accuracy": 0.834262129664421,
      "num_tokens": 56162304.0,
      "step": 46740
    },
    {
      "entropy": 1.9305769801139832,
      "epoch": 0.14492090960732634,
      "grad_norm": 6.763992786407471,
      "learning_rate": 6.6455632958360265e-06,
      "loss": 0.6003,
      "mean_token_accuracy": 0.8186121672391892,
      "num_tokens": 56173644.0,
      "step": 46750
    },
    {
      "entropy": 1.8460824131965636,
      "epoch": 0.14495190873237604,
      "grad_norm": 10.21903133392334,
      "learning_rate": 6.644852639209157e-06,
      "loss": 0.5388,
      "mean_token_accuracy": 0.8290549755096436,
      "num_tokens": 56185409.0,
      "step": 46760
    },
    {
      "entropy": 1.843972623348236,
      "epoch": 0.14498290785742574,
      "grad_norm": 5.34625244140625,
      "learning_rate": 6.6441422105200105e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8397470220923424,
      "num_tokens": 56197647.0,
      "step": 46770
    },
    {
      "entropy": 1.9071653231978416,
      "epoch": 0.14501390698247543,
      "grad_norm": 9.360774993896484,
      "learning_rate": 6.643432009646762e-06,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8233714982867241,
      "num_tokens": 56209322.0,
      "step": 46780
    },
    {
      "entropy": 1.867496982216835,
      "epoch": 0.14504490610752513,
      "grad_norm": 9.754353523254395,
      "learning_rate": 6.642722036467681e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8340319588780403,
      "num_tokens": 56221453.0,
      "step": 46790
    },
    {
      "entropy": 1.7567606747150422,
      "epoch": 0.1450759052325748,
      "grad_norm": 12.389626502990723,
      "learning_rate": 6.642012290861126e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8503746286034584,
      "num_tokens": 56234727.0,
      "step": 46800
    },
    {
      "entropy": 1.9208232283592224,
      "epoch": 0.1451069043576245,
      "grad_norm": 4.7340803146362305,
      "learning_rate": 6.641302772705548e-06,
      "loss": 0.5905,
      "mean_token_accuracy": 0.8217899233102799,
      "num_tokens": 56245924.0,
      "step": 46810
    },
    {
      "entropy": 1.7486246049404144,
      "epoch": 0.1451379034826742,
      "grad_norm": 4.023608207702637,
      "learning_rate": 6.640593481879488e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8474720388650894,
      "num_tokens": 56259967.0,
      "step": 46820
    },
    {
      "entropy": 1.9232235804200173,
      "epoch": 0.1451689026077239,
      "grad_norm": 9.536555290222168,
      "learning_rate": 6.63988441826158e-06,
      "loss": 0.5834,
      "mean_token_accuracy": 0.8297638326883316,
      "num_tokens": 56271400.0,
      "step": 46830
    },
    {
      "entropy": 1.862909409403801,
      "epoch": 0.14519990173277358,
      "grad_norm": 9.564486503601074,
      "learning_rate": 6.639175581730542e-06,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8356858551502228,
      "num_tokens": 56283115.0,
      "step": 46840
    },
    {
      "entropy": 1.8271246001124382,
      "epoch": 0.14523090085782328,
      "grad_norm": 4.301812648773193,
      "learning_rate": 6.638466972165192e-06,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8387002035975456,
      "num_tokens": 56295992.0,
      "step": 46850
    },
    {
      "entropy": 1.901492816209793,
      "epoch": 0.14526189998287298,
      "grad_norm": 8.603260040283203,
      "learning_rate": 6.63775858944443e-06,
      "loss": 0.5661,
      "mean_token_accuracy": 0.8132639810442924,
      "num_tokens": 56307074.0,
      "step": 46860
    },
    {
      "entropy": 1.7843642815947534,
      "epoch": 0.14529289910792267,
      "grad_norm": 7.738562107086182,
      "learning_rate": 6.637050433447254e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8477630734443664,
      "num_tokens": 56320727.0,
      "step": 46870
    },
    {
      "entropy": 1.8238538324832916,
      "epoch": 0.14532389823297237,
      "grad_norm": 4.08509635925293,
      "learning_rate": 6.636342504052748e-06,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8303026512265206,
      "num_tokens": 56333848.0,
      "step": 46880
    },
    {
      "entropy": 1.7951844319701196,
      "epoch": 0.14535489735802207,
      "grad_norm": 7.781752586364746,
      "learning_rate": 6.635634801140083e-06,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8464242547750473,
      "num_tokens": 56346965.0,
      "step": 46890
    },
    {
      "entropy": 1.8057351917028428,
      "epoch": 0.14538589648307176,
      "grad_norm": 3.1438589096069336,
      "learning_rate": 6.634927324588528e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8528795570135117,
      "num_tokens": 56359959.0,
      "step": 46900
    },
    {
      "entropy": 1.9472463458776474,
      "epoch": 0.14541689560812146,
      "grad_norm": 9.385503768920898,
      "learning_rate": 6.634220074277438e-06,
      "loss": 0.5786,
      "mean_token_accuracy": 0.82788744866848,
      "num_tokens": 56371468.0,
      "step": 46910
    },
    {
      "entropy": 1.8034564316272736,
      "epoch": 0.14544789473317116,
      "grad_norm": 6.253968238830566,
      "learning_rate": 6.633513050086256e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8509827584028244,
      "num_tokens": 56384646.0,
      "step": 46920
    },
    {
      "entropy": 1.8610297739505768,
      "epoch": 0.14547889385822085,
      "grad_norm": 9.084609031677246,
      "learning_rate": 6.6328062518945195e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8404456228017807,
      "num_tokens": 56396918.0,
      "step": 46930
    },
    {
      "entropy": 1.901959379762411,
      "epoch": 0.14550989298327055,
      "grad_norm": 8.576674461364746,
      "learning_rate": 6.63209967958185e-06,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8313698336482048,
      "num_tokens": 56409026.0,
      "step": 46940
    },
    {
      "entropy": 1.8410135254263877,
      "epoch": 0.14554089210832025,
      "grad_norm": 4.225868225097656,
      "learning_rate": 6.631393333027966e-06,
      "loss": 0.4869,
      "mean_token_accuracy": 0.838833749294281,
      "num_tokens": 56421316.0,
      "step": 46950
    },
    {
      "entropy": 1.8137712955474854,
      "epoch": 0.14557189123336994,
      "grad_norm": 7.888924598693848,
      "learning_rate": 6.630687212112668e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8446080341935158,
      "num_tokens": 56434495.0,
      "step": 46960
    },
    {
      "entropy": 1.9139295309782027,
      "epoch": 0.14560289035841964,
      "grad_norm": 7.617772579193115,
      "learning_rate": 6.629981316715853e-06,
      "loss": 0.5159,
      "mean_token_accuracy": 0.831653282046318,
      "num_tokens": 56447269.0,
      "step": 46970
    },
    {
      "entropy": 1.965524472296238,
      "epoch": 0.14563388948346934,
      "grad_norm": 9.798111915588379,
      "learning_rate": 6.629275646717503e-06,
      "loss": 0.5769,
      "mean_token_accuracy": 0.8235475957393646,
      "num_tokens": 56458884.0,
      "step": 46980
    },
    {
      "entropy": 1.8459963738918304,
      "epoch": 0.14566488860851903,
      "grad_norm": 8.215746879577637,
      "learning_rate": 6.628570201997693e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8458223536610603,
      "num_tokens": 56472093.0,
      "step": 46990
    },
    {
      "entropy": 1.956969639658928,
      "epoch": 0.14569588773356873,
      "grad_norm": 8.928169250488281,
      "learning_rate": 6.62786498243658e-06,
      "loss": 0.6133,
      "mean_token_accuracy": 0.8251758500933647,
      "num_tokens": 56482622.0,
      "step": 47000
    },
    {
      "entropy": 1.9019935339689256,
      "epoch": 0.14572688685861843,
      "grad_norm": 7.649944305419922,
      "learning_rate": 6.627159987914421e-06,
      "loss": 0.6096,
      "mean_token_accuracy": 0.8216905370354652,
      "num_tokens": 56495135.0,
      "step": 47010
    },
    {
      "entropy": 1.8709472745656968,
      "epoch": 0.14575788598366812,
      "grad_norm": 5.07206916809082,
      "learning_rate": 6.626455218311551e-06,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8332753911614418,
      "num_tokens": 56507439.0,
      "step": 47020
    },
    {
      "entropy": 1.9712065637111664,
      "epoch": 0.14578888510871782,
      "grad_norm": 8.687137603759766,
      "learning_rate": 6.6257506735084055e-06,
      "loss": 0.6034,
      "mean_token_accuracy": 0.8207579106092453,
      "num_tokens": 56518700.0,
      "step": 47030
    },
    {
      "entropy": 1.7840657129883766,
      "epoch": 0.1458198842337675,
      "grad_norm": 8.681142807006836,
      "learning_rate": 6.625046353385498e-06,
      "loss": 0.4615,
      "mean_token_accuracy": 0.8308488741517067,
      "num_tokens": 56531438.0,
      "step": 47040
    },
    {
      "entropy": 1.9545001640915871,
      "epoch": 0.14585088335881718,
      "grad_norm": 11.037637710571289,
      "learning_rate": 6.624342257823438e-06,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8239473134279252,
      "num_tokens": 56542655.0,
      "step": 47050
    },
    {
      "entropy": 1.8824947997927666,
      "epoch": 0.14588188248386688,
      "grad_norm": 8.642295837402344,
      "learning_rate": 6.623638386702921e-06,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8258578881621361,
      "num_tokens": 56555705.0,
      "step": 47060
    },
    {
      "entropy": 1.871096746623516,
      "epoch": 0.14591288160891658,
      "grad_norm": 9.842241287231445,
      "learning_rate": 6.622934739904732e-06,
      "loss": 0.4928,
      "mean_token_accuracy": 0.836942833662033,
      "num_tokens": 56568268.0,
      "step": 47070
    },
    {
      "entropy": 1.9141883179545403,
      "epoch": 0.14594388073396627,
      "grad_norm": 8.952548027038574,
      "learning_rate": 6.6222313173097454e-06,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8351935803890228,
      "num_tokens": 56579903.0,
      "step": 47080
    },
    {
      "entropy": 1.8572052717208862,
      "epoch": 0.14597487985901597,
      "grad_norm": 9.243520736694336,
      "learning_rate": 6.62152811879892e-06,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8366377666592598,
      "num_tokens": 56592500.0,
      "step": 47090
    },
    {
      "entropy": 1.8863570004701615,
      "epoch": 0.14600587898406567,
      "grad_norm": 8.63272762298584,
      "learning_rate": 6.620825144253312e-06,
      "loss": 0.528,
      "mean_token_accuracy": 0.8290054813027382,
      "num_tokens": 56604096.0,
      "step": 47100
    },
    {
      "entropy": 1.8670645967125892,
      "epoch": 0.14603687810911536,
      "grad_norm": 8.21442985534668,
      "learning_rate": 6.620122393554056e-06,
      "loss": 0.6717,
      "mean_token_accuracy": 0.8223129764199257,
      "num_tokens": 56616718.0,
      "step": 47110
    },
    {
      "entropy": 1.9217035099864006,
      "epoch": 0.14606787723416506,
      "grad_norm": 4.886983871459961,
      "learning_rate": 6.6194198665823796e-06,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8262522548437119,
      "num_tokens": 56627768.0,
      "step": 47120
    },
    {
      "entropy": 1.9420514121651649,
      "epoch": 0.14609887635921476,
      "grad_norm": 8.3319673538208,
      "learning_rate": 6.6187175632195985e-06,
      "loss": 0.602,
      "mean_token_accuracy": 0.8175177812576294,
      "num_tokens": 56638951.0,
      "step": 47130
    },
    {
      "entropy": 1.9218653574585915,
      "epoch": 0.14612987548426445,
      "grad_norm": 8.83938217163086,
      "learning_rate": 6.618015483347118e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.82304065823555,
      "num_tokens": 56651297.0,
      "step": 47140
    },
    {
      "entropy": 1.9661997646093368,
      "epoch": 0.14616087460931415,
      "grad_norm": 9.71408462524414,
      "learning_rate": 6.6173136268464276e-06,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8207120850682259,
      "num_tokens": 56662494.0,
      "step": 47150
    },
    {
      "entropy": 1.934261092543602,
      "epoch": 0.14619187373436385,
      "grad_norm": 8.855196952819824,
      "learning_rate": 6.616611993599109e-06,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8397008880972863,
      "num_tokens": 56673808.0,
      "step": 47160
    },
    {
      "entropy": 1.9365019842982292,
      "epoch": 0.14622287285941354,
      "grad_norm": 8.653056144714355,
      "learning_rate": 6.6159105834868275e-06,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8273678243160247,
      "num_tokens": 56685577.0,
      "step": 47170
    },
    {
      "entropy": 1.872480408847332,
      "epoch": 0.14625387198446324,
      "grad_norm": 8.392419815063477,
      "learning_rate": 6.615209396391338e-06,
      "loss": 0.511,
      "mean_token_accuracy": 0.8286007553339004,
      "num_tokens": 56698302.0,
      "step": 47180
    },
    {
      "entropy": 1.9219907984137534,
      "epoch": 0.14628487110951294,
      "grad_norm": 11.029183387756348,
      "learning_rate": 6.614508432194486e-06,
      "loss": 0.5493,
      "mean_token_accuracy": 0.8315992474555969,
      "num_tokens": 56709524.0,
      "step": 47190
    },
    {
      "entropy": 1.928644596040249,
      "epoch": 0.14631587023456263,
      "grad_norm": 8.550175666809082,
      "learning_rate": 6.613807690778199e-06,
      "loss": 0.5016,
      "mean_token_accuracy": 0.8338589072227478,
      "num_tokens": 56721586.0,
      "step": 47200
    },
    {
      "entropy": 1.8833053424954413,
      "epoch": 0.14634686935961233,
      "grad_norm": 4.840978622436523,
      "learning_rate": 6.613107172024497e-06,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8266768530011177,
      "num_tokens": 56733887.0,
      "step": 47210
    },
    {
      "entropy": 1.8826899453997612,
      "epoch": 0.14637786848466203,
      "grad_norm": 6.925222396850586,
      "learning_rate": 6.6124068758154836e-06,
      "loss": 0.5489,
      "mean_token_accuracy": 0.8324401840567589,
      "num_tokens": 56745661.0,
      "step": 47220
    },
    {
      "entropy": 1.8747637838125228,
      "epoch": 0.14640886760971172,
      "grad_norm": 9.450230598449707,
      "learning_rate": 6.611706802033354e-06,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8353126794099808,
      "num_tokens": 56758002.0,
      "step": 47230
    },
    {
      "entropy": 1.8948550507426263,
      "epoch": 0.14643986673476142,
      "grad_norm": 4.941351890563965,
      "learning_rate": 6.611006950560388e-06,
      "loss": 0.5553,
      "mean_token_accuracy": 0.8193787977099418,
      "num_tokens": 56770680.0,
      "step": 47240
    },
    {
      "entropy": 1.9120934292674066,
      "epoch": 0.14647086585981112,
      "grad_norm": 10.706798553466797,
      "learning_rate": 6.610307321278952e-06,
      "loss": 0.6107,
      "mean_token_accuracy": 0.8241473525762558,
      "num_tokens": 56782436.0,
      "step": 47250
    },
    {
      "entropy": 1.9104921489953994,
      "epoch": 0.1465018649848608,
      "grad_norm": 8.724833488464355,
      "learning_rate": 6.6096079140715005e-06,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8173437684774398,
      "num_tokens": 56794146.0,
      "step": 47260
    },
    {
      "entropy": 1.826416552066803,
      "epoch": 0.1465328641099105,
      "grad_norm": 7.878024578094482,
      "learning_rate": 6.6089087288205766e-06,
      "loss": 0.4277,
      "mean_token_accuracy": 0.8480478748679161,
      "num_tokens": 56806763.0,
      "step": 47270
    },
    {
      "entropy": 1.8449065156280995,
      "epoch": 0.1465638632349602,
      "grad_norm": 8.807769775390625,
      "learning_rate": 6.608209765408807e-06,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8426515519618988,
      "num_tokens": 56819546.0,
      "step": 47280
    },
    {
      "entropy": 1.9609465137124062,
      "epoch": 0.14659486236000988,
      "grad_norm": 9.253771781921387,
      "learning_rate": 6.607511023718909e-06,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8272478267550468,
      "num_tokens": 56830641.0,
      "step": 47290
    },
    {
      "entropy": 1.8133751511573792,
      "epoch": 0.14662586148505957,
      "grad_norm": 4.237439155578613,
      "learning_rate": 6.6068125036336824e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8408553779125214,
      "num_tokens": 56843286.0,
      "step": 47300
    },
    {
      "entropy": 1.8579260095953942,
      "epoch": 0.14665686061010927,
      "grad_norm": 8.59593677520752,
      "learning_rate": 6.6061142050360174e-06,
      "loss": 0.492,
      "mean_token_accuracy": 0.8416462868452073,
      "num_tokens": 56855446.0,
      "step": 47310
    },
    {
      "entropy": 1.9909762263298034,
      "epoch": 0.14668785973515897,
      "grad_norm": 9.792447090148926,
      "learning_rate": 6.60541612780889e-06,
      "loss": 0.5909,
      "mean_token_accuracy": 0.8207048043608666,
      "num_tokens": 56866314.0,
      "step": 47320
    },
    {
      "entropy": 1.8144005626440047,
      "epoch": 0.14671885886020866,
      "grad_norm": 6.8003644943237305,
      "learning_rate": 6.604718271835362e-06,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8494958788156509,
      "num_tokens": 56879467.0,
      "step": 47330
    },
    {
      "entropy": 1.934771877527237,
      "epoch": 0.14674985798525836,
      "grad_norm": 7.686990261077881,
      "learning_rate": 6.60402063699858e-06,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8413501441478729,
      "num_tokens": 56890100.0,
      "step": 47340
    },
    {
      "entropy": 1.8710382498800755,
      "epoch": 0.14678085711030806,
      "grad_norm": 9.023658752441406,
      "learning_rate": 6.603323223181781e-06,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8303638845682144,
      "num_tokens": 56903039.0,
      "step": 47350
    },
    {
      "entropy": 1.853154082596302,
      "epoch": 0.14681185623535775,
      "grad_norm": 7.490630626678467,
      "learning_rate": 6.6026260302682866e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8393293723464013,
      "num_tokens": 56915063.0,
      "step": 47360
    },
    {
      "entropy": 1.8372676715254783,
      "epoch": 0.14684285536040745,
      "grad_norm": 6.07124137878418,
      "learning_rate": 6.601929058141503e-06,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8287788331508636,
      "num_tokens": 56928744.0,
      "step": 47370
    },
    {
      "entropy": 1.831204354763031,
      "epoch": 0.14687385448545714,
      "grad_norm": 4.461570739746094,
      "learning_rate": 6.601232306684922e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8416302114725113,
      "num_tokens": 56941775.0,
      "step": 47380
    },
    {
      "entropy": 1.8040766946971416,
      "epoch": 0.14690485361050684,
      "grad_norm": 3.481854200363159,
      "learning_rate": 6.600535775782128e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8418020218610763,
      "num_tokens": 56955721.0,
      "step": 47390
    },
    {
      "entropy": 1.8790962159633637,
      "epoch": 0.14693585273555654,
      "grad_norm": 8.983044624328613,
      "learning_rate": 6.599839465316782e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8474810421466827,
      "num_tokens": 56968021.0,
      "step": 47400
    },
    {
      "entropy": 1.8795790463685988,
      "epoch": 0.14696685186060623,
      "grad_norm": 9.99634838104248,
      "learning_rate": 6.599143375172638e-06,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8363196149468421,
      "num_tokens": 56979842.0,
      "step": 47410
    },
    {
      "entropy": 1.9174206882715226,
      "epoch": 0.14699785098565593,
      "grad_norm": 8.22517204284668,
      "learning_rate": 6.598447505233533e-06,
      "loss": 0.5717,
      "mean_token_accuracy": 0.8306628108024597,
      "num_tokens": 56991796.0,
      "step": 47420
    },
    {
      "entropy": 1.9805866748094558,
      "epoch": 0.14702885011070563,
      "grad_norm": 9.243436813354492,
      "learning_rate": 6.59775185538339e-06,
      "loss": 0.5826,
      "mean_token_accuracy": 0.829287999868393,
      "num_tokens": 57003139.0,
      "step": 47430
    },
    {
      "entropy": 1.8765076369047164,
      "epoch": 0.14705984923575532,
      "grad_norm": 8.006113052368164,
      "learning_rate": 6.597056425506216e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8328446924686432,
      "num_tokens": 57015872.0,
      "step": 47440
    },
    {
      "entropy": 1.8864967197179794,
      "epoch": 0.14709084836080502,
      "grad_norm": 3.926126003265381,
      "learning_rate": 6.596361215486107e-06,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8418742284178734,
      "num_tokens": 57027981.0,
      "step": 47450
    },
    {
      "entropy": 1.8978651389479637,
      "epoch": 0.14712184748585472,
      "grad_norm": 4.1893134117126465,
      "learning_rate": 6.595666225207241e-06,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8199717015028,
      "num_tokens": 57039934.0,
      "step": 47460
    },
    {
      "entropy": 1.9076218917965888,
      "epoch": 0.14715284661090441,
      "grad_norm": 11.420867919921875,
      "learning_rate": 6.594971454553885e-06,
      "loss": 0.5421,
      "mean_token_accuracy": 0.830481293797493,
      "num_tokens": 57052007.0,
      "step": 47470
    },
    {
      "entropy": 2.0099540084600447,
      "epoch": 0.1471838457359541,
      "grad_norm": 8.73011589050293,
      "learning_rate": 6.5942769034103895e-06,
      "loss": 0.6084,
      "mean_token_accuracy": 0.8137494072318077,
      "num_tokens": 57062753.0,
      "step": 47480
    },
    {
      "entropy": 1.8646449223160744,
      "epoch": 0.1472148448610038,
      "grad_norm": 10.318194389343262,
      "learning_rate": 6.593582571661188e-06,
      "loss": 0.4844,
      "mean_token_accuracy": 0.842064967751503,
      "num_tokens": 57075443.0,
      "step": 47490
    },
    {
      "entropy": 1.8991115644574166,
      "epoch": 0.1472458439860535,
      "grad_norm": 6.439994812011719,
      "learning_rate": 6.592888459190802e-06,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8363965794444084,
      "num_tokens": 57088273.0,
      "step": 47500
    },
    {
      "entropy": 1.9018190592527389,
      "epoch": 0.1472768431111032,
      "grad_norm": 5.49556827545166,
      "learning_rate": 6.592194565883839e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8406558021903038,
      "num_tokens": 57100663.0,
      "step": 47510
    },
    {
      "entropy": 2.0055468559265135,
      "epoch": 0.1473078422361529,
      "grad_norm": 8.085359573364258,
      "learning_rate": 6.591500891624989e-06,
      "loss": 0.612,
      "mean_token_accuracy": 0.8196608617901802,
      "num_tokens": 57111772.0,
      "step": 47520
    },
    {
      "entropy": 1.944721657037735,
      "epoch": 0.1473388413612026,
      "grad_norm": 8.445950508117676,
      "learning_rate": 6.590807436299027e-06,
      "loss": 0.5381,
      "mean_token_accuracy": 0.824900957942009,
      "num_tokens": 57123451.0,
      "step": 47530
    },
    {
      "entropy": 1.9133238226175309,
      "epoch": 0.14736984048625226,
      "grad_norm": 4.461698532104492,
      "learning_rate": 6.590114199790815e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8348732620477677,
      "num_tokens": 57135025.0,
      "step": 47540
    },
    {
      "entropy": 1.862838363647461,
      "epoch": 0.14740083961130196,
      "grad_norm": 3.581024408340454,
      "learning_rate": 6.589421181985297e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8478377997875214,
      "num_tokens": 57146927.0,
      "step": 47550
    },
    {
      "entropy": 1.9030270993709564,
      "epoch": 0.14743183873635166,
      "grad_norm": 7.672686576843262,
      "learning_rate": 6.588728382767504e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8314098447561264,
      "num_tokens": 57158627.0,
      "step": 47560
    },
    {
      "entropy": 1.9589687079191207,
      "epoch": 0.14746283786140135,
      "grad_norm": 9.731719970703125,
      "learning_rate": 6.58803580202255e-06,
      "loss": 0.5612,
      "mean_token_accuracy": 0.8271705433726311,
      "num_tokens": 57169934.0,
      "step": 47570
    },
    {
      "entropy": 1.934072805941105,
      "epoch": 0.14749383698645105,
      "grad_norm": 8.698883056640625,
      "learning_rate": 6.587343439635634e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8397659227252007,
      "num_tokens": 57182220.0,
      "step": 47580
    },
    {
      "entropy": 1.9594342321157456,
      "epoch": 0.14752483611150075,
      "grad_norm": 9.804793357849121,
      "learning_rate": 6.586651295492042e-06,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8358360469341278,
      "num_tokens": 57193785.0,
      "step": 47590
    },
    {
      "entropy": 1.985189399123192,
      "epoch": 0.14755583523655044,
      "grad_norm": 8.428631782531738,
      "learning_rate": 6.585959369477139e-06,
      "loss": 0.5789,
      "mean_token_accuracy": 0.8262299597263336,
      "num_tokens": 57205399.0,
      "step": 47600
    },
    {
      "entropy": 2.0018600046634676,
      "epoch": 0.14758683436160014,
      "grad_norm": 7.76343297958374,
      "learning_rate": 6.585267661476379e-06,
      "loss": 0.5989,
      "mean_token_accuracy": 0.8301928460597991,
      "num_tokens": 57216010.0,
      "step": 47610
    },
    {
      "entropy": 1.9304006546735764,
      "epoch": 0.14761783348664984,
      "grad_norm": 7.311439514160156,
      "learning_rate": 6.584576171375298e-06,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8509150952100754,
      "num_tokens": 57227397.0,
      "step": 47620
    },
    {
      "entropy": 1.924649564921856,
      "epoch": 0.14764883261169953,
      "grad_norm": 8.823622703552246,
      "learning_rate": 6.5838848990595135e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8323667719960213,
      "num_tokens": 57239462.0,
      "step": 47630
    },
    {
      "entropy": 1.9695616766810418,
      "epoch": 0.14767983173674923,
      "grad_norm": 9.303346633911133,
      "learning_rate": 6.583193844414736e-06,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8252382263541221,
      "num_tokens": 57250697.0,
      "step": 47640
    },
    {
      "entropy": 1.9715566843748094,
      "epoch": 0.14771083086179893,
      "grad_norm": 8.258867263793945,
      "learning_rate": 6.582503007326752e-06,
      "loss": 0.579,
      "mean_token_accuracy": 0.8235922992229462,
      "num_tokens": 57261910.0,
      "step": 47650
    },
    {
      "entropy": 1.9681269809603692,
      "epoch": 0.14774182998684862,
      "grad_norm": 4.465610504150391,
      "learning_rate": 6.58181238768143e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8352336034178733,
      "num_tokens": 57273666.0,
      "step": 47660
    },
    {
      "entropy": 1.8807261288166046,
      "epoch": 0.14777282911189832,
      "grad_norm": 8.713099479675293,
      "learning_rate": 6.5811219853647315e-06,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8416456952691078,
      "num_tokens": 57286174.0,
      "step": 47670
    },
    {
      "entropy": 1.9564361080527306,
      "epoch": 0.14780382823694801,
      "grad_norm": 4.097822189331055,
      "learning_rate": 6.580431800262694e-06,
      "loss": 0.5526,
      "mean_token_accuracy": 0.8282778993248939,
      "num_tokens": 57298085.0,
      "step": 47680
    },
    {
      "entropy": 1.9153534591197967,
      "epoch": 0.1478348273619977,
      "grad_norm": 8.843733787536621,
      "learning_rate": 6.57974183226144e-06,
      "loss": 0.5322,
      "mean_token_accuracy": 0.832582226395607,
      "num_tokens": 57310164.0,
      "step": 47690
    },
    {
      "entropy": 1.8493407145142555,
      "epoch": 0.1478658264870474,
      "grad_norm": 10.445718765258789,
      "learning_rate": 6.579052081247181e-06,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8435613483190536,
      "num_tokens": 57323308.0,
      "step": 47700
    },
    {
      "entropy": 1.87943754196167,
      "epoch": 0.1478968256120971,
      "grad_norm": 7.931858062744141,
      "learning_rate": 6.578362547106202e-06,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8353148818016052,
      "num_tokens": 57336086.0,
      "step": 47710
    },
    {
      "entropy": 1.9901050910353661,
      "epoch": 0.1479278247371468,
      "grad_norm": 8.560962677001953,
      "learning_rate": 6.5776732297248805e-06,
      "loss": 0.5716,
      "mean_token_accuracy": 0.8269386544823647,
      "num_tokens": 57347719.0,
      "step": 47720
    },
    {
      "entropy": 2.0061254844069483,
      "epoch": 0.1479588238621965,
      "grad_norm": 7.655875205993652,
      "learning_rate": 6.576984128989673e-06,
      "loss": 0.5468,
      "mean_token_accuracy": 0.837123441696167,
      "num_tokens": 57359271.0,
      "step": 47730
    },
    {
      "entropy": 2.0044953674077988,
      "epoch": 0.1479898229872462,
      "grad_norm": 8.174701690673828,
      "learning_rate": 6.576295244787121e-06,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8275741189718246,
      "num_tokens": 57370109.0,
      "step": 47740
    },
    {
      "entropy": 1.8147762969136239,
      "epoch": 0.1480208221122959,
      "grad_norm": 9.397214889526367,
      "learning_rate": 6.575606577003847e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8475957497954368,
      "num_tokens": 57383452.0,
      "step": 47750
    },
    {
      "entropy": 1.9344044476747513,
      "epoch": 0.1480518212373456,
      "grad_norm": 9.833171844482422,
      "learning_rate": 6.574918125526558e-06,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8328167483210563,
      "num_tokens": 57394798.0,
      "step": 47760
    },
    {
      "entropy": 1.8636110588908195,
      "epoch": 0.14808282036239528,
      "grad_norm": 3.9326298236846924,
      "learning_rate": 6.574229890242045e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8495140254497529,
      "num_tokens": 57407317.0,
      "step": 47770
    },
    {
      "entropy": 1.8715795949101448,
      "epoch": 0.14811381948744495,
      "grad_norm": 4.289327621459961,
      "learning_rate": 6.57354187103718e-06,
      "loss": 0.517,
      "mean_token_accuracy": 0.8313918590545655,
      "num_tokens": 57419698.0,
      "step": 47780
    },
    {
      "entropy": 1.8456829696893693,
      "epoch": 0.14814481861249465,
      "grad_norm": 2.9080862998962402,
      "learning_rate": 6.57285406779892e-06,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8416327074170112,
      "num_tokens": 57432183.0,
      "step": 47790
    },
    {
      "entropy": 1.8631455272436142,
      "epoch": 0.14817581773754435,
      "grad_norm": 8.882286071777344,
      "learning_rate": 6.5721664804143015e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8395325362682342,
      "num_tokens": 57444434.0,
      "step": 47800
    },
    {
      "entropy": 1.8976523950695992,
      "epoch": 0.14820681686259404,
      "grad_norm": 9.217388153076172,
      "learning_rate": 6.5714791087704465e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8214605495333671,
      "num_tokens": 57456138.0,
      "step": 47810
    },
    {
      "entropy": 1.9854907482862472,
      "epoch": 0.14823781598764374,
      "grad_norm": 9.021675109863281,
      "learning_rate": 6.570791952754559e-06,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8244616031646729,
      "num_tokens": 57466708.0,
      "step": 47820
    },
    {
      "entropy": 1.9569459453225135,
      "epoch": 0.14826881511269344,
      "grad_norm": 9.533801078796387,
      "learning_rate": 6.570105012253927e-06,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8229658871889114,
      "num_tokens": 57478273.0,
      "step": 47830
    },
    {
      "entropy": 1.85792086571455,
      "epoch": 0.14829981423774313,
      "grad_norm": 9.635293960571289,
      "learning_rate": 6.569418287155915e-06,
      "loss": 0.518,
      "mean_token_accuracy": 0.8407026007771492,
      "num_tokens": 57491220.0,
      "step": 47840
    },
    {
      "entropy": 1.906779670715332,
      "epoch": 0.14833081336279283,
      "grad_norm": 9.097919464111328,
      "learning_rate": 6.568731777347978e-06,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8338428869843483,
      "num_tokens": 57503621.0,
      "step": 47850
    },
    {
      "entropy": 1.8580841287970542,
      "epoch": 0.14836181248784253,
      "grad_norm": 9.157151222229004,
      "learning_rate": 6.568045482717649e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8528664171695709,
      "num_tokens": 57515610.0,
      "step": 47860
    },
    {
      "entropy": 1.839845283329487,
      "epoch": 0.14839281161289222,
      "grad_norm": 4.290517807006836,
      "learning_rate": 6.567359403152542e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8425501629710197,
      "num_tokens": 57528868.0,
      "step": 47870
    },
    {
      "entropy": 1.8454848155379295,
      "epoch": 0.14842381073794192,
      "grad_norm": 4.269301891326904,
      "learning_rate": 6.566673538540357e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8409277200698853,
      "num_tokens": 57542104.0,
      "step": 47880
    },
    {
      "entropy": 1.8853970304131509,
      "epoch": 0.14845480986299162,
      "grad_norm": 8.715253829956055,
      "learning_rate": 6.5659878887688726e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8417058885097504,
      "num_tokens": 57555226.0,
      "step": 47890
    },
    {
      "entropy": 1.905742235481739,
      "epoch": 0.1484858089880413,
      "grad_norm": 8.63672161102295,
      "learning_rate": 6.56530245372595e-06,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8448764130473136,
      "num_tokens": 57567249.0,
      "step": 47900
    },
    {
      "entropy": 1.9830147728323937,
      "epoch": 0.148516808113091,
      "grad_norm": 7.22813606262207,
      "learning_rate": 6.564617233299536e-06,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8194261863827705,
      "num_tokens": 57579388.0,
      "step": 47910
    },
    {
      "entropy": 1.769668859243393,
      "epoch": 0.1485478072381407,
      "grad_norm": 4.5545125007629395,
      "learning_rate": 6.563932227377654e-06,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8449043944478035,
      "num_tokens": 57594365.0,
      "step": 47920
    },
    {
      "entropy": 1.802185168862343,
      "epoch": 0.1485788063631904,
      "grad_norm": 10.426836013793945,
      "learning_rate": 6.563247435848412e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8285350769758224,
      "num_tokens": 57608164.0,
      "step": 47930
    },
    {
      "entropy": 1.88195166811347,
      "epoch": 0.1486098054882401,
      "grad_norm": 7.976967811584473,
      "learning_rate": 6.5625628586e-06,
      "loss": 0.5016,
      "mean_token_accuracy": 0.8379166424274445,
      "num_tokens": 57621398.0,
      "step": 47940
    },
    {
      "entropy": 1.8815200373530387,
      "epoch": 0.1486408046132898,
      "grad_norm": 4.0617876052856445,
      "learning_rate": 6.561878495520689e-06,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8357270896434784,
      "num_tokens": 57633368.0,
      "step": 47950
    },
    {
      "entropy": 1.8966475576162338,
      "epoch": 0.1486718037383395,
      "grad_norm": 7.131133079528809,
      "learning_rate": 6.56119434649883e-06,
      "loss": 0.5513,
      "mean_token_accuracy": 0.8241571202874184,
      "num_tokens": 57645429.0,
      "step": 47960
    },
    {
      "entropy": 1.970197968184948,
      "epoch": 0.1487028028633892,
      "grad_norm": 8.698470115661621,
      "learning_rate": 6.5605104114228565e-06,
      "loss": 0.5695,
      "mean_token_accuracy": 0.8253947734832764,
      "num_tokens": 57656725.0,
      "step": 47970
    },
    {
      "entropy": 1.8682085782289506,
      "epoch": 0.14873380198843889,
      "grad_norm": 7.869755744934082,
      "learning_rate": 6.5598266901812866e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8386846721172333,
      "num_tokens": 57668875.0,
      "step": 47980
    },
    {
      "entropy": 1.9324923619627952,
      "epoch": 0.14876480111348858,
      "grad_norm": 8.918335914611816,
      "learning_rate": 6.559143182662716e-06,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8326233476400375,
      "num_tokens": 57680511.0,
      "step": 47990
    },
    {
      "entropy": 1.8512993976473808,
      "epoch": 0.14879580023853828,
      "grad_norm": 3.6257503032684326,
      "learning_rate": 6.55845988875582e-06,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8437419295310974,
      "num_tokens": 57693483.0,
      "step": 48000
    },
    {
      "entropy": 1.877199736237526,
      "epoch": 0.14882679936358797,
      "grad_norm": 4.235811233520508,
      "learning_rate": 6.557776808349361e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8381939142942428,
      "num_tokens": 57705721.0,
      "step": 48010
    },
    {
      "entropy": 1.969959369301796,
      "epoch": 0.14885779848863767,
      "grad_norm": 8.612361907958984,
      "learning_rate": 6.557093941332177e-06,
      "loss": 0.594,
      "mean_token_accuracy": 0.8242570266127587,
      "num_tokens": 57716430.0,
      "step": 48020
    },
    {
      "entropy": 1.9459771722555161,
      "epoch": 0.14888879761368734,
      "grad_norm": 4.5910749435424805,
      "learning_rate": 6.556411287593189e-06,
      "loss": 0.5514,
      "mean_token_accuracy": 0.824884532392025,
      "num_tokens": 57728220.0,
      "step": 48030
    },
    {
      "entropy": 1.8936843484640122,
      "epoch": 0.14891979673873704,
      "grad_norm": 9.532984733581543,
      "learning_rate": 6.5557288470214e-06,
      "loss": 0.5743,
      "mean_token_accuracy": 0.8205350682139396,
      "num_tokens": 57740471.0,
      "step": 48040
    },
    {
      "entropy": 1.8339029610157014,
      "epoch": 0.14895079586378673,
      "grad_norm": 3.72184157371521,
      "learning_rate": 6.555046619505892e-06,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8391190290451049,
      "num_tokens": 57752802.0,
      "step": 48050
    },
    {
      "entropy": 1.9061024576425551,
      "epoch": 0.14898179498883643,
      "grad_norm": 7.634225845336914,
      "learning_rate": 6.554364604935828e-06,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8316650420427323,
      "num_tokens": 57764533.0,
      "step": 48060
    },
    {
      "entropy": 1.9208767369389534,
      "epoch": 0.14901279411388613,
      "grad_norm": 9.036537170410156,
      "learning_rate": 6.5536828032004554e-06,
      "loss": 0.5611,
      "mean_token_accuracy": 0.831520353257656,
      "num_tokens": 57776014.0,
      "step": 48070
    },
    {
      "entropy": 1.8474348559975624,
      "epoch": 0.14904379323893582,
      "grad_norm": 4.276791095733643,
      "learning_rate": 6.553001214189095e-06,
      "loss": 0.507,
      "mean_token_accuracy": 0.8321506321430207,
      "num_tokens": 57788876.0,
      "step": 48080
    },
    {
      "entropy": 1.8484969601035117,
      "epoch": 0.14907479236398552,
      "grad_norm": 9.312860488891602,
      "learning_rate": 6.552319837791156e-06,
      "loss": 0.5018,
      "mean_token_accuracy": 0.839774203300476,
      "num_tokens": 57801080.0,
      "step": 48090
    },
    {
      "entropy": 1.8271625474095345,
      "epoch": 0.14910579148903522,
      "grad_norm": 3.9336740970611572,
      "learning_rate": 6.551638673896124e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8387017279863358,
      "num_tokens": 57813838.0,
      "step": 48100
    },
    {
      "entropy": 2.0055576503276824,
      "epoch": 0.1491367906140849,
      "grad_norm": 8.840508460998535,
      "learning_rate": 6.550957722393561e-06,
      "loss": 0.602,
      "mean_token_accuracy": 0.823385763168335,
      "num_tokens": 57824304.0,
      "step": 48110
    },
    {
      "entropy": 1.8011847533285619,
      "epoch": 0.1491677897391346,
      "grad_norm": 7.309523582458496,
      "learning_rate": 6.55027698317312e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8421018213033676,
      "num_tokens": 57838533.0,
      "step": 48120
    },
    {
      "entropy": 1.8618897780776025,
      "epoch": 0.1491987888641843,
      "grad_norm": 8.072103500366211,
      "learning_rate": 6.549596456124524e-06,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8348542749881744,
      "num_tokens": 57850689.0,
      "step": 48130
    },
    {
      "entropy": 1.9547663122415542,
      "epoch": 0.149229787989234,
      "grad_norm": 7.6012349128723145,
      "learning_rate": 6.548916141137581e-06,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8345204189419746,
      "num_tokens": 57861907.0,
      "step": 48140
    },
    {
      "entropy": 1.9206694543361664,
      "epoch": 0.1492607871142837,
      "grad_norm": 9.59740924835205,
      "learning_rate": 6.548236038102178e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8426736682653427,
      "num_tokens": 57873874.0,
      "step": 48150
    },
    {
      "entropy": 1.9666424721479416,
      "epoch": 0.1492917862393334,
      "grad_norm": 9.023880958557129,
      "learning_rate": 6.547556146908285e-06,
      "loss": 0.5622,
      "mean_token_accuracy": 0.8249856010079384,
      "num_tokens": 57885436.0,
      "step": 48160
    },
    {
      "entropy": 1.926368948817253,
      "epoch": 0.1493227853643831,
      "grad_norm": 9.346840858459473,
      "learning_rate": 6.546876467445947e-06,
      "loss": 0.6086,
      "mean_token_accuracy": 0.8210124105215073,
      "num_tokens": 57896563.0,
      "step": 48170
    },
    {
      "entropy": 1.967759844660759,
      "epoch": 0.1493537844894328,
      "grad_norm": 8.98388385772705,
      "learning_rate": 6.546196999605291e-06,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8355759829282761,
      "num_tokens": 57907605.0,
      "step": 48180
    },
    {
      "entropy": 1.9306065008044242,
      "epoch": 0.14938478361448249,
      "grad_norm": 6.92803955078125,
      "learning_rate": 6.545517743276522e-06,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8288378581404686,
      "num_tokens": 57918939.0,
      "step": 48190
    },
    {
      "entropy": 1.871036571264267,
      "epoch": 0.14941578273953218,
      "grad_norm": 5.162753582000732,
      "learning_rate": 6.544838698349932e-06,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8356782793998718,
      "num_tokens": 57931205.0,
      "step": 48200
    },
    {
      "entropy": 1.8842888280749321,
      "epoch": 0.14944678186458188,
      "grad_norm": 9.374017715454102,
      "learning_rate": 6.5441598647158835e-06,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8327220484614373,
      "num_tokens": 57942778.0,
      "step": 48210
    },
    {
      "entropy": 1.8962712571024896,
      "epoch": 0.14947778098963158,
      "grad_norm": 8.113261222839355,
      "learning_rate": 6.543481242264823e-06,
      "loss": 0.5285,
      "mean_token_accuracy": 0.8363337010145188,
      "num_tokens": 57955128.0,
      "step": 48220
    },
    {
      "entropy": 1.9268214851617813,
      "epoch": 0.14950878011468127,
      "grad_norm": 9.561257362365723,
      "learning_rate": 6.542802830887277e-06,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8345524996519089,
      "num_tokens": 57966427.0,
      "step": 48230
    },
    {
      "entropy": 1.8193738594651223,
      "epoch": 0.14953977923973097,
      "grad_norm": 7.488965034484863,
      "learning_rate": 6.542124630473848e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8413304805755615,
      "num_tokens": 57980642.0,
      "step": 48240
    },
    {
      "entropy": 1.8412556283175945,
      "epoch": 0.14957077836478067,
      "grad_norm": 4.181149959564209,
      "learning_rate": 6.541446640915224e-06,
      "loss": 0.4584,
      "mean_token_accuracy": 0.840462064743042,
      "num_tokens": 57994002.0,
      "step": 48250
    },
    {
      "entropy": 1.8988297596573829,
      "epoch": 0.14960177748983036,
      "grad_norm": 7.6107025146484375,
      "learning_rate": 6.540768862102166e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8412007540464401,
      "num_tokens": 58005701.0,
      "step": 48260
    },
    {
      "entropy": 1.860750602185726,
      "epoch": 0.14963277661488006,
      "grad_norm": 9.468915939331055,
      "learning_rate": 6.5400912939255156e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8450239017605782,
      "num_tokens": 58018246.0,
      "step": 48270
    },
    {
      "entropy": 1.9176729574799538,
      "epoch": 0.14966377573992973,
      "grad_norm": 5.558367729187012,
      "learning_rate": 6.5394139362761964e-06,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8284741297364235,
      "num_tokens": 58030012.0,
      "step": 48280
    },
    {
      "entropy": 1.8357847198843955,
      "epoch": 0.14969477486497942,
      "grad_norm": 4.411746501922607,
      "learning_rate": 6.5387367890452105e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8405596747994423,
      "num_tokens": 58042890.0,
      "step": 48290
    },
    {
      "entropy": 1.8932993397116662,
      "epoch": 0.14972577399002912,
      "grad_norm": 8.260334014892578,
      "learning_rate": 6.538059852123636e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8374821558594704,
      "num_tokens": 58055225.0,
      "step": 48300
    },
    {
      "entropy": 1.7874238356947898,
      "epoch": 0.14975677311507882,
      "grad_norm": 10.80783462524414,
      "learning_rate": 6.537383125402632e-06,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8481561884284019,
      "num_tokens": 58068456.0,
      "step": 48310
    },
    {
      "entropy": 1.9409461870789528,
      "epoch": 0.14978777224012851,
      "grad_norm": 8.024087905883789,
      "learning_rate": 6.536706608773437e-06,
      "loss": 0.5432,
      "mean_token_accuracy": 0.835419024527073,
      "num_tokens": 58079470.0,
      "step": 48320
    },
    {
      "entropy": 1.9292469948530198,
      "epoch": 0.1498187713651782,
      "grad_norm": 7.896714687347412,
      "learning_rate": 6.536030302127366e-06,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8290838211774826,
      "num_tokens": 58090946.0,
      "step": 48330
    },
    {
      "entropy": 1.878405897319317,
      "epoch": 0.1498497704902279,
      "grad_norm": 11.065820693969727,
      "learning_rate": 6.535354205355815e-06,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8164193764328956,
      "num_tokens": 58102826.0,
      "step": 48340
    },
    {
      "entropy": 1.9426892310380937,
      "epoch": 0.1498807696152776,
      "grad_norm": 8.384618759155273,
      "learning_rate": 6.534678318350258e-06,
      "loss": 0.5697,
      "mean_token_accuracy": 0.825680835545063,
      "num_tokens": 58113587.0,
      "step": 48350
    },
    {
      "entropy": 1.8836750328540801,
      "epoch": 0.1499117687403273,
      "grad_norm": 9.055339813232422,
      "learning_rate": 6.534002641002247e-06,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8372475564479828,
      "num_tokens": 58125471.0,
      "step": 48360
    },
    {
      "entropy": 1.9558273077011108,
      "epoch": 0.149942767865377,
      "grad_norm": 8.898198127746582,
      "learning_rate": 6.533327173203413e-06,
      "loss": 0.5413,
      "mean_token_accuracy": 0.8336478710174561,
      "num_tokens": 58137376.0,
      "step": 48370
    },
    {
      "entropy": 1.9481865465641022,
      "epoch": 0.1499737669904267,
      "grad_norm": 10.591567993164062,
      "learning_rate": 6.532651914845465e-06,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8392414048314094,
      "num_tokens": 58148589.0,
      "step": 48380
    },
    {
      "entropy": 1.8466496154665948,
      "epoch": 0.1500047661154764,
      "grad_norm": 9.16861629486084,
      "learning_rate": 6.531976865820191e-06,
      "loss": 0.5262,
      "mean_token_accuracy": 0.836252911388874,
      "num_tokens": 58160303.0,
      "step": 48390
    },
    {
      "entropy": 1.8699641510844232,
      "epoch": 0.1500357652405261,
      "grad_norm": 8.319191932678223,
      "learning_rate": 6.531302026019457e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8408505141735076,
      "num_tokens": 58172442.0,
      "step": 48400
    },
    {
      "entropy": 1.9123701184988022,
      "epoch": 0.15006676436557578,
      "grad_norm": 7.481304168701172,
      "learning_rate": 6.530627395335206e-06,
      "loss": 0.5081,
      "mean_token_accuracy": 0.83611781001091,
      "num_tokens": 58184383.0,
      "step": 48410
    },
    {
      "entropy": 1.9331264093518257,
      "epoch": 0.15009776349062548,
      "grad_norm": 8.620382308959961,
      "learning_rate": 6.529952973659459e-06,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8453733563423157,
      "num_tokens": 58195683.0,
      "step": 48420
    },
    {
      "entropy": 1.8736334875226022,
      "epoch": 0.15012876261567518,
      "grad_norm": 7.600832939147949,
      "learning_rate": 6.52927876088432e-06,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8397956430912018,
      "num_tokens": 58207322.0,
      "step": 48430
    },
    {
      "entropy": 1.887729911506176,
      "epoch": 0.15015976174072487,
      "grad_norm": 9.004258155822754,
      "learning_rate": 6.5286047569019626e-06,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8307413533329964,
      "num_tokens": 58219063.0,
      "step": 48440
    },
    {
      "entropy": 1.8791867524385453,
      "epoch": 0.15019076086577457,
      "grad_norm": 8.612053871154785,
      "learning_rate": 6.5279309616046475e-06,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8294597789645195,
      "num_tokens": 58230800.0,
      "step": 48450
    },
    {
      "entropy": 1.8117628186941146,
      "epoch": 0.15022175999082427,
      "grad_norm": 7.665522575378418,
      "learning_rate": 6.527257374884704e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8396688923239708,
      "num_tokens": 58243429.0,
      "step": 48460
    },
    {
      "entropy": 1.961036217212677,
      "epoch": 0.15025275911587396,
      "grad_norm": 9.192258834838867,
      "learning_rate": 6.5265839966345466e-06,
      "loss": 0.5953,
      "mean_token_accuracy": 0.819936765730381,
      "num_tokens": 58254745.0,
      "step": 48470
    },
    {
      "entropy": 1.671479968726635,
      "epoch": 0.15028375824092366,
      "grad_norm": 4.934950828552246,
      "learning_rate": 6.5259108267466635e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8538110002875328,
      "num_tokens": 58270043.0,
      "step": 48480
    },
    {
      "entropy": 1.9253578931093216,
      "epoch": 0.15031475736597336,
      "grad_norm": 8.208556175231934,
      "learning_rate": 6.525237865113621e-06,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8374222457408905,
      "num_tokens": 58281799.0,
      "step": 48490
    },
    {
      "entropy": 1.8729837134480476,
      "epoch": 0.15034575649102305,
      "grad_norm": 9.998933792114258,
      "learning_rate": 6.524565111628065e-06,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8276918828487396,
      "num_tokens": 58293623.0,
      "step": 48500
    },
    {
      "entropy": 1.7981028646230697,
      "epoch": 0.15037675561607275,
      "grad_norm": 5.204981327056885,
      "learning_rate": 6.523892566182717e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8532559484243393,
      "num_tokens": 58306840.0,
      "step": 48510
    },
    {
      "entropy": 1.884388868510723,
      "epoch": 0.15040775474112242,
      "grad_norm": 9.6846342086792,
      "learning_rate": 6.523220228670375e-06,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8274782180786133,
      "num_tokens": 58318898.0,
      "step": 48520
    },
    {
      "entropy": 1.9351076558232307,
      "epoch": 0.15043875386617211,
      "grad_norm": 8.531515121459961,
      "learning_rate": 6.522548098983917e-06,
      "loss": 0.5733,
      "mean_token_accuracy": 0.8250018388032914,
      "num_tokens": 58330362.0,
      "step": 48530
    },
    {
      "entropy": 1.8293303191661834,
      "epoch": 0.1504697529912218,
      "grad_norm": 9.502824783325195,
      "learning_rate": 6.521876177016295e-06,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8435877069830895,
      "num_tokens": 58343242.0,
      "step": 48540
    },
    {
      "entropy": 1.890285351872444,
      "epoch": 0.1505007521162715,
      "grad_norm": 9.041160583496094,
      "learning_rate": 6.521204462660542e-06,
      "loss": 0.501,
      "mean_token_accuracy": 0.8346679538488389,
      "num_tokens": 58355866.0,
      "step": 48550
    },
    {
      "entropy": 1.8444904461503029,
      "epoch": 0.1505317512413212,
      "grad_norm": 3.8206639289855957,
      "learning_rate": 6.520532955809765e-06,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8404908493161202,
      "num_tokens": 58368640.0,
      "step": 48560
    },
    {
      "entropy": 1.835204230248928,
      "epoch": 0.1505627503663709,
      "grad_norm": 8.419897079467773,
      "learning_rate": 6.5198616563571505e-06,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8446251839399338,
      "num_tokens": 58381253.0,
      "step": 48570
    },
    {
      "entropy": 1.8946281239390372,
      "epoch": 0.1505937494914206,
      "grad_norm": 9.88506031036377,
      "learning_rate": 6.519190564195959e-06,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8362071871757507,
      "num_tokens": 58393895.0,
      "step": 48580
    },
    {
      "entropy": 1.82159626185894,
      "epoch": 0.1506247486164703,
      "grad_norm": 3.63238787651062,
      "learning_rate": 6.518519679219528e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8525608211755753,
      "num_tokens": 58407125.0,
      "step": 48590
    },
    {
      "entropy": 1.8428357735276222,
      "epoch": 0.15065574774152,
      "grad_norm": 9.18499755859375,
      "learning_rate": 6.517849001321278e-06,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8396696642041206,
      "num_tokens": 58420081.0,
      "step": 48600
    },
    {
      "entropy": 1.8873658314347268,
      "epoch": 0.1506867468665697,
      "grad_norm": 10.179847717285156,
      "learning_rate": 6.517178530394698e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8444565415382386,
      "num_tokens": 58432042.0,
      "step": 48610
    },
    {
      "entropy": 1.844616176187992,
      "epoch": 0.15071774599161938,
      "grad_norm": 3.9476633071899414,
      "learning_rate": 6.516508266333358e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.844041819870472,
      "num_tokens": 58444698.0,
      "step": 48620
    },
    {
      "entropy": 1.8296439558267594,
      "epoch": 0.15074874511666908,
      "grad_norm": 9.040117263793945,
      "learning_rate": 6.5158382090309035e-06,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8471789911389351,
      "num_tokens": 58456911.0,
      "step": 48630
    },
    {
      "entropy": 1.891487744450569,
      "epoch": 0.15077974424171878,
      "grad_norm": 8.067523956298828,
      "learning_rate": 6.515168358381057e-06,
      "loss": 0.5691,
      "mean_token_accuracy": 0.822045773267746,
      "num_tokens": 58468808.0,
      "step": 48640
    },
    {
      "entropy": 1.9148033902049064,
      "epoch": 0.15081074336676847,
      "grad_norm": 8.858370780944824,
      "learning_rate": 6.514498714277619e-06,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8241749078035354,
      "num_tokens": 58480696.0,
      "step": 48650
    },
    {
      "entropy": 1.9157651707530021,
      "epoch": 0.15084174249181817,
      "grad_norm": 9.38106918334961,
      "learning_rate": 6.5138292766144615e-06,
      "loss": 0.54,
      "mean_token_accuracy": 0.8287394896149636,
      "num_tokens": 58492144.0,
      "step": 48660
    },
    {
      "entropy": 1.9842921257019044,
      "epoch": 0.15087274161686787,
      "grad_norm": 8.547111511230469,
      "learning_rate": 6.5131600452855394e-06,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8356396153569221,
      "num_tokens": 58502452.0,
      "step": 48670
    },
    {
      "entropy": 1.972530573606491,
      "epoch": 0.15090374074191756,
      "grad_norm": 9.455718040466309,
      "learning_rate": 6.512491020184877e-06,
      "loss": 0.5894,
      "mean_token_accuracy": 0.8308357432484627,
      "num_tokens": 58513985.0,
      "step": 48680
    },
    {
      "entropy": 1.9101537883281707,
      "epoch": 0.15093473986696726,
      "grad_norm": 8.880309104919434,
      "learning_rate": 6.5118222012065825e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8405994296073913,
      "num_tokens": 58525140.0,
      "step": 48690
    },
    {
      "entropy": 1.946905219554901,
      "epoch": 0.15096573899201696,
      "grad_norm": 7.35371732711792,
      "learning_rate": 6.511153588244832e-06,
      "loss": 0.5481,
      "mean_token_accuracy": 0.8271761432290077,
      "num_tokens": 58536580.0,
      "step": 48700
    },
    {
      "entropy": 1.947824102640152,
      "epoch": 0.15099673811706665,
      "grad_norm": 9.245978355407715,
      "learning_rate": 6.510485181193884e-06,
      "loss": 0.6137,
      "mean_token_accuracy": 0.821645250916481,
      "num_tokens": 58548298.0,
      "step": 48710
    },
    {
      "entropy": 1.8544338196516037,
      "epoch": 0.15102773724211635,
      "grad_norm": 10.973575592041016,
      "learning_rate": 6.50981697994807e-06,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8355946585536003,
      "num_tokens": 58560605.0,
      "step": 48720
    },
    {
      "entropy": 1.8731532111763953,
      "epoch": 0.15105873636716605,
      "grad_norm": 8.846532821655273,
      "learning_rate": 6.5091489844017984e-06,
      "loss": 0.5354,
      "mean_token_accuracy": 0.838229563832283,
      "num_tokens": 58573583.0,
      "step": 48730
    },
    {
      "entropy": 1.8979535773396492,
      "epoch": 0.15108973549221574,
      "grad_norm": 8.72420883178711,
      "learning_rate": 6.5084811944495515e-06,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8275813281536102,
      "num_tokens": 58585343.0,
      "step": 48740
    },
    {
      "entropy": 1.9084653094410897,
      "epoch": 0.15112073461726544,
      "grad_norm": 8.924386978149414,
      "learning_rate": 6.50781360998589e-06,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8385712206363678,
      "num_tokens": 58596594.0,
      "step": 48750
    },
    {
      "entropy": 1.9596381425857543,
      "epoch": 0.15115173374231514,
      "grad_norm": 8.702342987060547,
      "learning_rate": 6.50714623090545e-06,
      "loss": 0.5923,
      "mean_token_accuracy": 0.8301157906651497,
      "num_tokens": 58607471.0,
      "step": 48760
    },
    {
      "entropy": 1.8487385302782058,
      "epoch": 0.1511827328673648,
      "grad_norm": 8.954679489135742,
      "learning_rate": 6.506479057102942e-06,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8388300389051437,
      "num_tokens": 58619562.0,
      "step": 48770
    },
    {
      "entropy": 1.910484978556633,
      "epoch": 0.1512137319924145,
      "grad_norm": 9.405644416809082,
      "learning_rate": 6.505812088473151e-06,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8201380014419556,
      "num_tokens": 58630856.0,
      "step": 48780
    },
    {
      "entropy": 1.816239494085312,
      "epoch": 0.1512447311174642,
      "grad_norm": 4.988278388977051,
      "learning_rate": 6.505145324910941e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8467860907316208,
      "num_tokens": 58643774.0,
      "step": 48790
    },
    {
      "entropy": 1.889238955080509,
      "epoch": 0.1512757302425139,
      "grad_norm": 8.861750602722168,
      "learning_rate": 6.504478766311248e-06,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8347546219825744,
      "num_tokens": 58655783.0,
      "step": 48800
    },
    {
      "entropy": 1.9085273638367652,
      "epoch": 0.1513067293675636,
      "grad_norm": 9.441707611083984,
      "learning_rate": 6.503812412569084e-06,
      "loss": 0.523,
      "mean_token_accuracy": 0.8369966968894005,
      "num_tokens": 58667819.0,
      "step": 48810
    },
    {
      "entropy": 1.9685264229774475,
      "epoch": 0.1513377284926133,
      "grad_norm": 10.141180038452148,
      "learning_rate": 6.503146263579539e-06,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8229795888066291,
      "num_tokens": 58679528.0,
      "step": 48820
    },
    {
      "entropy": 1.8758042559027672,
      "epoch": 0.15136872761766298,
      "grad_norm": 11.18869400024414,
      "learning_rate": 6.502480319237775e-06,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8300367653369903,
      "num_tokens": 58691911.0,
      "step": 48830
    },
    {
      "entropy": 1.8802335485816002,
      "epoch": 0.15139972674271268,
      "grad_norm": 8.065637588500977,
      "learning_rate": 6.5018145794390305e-06,
      "loss": 0.501,
      "mean_token_accuracy": 0.8387890234589577,
      "num_tokens": 58704390.0,
      "step": 48840
    },
    {
      "entropy": 1.8963167145848274,
      "epoch": 0.15143072586776238,
      "grad_norm": 5.054469585418701,
      "learning_rate": 6.501149044078618e-06,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8216091677546501,
      "num_tokens": 58717245.0,
      "step": 48850
    },
    {
      "entropy": 1.8458565592765808,
      "epoch": 0.15146172499281207,
      "grad_norm": 4.077347278594971,
      "learning_rate": 6.500483713051927e-06,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8441688820719719,
      "num_tokens": 58729499.0,
      "step": 48860
    },
    {
      "entropy": 1.9472675323486328,
      "epoch": 0.15149272411786177,
      "grad_norm": 8.823904991149902,
      "learning_rate": 6.499818586254422e-06,
      "loss": 0.513,
      "mean_token_accuracy": 0.8328737184405327,
      "num_tokens": 58741160.0,
      "step": 48870
    },
    {
      "entropy": 1.9384382277727128,
      "epoch": 0.15152372324291147,
      "grad_norm": 8.922835350036621,
      "learning_rate": 6.499153663581638e-06,
      "loss": 0.49,
      "mean_token_accuracy": 0.8393163681030273,
      "num_tokens": 58752947.0,
      "step": 48880
    },
    {
      "entropy": 1.9302797958254814,
      "epoch": 0.15155472236796116,
      "grad_norm": 8.566751480102539,
      "learning_rate": 6.49848894492919e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8295559108257293,
      "num_tokens": 58765340.0,
      "step": 48890
    },
    {
      "entropy": 1.8543498650193215,
      "epoch": 0.15158572149301086,
      "grad_norm": 8.253662109375,
      "learning_rate": 6.497824430192765e-06,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8355495229363441,
      "num_tokens": 58778000.0,
      "step": 48900
    },
    {
      "entropy": 1.874028617143631,
      "epoch": 0.15161672061806056,
      "grad_norm": 5.289228916168213,
      "learning_rate": 6.497160119268126e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8463690742850304,
      "num_tokens": 58791118.0,
      "step": 48910
    },
    {
      "entropy": 1.9725682616233826,
      "epoch": 0.15164771974311025,
      "grad_norm": 8.66529655456543,
      "learning_rate": 6.496496012051109e-06,
      "loss": 0.6059,
      "mean_token_accuracy": 0.8233085095882415,
      "num_tokens": 58802602.0,
      "step": 48920
    },
    {
      "entropy": 1.9239532873034477,
      "epoch": 0.15167871886815995,
      "grad_norm": 9.758983612060547,
      "learning_rate": 6.495832108437626e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8342737898230552,
      "num_tokens": 58814927.0,
      "step": 48930
    },
    {
      "entropy": 1.846892774105072,
      "epoch": 0.15170971799320965,
      "grad_norm": 4.085844993591309,
      "learning_rate": 6.495168408323665e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8468069404363632,
      "num_tokens": 58827118.0,
      "step": 48940
    },
    {
      "entropy": 1.9423293888568878,
      "epoch": 0.15174071711825934,
      "grad_norm": 9.48098373413086,
      "learning_rate": 6.4945049116052795e-06,
      "loss": 0.5816,
      "mean_token_accuracy": 0.821548655629158,
      "num_tokens": 58838717.0,
      "step": 48950
    },
    {
      "entropy": 1.9620625630021096,
      "epoch": 0.15177171624330904,
      "grad_norm": 7.364303112030029,
      "learning_rate": 6.493841618178611e-06,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8329650431871414,
      "num_tokens": 58850165.0,
      "step": 48960
    },
    {
      "entropy": 1.7987186387181282,
      "epoch": 0.15180271536835874,
      "grad_norm": 8.708568572998047,
      "learning_rate": 6.4931785279398666e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8545974805951119,
      "num_tokens": 58863232.0,
      "step": 48970
    },
    {
      "entropy": 1.92343827188015,
      "epoch": 0.15183371449340843,
      "grad_norm": 8.546026229858398,
      "learning_rate": 6.4925156407853275e-06,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8306387692689896,
      "num_tokens": 58874670.0,
      "step": 48980
    },
    {
      "entropy": 1.8594204097986222,
      "epoch": 0.15186471361845813,
      "grad_norm": 3.8005247116088867,
      "learning_rate": 6.491852956611351e-06,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8334858074784279,
      "num_tokens": 58887318.0,
      "step": 48990
    },
    {
      "entropy": 1.939525455236435,
      "epoch": 0.15189571274350783,
      "grad_norm": 7.96549129486084,
      "learning_rate": 6.4911904753143696e-06,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8367563143372536,
      "num_tokens": 58898840.0,
      "step": 49000
    },
    {
      "entropy": 1.8316703870892526,
      "epoch": 0.15192671186855752,
      "grad_norm": 7.305845737457275,
      "learning_rate": 6.490528196790886e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8301277205348014,
      "num_tokens": 58911661.0,
      "step": 49010
    },
    {
      "entropy": 1.808566428720951,
      "epoch": 0.1519577109936072,
      "grad_norm": 7.7132792472839355,
      "learning_rate": 6.489866120937483e-06,
      "loss": 0.4316,
      "mean_token_accuracy": 0.851980808377266,
      "num_tokens": 58924530.0,
      "step": 49020
    },
    {
      "entropy": 1.9115712746977807,
      "epoch": 0.1519887101186569,
      "grad_norm": 8.088088989257812,
      "learning_rate": 6.489204247650809e-06,
      "loss": 0.5556,
      "mean_token_accuracy": 0.832783767580986,
      "num_tokens": 58936710.0,
      "step": 49030
    },
    {
      "entropy": 1.7950451903045177,
      "epoch": 0.15201970924370659,
      "grad_norm": 7.494925022125244,
      "learning_rate": 6.4885425768275945e-06,
      "loss": 0.4237,
      "mean_token_accuracy": 0.8493603855371475,
      "num_tokens": 58950735.0,
      "step": 49040
    },
    {
      "entropy": 1.8894059650599957,
      "epoch": 0.15205070836875628,
      "grad_norm": 7.540571212768555,
      "learning_rate": 6.487881108364637e-06,
      "loss": 0.5165,
      "mean_token_accuracy": 0.835996463894844,
      "num_tokens": 58962971.0,
      "step": 49050
    },
    {
      "entropy": 1.7538513764739037,
      "epoch": 0.15208170749380598,
      "grad_norm": 4.750351905822754,
      "learning_rate": 6.487219842158812e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.8489253923296929,
      "num_tokens": 58976766.0,
      "step": 49060
    },
    {
      "entropy": 1.7878063380718232,
      "epoch": 0.15211270661885568,
      "grad_norm": 3.8629848957061768,
      "learning_rate": 6.486558778107066e-06,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8460006847977638,
      "num_tokens": 58990175.0,
      "step": 49070
    },
    {
      "entropy": 1.8979044020175935,
      "epoch": 0.15214370574390537,
      "grad_norm": 10.183228492736816,
      "learning_rate": 6.485897916106419e-06,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8209685668349266,
      "num_tokens": 59002434.0,
      "step": 49080
    },
    {
      "entropy": 1.936001867055893,
      "epoch": 0.15217470486895507,
      "grad_norm": 10.987131118774414,
      "learning_rate": 6.485237256053968e-06,
      "loss": 0.5699,
      "mean_token_accuracy": 0.8229846864938736,
      "num_tokens": 59013865.0,
      "step": 49090
    },
    {
      "entropy": 1.7930354595184326,
      "epoch": 0.15220570399400477,
      "grad_norm": 8.664037704467773,
      "learning_rate": 6.484576797846879e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8444106802344322,
      "num_tokens": 59026905.0,
      "step": 49100
    },
    {
      "entropy": 1.848856683075428,
      "epoch": 0.15223670311905446,
      "grad_norm": 8.302072525024414,
      "learning_rate": 6.4839165413823935e-06,
      "loss": 0.551,
      "mean_token_accuracy": 0.8347058981657028,
      "num_tokens": 59039328.0,
      "step": 49110
    },
    {
      "entropy": 1.754729336500168,
      "epoch": 0.15226770224410416,
      "grad_norm": 8.616592407226562,
      "learning_rate": 6.483256486557824e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8488777309656144,
      "num_tokens": 59052355.0,
      "step": 49120
    },
    {
      "entropy": 1.854713924229145,
      "epoch": 0.15229870136915385,
      "grad_norm": 10.268636703491211,
      "learning_rate": 6.482596633270561e-06,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8265779912471771,
      "num_tokens": 59063971.0,
      "step": 49130
    },
    {
      "entropy": 1.860026153922081,
      "epoch": 0.15232970049420355,
      "grad_norm": 7.569554805755615,
      "learning_rate": 6.481936981418064e-06,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8441597208380699,
      "num_tokens": 59075534.0,
      "step": 49140
    },
    {
      "entropy": 1.9424165233969688,
      "epoch": 0.15236069961925325,
      "grad_norm": 7.921428680419922,
      "learning_rate": 6.481277530897865e-06,
      "loss": 0.603,
      "mean_token_accuracy": 0.8199049532413483,
      "num_tokens": 59086811.0,
      "step": 49150
    },
    {
      "entropy": 1.862628909945488,
      "epoch": 0.15239169874430294,
      "grad_norm": 7.977451801300049,
      "learning_rate": 6.480618281607572e-06,
      "loss": 0.494,
      "mean_token_accuracy": 0.8402055114507675,
      "num_tokens": 59099011.0,
      "step": 49160
    },
    {
      "entropy": 1.8622022837400436,
      "epoch": 0.15242269786935264,
      "grad_norm": 8.97208309173584,
      "learning_rate": 6.479959233444862e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8342375844717026,
      "num_tokens": 59110548.0,
      "step": 49170
    },
    {
      "entropy": 1.8860869765281678,
      "epoch": 0.15245369699440234,
      "grad_norm": 8.571884155273438,
      "learning_rate": 6.47930038630749e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8293202951550483,
      "num_tokens": 59121640.0,
      "step": 49180
    },
    {
      "entropy": 1.8871512919664384,
      "epoch": 0.15248469611945203,
      "grad_norm": 8.496338844299316,
      "learning_rate": 6.478641740093281e-06,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8324317663908005,
      "num_tokens": 59133528.0,
      "step": 49190
    },
    {
      "entropy": 1.9237959653139114,
      "epoch": 0.15251569524450173,
      "grad_norm": 7.444377422332764,
      "learning_rate": 6.4779832947001306e-06,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8336405113339425,
      "num_tokens": 59144958.0,
      "step": 49200
    },
    {
      "entropy": 1.9096000641584396,
      "epoch": 0.15254669436955143,
      "grad_norm": 10.638242721557617,
      "learning_rate": 6.47732505002601e-06,
      "loss": 0.5257,
      "mean_token_accuracy": 0.836242513358593,
      "num_tokens": 59155880.0,
      "step": 49210
    },
    {
      "entropy": 1.895295462012291,
      "epoch": 0.15257769349460112,
      "grad_norm": 8.52206802368164,
      "learning_rate": 6.4766670059689615e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.834420631825924,
      "num_tokens": 59168200.0,
      "step": 49220
    },
    {
      "entropy": 1.8659065082669257,
      "epoch": 0.15260869261965082,
      "grad_norm": 8.12820816040039,
      "learning_rate": 6.476009162427102e-06,
      "loss": 0.6056,
      "mean_token_accuracy": 0.8158292233943939,
      "num_tokens": 59180782.0,
      "step": 49230
    },
    {
      "entropy": 1.8979418486356736,
      "epoch": 0.15263969174470052,
      "grad_norm": 10.203886985778809,
      "learning_rate": 6.475351519298617e-06,
      "loss": 0.5562,
      "mean_token_accuracy": 0.836780446767807,
      "num_tokens": 59192007.0,
      "step": 49240
    },
    {
      "entropy": 1.7612644746899604,
      "epoch": 0.15267069086975021,
      "grad_norm": 7.155702114105225,
      "learning_rate": 6.474694076481769e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8396282330155372,
      "num_tokens": 59205137.0,
      "step": 49250
    },
    {
      "entropy": 1.8855042546987533,
      "epoch": 0.15270168999479988,
      "grad_norm": 10.510095596313477,
      "learning_rate": 6.474036833874888e-06,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8316191598773003,
      "num_tokens": 59216398.0,
      "step": 49260
    },
    {
      "entropy": 1.9063111320137978,
      "epoch": 0.15273268911984958,
      "grad_norm": 8.204238891601562,
      "learning_rate": 6.4733797913763806e-06,
      "loss": 0.5644,
      "mean_token_accuracy": 0.8286448165774345,
      "num_tokens": 59228148.0,
      "step": 49270
    },
    {
      "entropy": 1.8491440996527673,
      "epoch": 0.15276368824489928,
      "grad_norm": 4.404919147491455,
      "learning_rate": 6.472722948884723e-06,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8485887929797172,
      "num_tokens": 59240379.0,
      "step": 49280
    },
    {
      "entropy": 1.8710645034909248,
      "epoch": 0.15279468736994897,
      "grad_norm": 4.196066856384277,
      "learning_rate": 6.472066306298462e-06,
      "loss": 0.5844,
      "mean_token_accuracy": 0.821666119992733,
      "num_tokens": 59251646.0,
      "step": 49290
    },
    {
      "entropy": 1.8651590749621392,
      "epoch": 0.15282568649499867,
      "grad_norm": 9.875333786010742,
      "learning_rate": 6.471409863516221e-06,
      "loss": 0.5306,
      "mean_token_accuracy": 0.8275208979845047,
      "num_tokens": 59263652.0,
      "step": 49300
    },
    {
      "entropy": 1.9215037196874618,
      "epoch": 0.15285668562004837,
      "grad_norm": 9.746886253356934,
      "learning_rate": 6.470753620436694e-06,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8245725408196449,
      "num_tokens": 59275186.0,
      "step": 49310
    },
    {
      "entropy": 1.9342552363872527,
      "epoch": 0.15288768474509806,
      "grad_norm": 8.793591499328613,
      "learning_rate": 6.470097576958641e-06,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8296951711177826,
      "num_tokens": 59286058.0,
      "step": 49320
    },
    {
      "entropy": 1.8291342303156852,
      "epoch": 0.15291868387014776,
      "grad_norm": 8.690282821655273,
      "learning_rate": 6.469441732980904e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8339086413383484,
      "num_tokens": 59298932.0,
      "step": 49330
    },
    {
      "entropy": 1.8255377933382988,
      "epoch": 0.15294968299519746,
      "grad_norm": 10.050409317016602,
      "learning_rate": 6.468786088402388e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8509842753410339,
      "num_tokens": 59311303.0,
      "step": 49340
    },
    {
      "entropy": 1.871800681948662,
      "epoch": 0.15298068212024715,
      "grad_norm": 8.856807708740234,
      "learning_rate": 6.468130643122074e-06,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8348386570811271,
      "num_tokens": 59322970.0,
      "step": 49350
    },
    {
      "entropy": 1.8413175642490387,
      "epoch": 0.15301168124529685,
      "grad_norm": 10.395611763000488,
      "learning_rate": 6.4674753970390126e-06,
      "loss": 0.5536,
      "mean_token_accuracy": 0.825497391819954,
      "num_tokens": 59335548.0,
      "step": 49360
    },
    {
      "entropy": 1.9431413426995277,
      "epoch": 0.15304268037034655,
      "grad_norm": 8.011473655700684,
      "learning_rate": 6.46682035005233e-06,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8263207510113716,
      "num_tokens": 59347158.0,
      "step": 49370
    },
    {
      "entropy": 1.8601020842790603,
      "epoch": 0.15307367949539624,
      "grad_norm": 10.249993324279785,
      "learning_rate": 6.466165502061217e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.833003306388855,
      "num_tokens": 59359361.0,
      "step": 49380
    },
    {
      "entropy": 1.8623602211475372,
      "epoch": 0.15310467862044594,
      "grad_norm": 9.118306159973145,
      "learning_rate": 6.465510852964943e-06,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8340568155050277,
      "num_tokens": 59371827.0,
      "step": 49390
    },
    {
      "entropy": 1.9150095269083978,
      "epoch": 0.15313567774549564,
      "grad_norm": 8.470681190490723,
      "learning_rate": 6.464856402662844e-06,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8367661386728287,
      "num_tokens": 59383425.0,
      "step": 49400
    },
    {
      "entropy": 1.9471410617232323,
      "epoch": 0.15316667687054533,
      "grad_norm": 3.6562201976776123,
      "learning_rate": 6.4642021510543284e-06,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8367534473538398,
      "num_tokens": 59394778.0,
      "step": 49410
    },
    {
      "entropy": 1.831431895494461,
      "epoch": 0.15319767599559503,
      "grad_norm": 7.588027000427246,
      "learning_rate": 6.463548098038879e-06,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8452267602086068,
      "num_tokens": 59407172.0,
      "step": 49420
    },
    {
      "entropy": 1.8925320595502853,
      "epoch": 0.15322867512064473,
      "grad_norm": 8.88277816772461,
      "learning_rate": 6.462894243516044e-06,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8345058739185334,
      "num_tokens": 59418806.0,
      "step": 49430
    },
    {
      "entropy": 1.8566480353474617,
      "epoch": 0.15325967424569442,
      "grad_norm": 4.175265789031982,
      "learning_rate": 6.462240587385448e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8374285340309143,
      "num_tokens": 59430897.0,
      "step": 49440
    },
    {
      "entropy": 1.8792894035577774,
      "epoch": 0.15329067337074412,
      "grad_norm": 11.97266674041748,
      "learning_rate": 6.461587129546784e-06,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8309213414788246,
      "num_tokens": 59442625.0,
      "step": 49450
    },
    {
      "entropy": 1.8707554250955583,
      "epoch": 0.15332167249579381,
      "grad_norm": 9.689730644226074,
      "learning_rate": 6.460933869899815e-06,
      "loss": 0.5504,
      "mean_token_accuracy": 0.8266577154397965,
      "num_tokens": 59454969.0,
      "step": 49460
    },
    {
      "entropy": 1.912343481183052,
      "epoch": 0.1533526716208435,
      "grad_norm": 9.543248176574707,
      "learning_rate": 6.460280808344378e-06,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8356236189603805,
      "num_tokens": 59467285.0,
      "step": 49470
    },
    {
      "entropy": 1.8799829974770546,
      "epoch": 0.1533836707458932,
      "grad_norm": 7.834064960479736,
      "learning_rate": 6.459627944780378e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8366628587245941,
      "num_tokens": 59480058.0,
      "step": 49480
    },
    {
      "entropy": 1.8075303509831429,
      "epoch": 0.1534146698709429,
      "grad_norm": 3.9849984645843506,
      "learning_rate": 6.458975279107794e-06,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8453790143132209,
      "num_tokens": 59493616.0,
      "step": 49490
    },
    {
      "entropy": 1.8907064393162727,
      "epoch": 0.1534456689959926,
      "grad_norm": 8.766815185546875,
      "learning_rate": 6.458322811226673e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.838210554420948,
      "num_tokens": 59506402.0,
      "step": 49500
    },
    {
      "entropy": 1.8980057820677758,
      "epoch": 0.15347666812104227,
      "grad_norm": 9.522225379943848,
      "learning_rate": 6.457670541037133e-06,
      "loss": 0.522,
      "mean_token_accuracy": 0.8347028121352196,
      "num_tokens": 59518065.0,
      "step": 49510
    },
    {
      "entropy": 1.9182281613349914,
      "epoch": 0.15350766724609197,
      "grad_norm": 8.175332069396973,
      "learning_rate": 6.457018468439363e-06,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8285543143749237,
      "num_tokens": 59528843.0,
      "step": 49520
    },
    {
      "entropy": 1.8754586443305015,
      "epoch": 0.15353866637114166,
      "grad_norm": 8.504158020019531,
      "learning_rate": 6.456366593333622e-06,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8261403411626815,
      "num_tokens": 59541901.0,
      "step": 49530
    },
    {
      "entropy": 1.816955418884754,
      "epoch": 0.15356966549619136,
      "grad_norm": 8.501500129699707,
      "learning_rate": 6.455714915620241e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.840816356241703,
      "num_tokens": 59555193.0,
      "step": 49540
    },
    {
      "entropy": 1.8059712588787078,
      "epoch": 0.15360066462124106,
      "grad_norm": 4.024796009063721,
      "learning_rate": 6.45506343519962e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8564706295728683,
      "num_tokens": 59568219.0,
      "step": 49550
    },
    {
      "entropy": 1.932602970302105,
      "epoch": 0.15363166374629075,
      "grad_norm": 8.853862762451172,
      "learning_rate": 6.4544121519722305e-06,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8353964149951935,
      "num_tokens": 59579262.0,
      "step": 49560
    },
    {
      "entropy": 1.8641065925359726,
      "epoch": 0.15366266287134045,
      "grad_norm": 7.992645263671875,
      "learning_rate": 6.453761065838612e-06,
      "loss": 0.5253,
      "mean_token_accuracy": 0.842236676812172,
      "num_tokens": 59590979.0,
      "step": 49570
    },
    {
      "entropy": 1.8351068049669266,
      "epoch": 0.15369366199639015,
      "grad_norm": 11.022222518920898,
      "learning_rate": 6.453110176699378e-06,
      "loss": 0.5175,
      "mean_token_accuracy": 0.836832246184349,
      "num_tokens": 59603268.0,
      "step": 49580
    },
    {
      "entropy": 1.8691863656044005,
      "epoch": 0.15372466112143984,
      "grad_norm": 10.182769775390625,
      "learning_rate": 6.452459484455208e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8315588355064392,
      "num_tokens": 59615230.0,
      "step": 49590
    },
    {
      "entropy": 1.867138534784317,
      "epoch": 0.15375566024648954,
      "grad_norm": 8.996567726135254,
      "learning_rate": 6.451808989006854e-06,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8165970057249069,
      "num_tokens": 59628275.0,
      "step": 49600
    },
    {
      "entropy": 1.8459146052598954,
      "epoch": 0.15378665937153924,
      "grad_norm": 4.567706108093262,
      "learning_rate": 6.451158690255139e-06,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8464686393737793,
      "num_tokens": 59640506.0,
      "step": 49610
    },
    {
      "entropy": 1.9115084454417228,
      "epoch": 0.15381765849658893,
      "grad_norm": 8.155994415283203,
      "learning_rate": 6.450508588100953e-06,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8274505391716958,
      "num_tokens": 59651786.0,
      "step": 49620
    },
    {
      "entropy": 1.8388926222920419,
      "epoch": 0.15384865762163863,
      "grad_norm": 3.681293487548828,
      "learning_rate": 6.449858682445258e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.842620424926281,
      "num_tokens": 59664048.0,
      "step": 49630
    },
    {
      "entropy": 1.836945366859436,
      "epoch": 0.15387965674668833,
      "grad_norm": 9.795785903930664,
      "learning_rate": 6.449208973189086e-06,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8397533014416695,
      "num_tokens": 59676318.0,
      "step": 49640
    },
    {
      "entropy": 1.9197256535291671,
      "epoch": 0.15391065587173802,
      "grad_norm": 9.240389823913574,
      "learning_rate": 6.448559460233536e-06,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8350172653794289,
      "num_tokens": 59686954.0,
      "step": 49650
    },
    {
      "entropy": 1.9680290162563323,
      "epoch": 0.15394165499678772,
      "grad_norm": 8.425545692443848,
      "learning_rate": 6.447910143479779e-06,
      "loss": 0.568,
      "mean_token_accuracy": 0.8262524694204331,
      "num_tokens": 59697663.0,
      "step": 49660
    },
    {
      "entropy": 1.8470780551433563,
      "epoch": 0.15397265412183742,
      "grad_norm": 4.306032180786133,
      "learning_rate": 6.447261022829057e-06,
      "loss": 0.4315,
      "mean_token_accuracy": 0.847578053176403,
      "num_tokens": 59710534.0,
      "step": 49670
    },
    {
      "entropy": 1.9209625497460365,
      "epoch": 0.1540036532468871,
      "grad_norm": 8.741031646728516,
      "learning_rate": 6.446612098182679e-06,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8277770847082138,
      "num_tokens": 59721380.0,
      "step": 49680
    },
    {
      "entropy": 1.7767624616622926,
      "epoch": 0.1540346523719368,
      "grad_norm": 9.159987449645996,
      "learning_rate": 6.445963369442024e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8386806547641754,
      "num_tokens": 59734270.0,
      "step": 49690
    },
    {
      "entropy": 1.9001009285449981,
      "epoch": 0.1540656514969865,
      "grad_norm": 13.77109432220459,
      "learning_rate": 6.4453148365085425e-06,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8450674623250961,
      "num_tokens": 59746960.0,
      "step": 49700
    },
    {
      "entropy": 1.9309724509716033,
      "epoch": 0.1540966506220362,
      "grad_norm": 4.395373821258545,
      "learning_rate": 6.444666499283752e-06,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8193286895751953,
      "num_tokens": 59758480.0,
      "step": 49710
    },
    {
      "entropy": 1.8494668424129486,
      "epoch": 0.1541276497470859,
      "grad_norm": 4.505528926849365,
      "learning_rate": 6.444018357669239e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8275418624281883,
      "num_tokens": 59772380.0,
      "step": 49720
    },
    {
      "entropy": 1.9033274337649346,
      "epoch": 0.1541586488721356,
      "grad_norm": 9.880914688110352,
      "learning_rate": 6.443370411566663e-06,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8315735951066017,
      "num_tokens": 59784089.0,
      "step": 49730
    },
    {
      "entropy": 1.888562636077404,
      "epoch": 0.1541896479971853,
      "grad_norm": 8.08756160736084,
      "learning_rate": 6.442722660877747e-06,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8301805481314659,
      "num_tokens": 59796171.0,
      "step": 49740
    },
    {
      "entropy": 1.90590338408947,
      "epoch": 0.154220647122235,
      "grad_norm": 9.111113548278809,
      "learning_rate": 6.44207510550429e-06,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8257271036505699,
      "num_tokens": 59807917.0,
      "step": 49750
    },
    {
      "entropy": 1.9074049085378646,
      "epoch": 0.15425164624728466,
      "grad_norm": 9.030037879943848,
      "learning_rate": 6.441427745348153e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8337802454829216,
      "num_tokens": 59818916.0,
      "step": 49760
    },
    {
      "entropy": 1.8746358096599578,
      "epoch": 0.15428264537233435,
      "grad_norm": 7.909918308258057,
      "learning_rate": 6.440780580311269e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8407538250088692,
      "num_tokens": 59830862.0,
      "step": 49770
    },
    {
      "entropy": 1.875384160876274,
      "epoch": 0.15431364449738405,
      "grad_norm": 8.324300765991211,
      "learning_rate": 6.4401336102956434e-06,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8312161207199097,
      "num_tokens": 59842486.0,
      "step": 49780
    },
    {
      "entropy": 1.9422027677297593,
      "epoch": 0.15434464362243375,
      "grad_norm": 9.501036643981934,
      "learning_rate": 6.439486835203346e-06,
      "loss": 0.5842,
      "mean_token_accuracy": 0.8298379242420196,
      "num_tokens": 59853228.0,
      "step": 49790
    },
    {
      "entropy": 1.854721449315548,
      "epoch": 0.15437564274748344,
      "grad_norm": 7.378803730010986,
      "learning_rate": 6.438840254936516e-06,
      "loss": 0.509,
      "mean_token_accuracy": 0.8368782594799995,
      "num_tokens": 59864876.0,
      "step": 49800
    },
    {
      "entropy": 1.8604190409183503,
      "epoch": 0.15440664187253314,
      "grad_norm": 11.659805297851562,
      "learning_rate": 6.438193869397364e-06,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8247652351856232,
      "num_tokens": 59877865.0,
      "step": 49810
    },
    {
      "entropy": 1.8469142317771912,
      "epoch": 0.15443764099758284,
      "grad_norm": 4.957024097442627,
      "learning_rate": 6.437547678488166e-06,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8259366884827614,
      "num_tokens": 59890976.0,
      "step": 49820
    },
    {
      "entropy": 1.8666848599910737,
      "epoch": 0.15446864012263253,
      "grad_norm": 9.470239639282227,
      "learning_rate": 6.436901682111268e-06,
      "loss": 0.4959,
      "mean_token_accuracy": 0.8429308265447617,
      "num_tokens": 59903295.0,
      "step": 49830
    },
    {
      "entropy": 1.9163474783301353,
      "epoch": 0.15449963924768223,
      "grad_norm": 9.075465202331543,
      "learning_rate": 6.436255880169087e-06,
      "loss": 0.5747,
      "mean_token_accuracy": 0.8243316933512688,
      "num_tokens": 59914765.0,
      "step": 49840
    },
    {
      "entropy": 1.8828782141208649,
      "epoch": 0.15453063837273193,
      "grad_norm": 9.262798309326172,
      "learning_rate": 6.4356102725641035e-06,
      "loss": 0.5203,
      "mean_token_accuracy": 0.837623517215252,
      "num_tokens": 59925835.0,
      "step": 49850
    },
    {
      "entropy": 1.8749229982495308,
      "epoch": 0.15456163749778162,
      "grad_norm": 8.3054780960083,
      "learning_rate": 6.434964859198871e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8357605487108231,
      "num_tokens": 59937496.0,
      "step": 49860
    },
    {
      "entropy": 1.9846446454524993,
      "epoch": 0.15459263662283132,
      "grad_norm": 7.847235679626465,
      "learning_rate": 6.434319639976007e-06,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8296759322285652,
      "num_tokens": 59948007.0,
      "step": 49870
    },
    {
      "entropy": 1.8958945728838443,
      "epoch": 0.15462363574788102,
      "grad_norm": 8.949000358581543,
      "learning_rate": 6.433674614798204e-06,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8374200582504272,
      "num_tokens": 59960116.0,
      "step": 49880
    },
    {
      "entropy": 1.8910530745983123,
      "epoch": 0.1546546348729307,
      "grad_norm": 11.521330833435059,
      "learning_rate": 6.433029783568216e-06,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8429587453603744,
      "num_tokens": 59971792.0,
      "step": 49890
    },
    {
      "entropy": 1.9622877299785615,
      "epoch": 0.1546856339979804,
      "grad_norm": 8.818026542663574,
      "learning_rate": 6.4323851461888694e-06,
      "loss": 0.5706,
      "mean_token_accuracy": 0.8334489464759827,
      "num_tokens": 59982734.0,
      "step": 49900
    },
    {
      "entropy": 1.9317353338003158,
      "epoch": 0.1547166331230301,
      "grad_norm": 9.778739929199219,
      "learning_rate": 6.431740702563056e-06,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8346163481473923,
      "num_tokens": 59993700.0,
      "step": 49910
    },
    {
      "entropy": 1.8633641496300697,
      "epoch": 0.1547476322480798,
      "grad_norm": 8.956582069396973,
      "learning_rate": 6.431096452593738e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8405205994844437,
      "num_tokens": 60005298.0,
      "step": 49920
    },
    {
      "entropy": 1.9589593350887298,
      "epoch": 0.1547786313731295,
      "grad_norm": 7.157927989959717,
      "learning_rate": 6.4304523961839436e-06,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8311053574085235,
      "num_tokens": 60016287.0,
      "step": 49930
    },
    {
      "entropy": 1.9007370486855506,
      "epoch": 0.1548096304981792,
      "grad_norm": 11.051389694213867,
      "learning_rate": 6.429808533236771e-06,
      "loss": 0.536,
      "mean_token_accuracy": 0.830623884499073,
      "num_tokens": 60028218.0,
      "step": 49940
    },
    {
      "entropy": 1.8595467865467072,
      "epoch": 0.1548406296232289,
      "grad_norm": 10.759744644165039,
      "learning_rate": 6.429164863655384e-06,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8303674578666687,
      "num_tokens": 60040116.0,
      "step": 49950
    },
    {
      "entropy": 1.8307231336832046,
      "epoch": 0.1548716287482786,
      "grad_norm": 9.683849334716797,
      "learning_rate": 6.428521387343016e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.847040268778801,
      "num_tokens": 60053053.0,
      "step": 49960
    },
    {
      "entropy": 1.855046309530735,
      "epoch": 0.15490262787332829,
      "grad_norm": 7.891843318939209,
      "learning_rate": 6.427878104202968e-06,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8327124953269959,
      "num_tokens": 60065345.0,
      "step": 49970
    },
    {
      "entropy": 1.9063360676169396,
      "epoch": 0.15493362699837798,
      "grad_norm": 10.191269874572754,
      "learning_rate": 6.4272350141386095e-06,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8289344042539597,
      "num_tokens": 60077605.0,
      "step": 49980
    },
    {
      "entropy": 1.8241621538996697,
      "epoch": 0.15496462612342768,
      "grad_norm": 10.132854461669922,
      "learning_rate": 6.4265921170533755e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8447253108024597,
      "num_tokens": 60089882.0,
      "step": 49990
    },
    {
      "entropy": 1.9241633802652358,
      "epoch": 0.15499562524847738,
      "grad_norm": 8.673702239990234,
      "learning_rate": 6.425949412850768e-06,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8289055705070496,
      "num_tokens": 60100890.0,
      "step": 50000
    },
    {
      "entropy": 1.9505061358213425,
      "epoch": 0.15502662437352704,
      "grad_norm": 8.550101280212402,
      "learning_rate": 6.4253069014343615e-06,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8236529782414437,
      "num_tokens": 60112909.0,
      "step": 50010
    },
    {
      "entropy": 1.9289822548627853,
      "epoch": 0.15505762349857674,
      "grad_norm": 8.745688438415527,
      "learning_rate": 6.424664582707793e-06,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8206963390111923,
      "num_tokens": 60124202.0,
      "step": 50020
    },
    {
      "entropy": 1.7902291625738145,
      "epoch": 0.15508862262362644,
      "grad_norm": 3.6163487434387207,
      "learning_rate": 6.424022456574768e-06,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8540971964597702,
      "num_tokens": 60137010.0,
      "step": 50030
    },
    {
      "entropy": 1.907183986902237,
      "epoch": 0.15511962174867613,
      "grad_norm": 9.091652870178223,
      "learning_rate": 6.42338052293906e-06,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8403370261192322,
      "num_tokens": 60148040.0,
      "step": 50040
    },
    {
      "entropy": 1.8011170402169228,
      "epoch": 0.15515062087372583,
      "grad_norm": 4.003586769104004,
      "learning_rate": 6.4227387817045115e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8478186964988709,
      "num_tokens": 60160541.0,
      "step": 50050
    },
    {
      "entropy": 1.8726102083921432,
      "epoch": 0.15518161999877553,
      "grad_norm": 9.244813919067383,
      "learning_rate": 6.42209723277503e-06,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8411934614181519,
      "num_tokens": 60171909.0,
      "step": 50060
    },
    {
      "entropy": 1.8079619467258454,
      "epoch": 0.15521261912382522,
      "grad_norm": 8.226319313049316,
      "learning_rate": 6.421455876054589e-06,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8530865862965584,
      "num_tokens": 60184414.0,
      "step": 50070
    },
    {
      "entropy": 1.9148479044437408,
      "epoch": 0.15524361824887492,
      "grad_norm": 9.0443696975708,
      "learning_rate": 6.420814711447232e-06,
      "loss": 0.549,
      "mean_token_accuracy": 0.8360567212104797,
      "num_tokens": 60196202.0,
      "step": 50080
    },
    {
      "entropy": 1.754614818096161,
      "epoch": 0.15527461737392462,
      "grad_norm": 9.89665412902832,
      "learning_rate": 6.42017373885707e-06,
      "loss": 0.4003,
      "mean_token_accuracy": 0.849841496348381,
      "num_tokens": 60209012.0,
      "step": 50090
    },
    {
      "entropy": 1.7963035687804223,
      "epoch": 0.1553056164989743,
      "grad_norm": 3.856391429901123,
      "learning_rate": 6.419532958188275e-06,
      "loss": 0.4344,
      "mean_token_accuracy": 0.8433475911617279,
      "num_tokens": 60221933.0,
      "step": 50100
    },
    {
      "entropy": 1.9359655752778053,
      "epoch": 0.155336615624024,
      "grad_norm": 8.836137771606445,
      "learning_rate": 6.418892369345093e-06,
      "loss": 0.5711,
      "mean_token_accuracy": 0.823367503285408,
      "num_tokens": 60232952.0,
      "step": 50110
    },
    {
      "entropy": 1.9027406126260757,
      "epoch": 0.1553676147490737,
      "grad_norm": 8.101412773132324,
      "learning_rate": 6.418251972231836e-06,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8346571624279022,
      "num_tokens": 60244389.0,
      "step": 50120
    },
    {
      "entropy": 1.8543214410543443,
      "epoch": 0.1553986138741234,
      "grad_norm": 4.314391613006592,
      "learning_rate": 6.417611766752878e-06,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8420602694153786,
      "num_tokens": 60256224.0,
      "step": 50130
    },
    {
      "entropy": 1.9107976794242858,
      "epoch": 0.1554296129991731,
      "grad_norm": 9.208541870117188,
      "learning_rate": 6.416971752812663e-06,
      "loss": 0.5544,
      "mean_token_accuracy": 0.8198239028453826,
      "num_tokens": 60267830.0,
      "step": 50140
    },
    {
      "entropy": 1.8322341233491897,
      "epoch": 0.1554606121242228,
      "grad_norm": 4.257328033447266,
      "learning_rate": 6.416331930315704e-06,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8367450326681137,
      "num_tokens": 60281007.0,
      "step": 50150
    },
    {
      "entropy": 1.9688672095537185,
      "epoch": 0.1554916112492725,
      "grad_norm": 7.649997711181641,
      "learning_rate": 6.415692299166574e-06,
      "loss": 0.6013,
      "mean_token_accuracy": 0.8196437805891037,
      "num_tokens": 60291878.0,
      "step": 50160
    },
    {
      "entropy": 1.9667943209409713,
      "epoch": 0.1555226103743222,
      "grad_norm": 9.782792091369629,
      "learning_rate": 6.41505285926992e-06,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8302391171455383,
      "num_tokens": 60302517.0,
      "step": 50170
    },
    {
      "entropy": 1.8588977128267288,
      "epoch": 0.1555536094993719,
      "grad_norm": 7.907834053039551,
      "learning_rate": 6.41441361053045e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8453133404254913,
      "num_tokens": 60314120.0,
      "step": 50180
    },
    {
      "entropy": 1.8701641455292701,
      "epoch": 0.15558460862442158,
      "grad_norm": 8.567540168762207,
      "learning_rate": 6.413774552852943e-06,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8220806911587715,
      "num_tokens": 60326477.0,
      "step": 50190
    },
    {
      "entropy": 1.8901820540428163,
      "epoch": 0.15561560774947128,
      "grad_norm": 13.7083740234375,
      "learning_rate": 6.41313568614224e-06,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8234784409403801,
      "num_tokens": 60338791.0,
      "step": 50200
    },
    {
      "entropy": 1.8873229175806046,
      "epoch": 0.15564660687452098,
      "grad_norm": 7.920546054840088,
      "learning_rate": 6.4124970103032505e-06,
      "loss": 0.5607,
      "mean_token_accuracy": 0.8268739849328994,
      "num_tokens": 60350606.0,
      "step": 50210
    },
    {
      "entropy": 1.8220305427908898,
      "epoch": 0.15567760599957067,
      "grad_norm": 9.152536392211914,
      "learning_rate": 6.411858525240952e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8429689288139344,
      "num_tokens": 60362867.0,
      "step": 50220
    },
    {
      "entropy": 1.9848755061626435,
      "epoch": 0.15570860512462037,
      "grad_norm": 8.138653755187988,
      "learning_rate": 6.411220230860381e-06,
      "loss": 0.5795,
      "mean_token_accuracy": 0.8322038248181343,
      "num_tokens": 60373606.0,
      "step": 50230
    },
    {
      "entropy": 1.9133560702204704,
      "epoch": 0.15573960424967007,
      "grad_norm": 8.208351135253906,
      "learning_rate": 6.410582127066652e-06,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8335413232445716,
      "num_tokens": 60385315.0,
      "step": 50240
    },
    {
      "entropy": 1.8043673783540726,
      "epoch": 0.15577060337471973,
      "grad_norm": 8.482664108276367,
      "learning_rate": 6.4099442137649356e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.8454105794429779,
      "num_tokens": 60398022.0,
      "step": 50250
    },
    {
      "entropy": 1.9254692614078521,
      "epoch": 0.15580160249976943,
      "grad_norm": 10.395545959472656,
      "learning_rate": 6.409306490860473e-06,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8282811924815178,
      "num_tokens": 60409280.0,
      "step": 50260
    },
    {
      "entropy": 1.8486803263425826,
      "epoch": 0.15583260162481913,
      "grad_norm": 10.327423095703125,
      "learning_rate": 6.408668958258571e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8356290921568871,
      "num_tokens": 60421673.0,
      "step": 50270
    },
    {
      "entropy": 1.8446889415383338,
      "epoch": 0.15586360074986882,
      "grad_norm": 10.985127449035645,
      "learning_rate": 6.408031615864598e-06,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8322935044765473,
      "num_tokens": 60434775.0,
      "step": 50280
    },
    {
      "entropy": 1.943482118844986,
      "epoch": 0.15589459987491852,
      "grad_norm": 9.617402076721191,
      "learning_rate": 6.407394463583996e-06,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8407940044999123,
      "num_tokens": 60445972.0,
      "step": 50290
    },
    {
      "entropy": 1.85861434340477,
      "epoch": 0.15592559899996822,
      "grad_norm": 9.958732604980469,
      "learning_rate": 6.406757501322266e-06,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8299056336283683,
      "num_tokens": 60458006.0,
      "step": 50300
    },
    {
      "entropy": 1.887197096645832,
      "epoch": 0.15595659812501791,
      "grad_norm": 11.390666961669922,
      "learning_rate": 6.406120728984979e-06,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8242016166448594,
      "num_tokens": 60470382.0,
      "step": 50310
    },
    {
      "entropy": 1.8762848794460296,
      "epoch": 0.1559875972500676,
      "grad_norm": 7.51845121383667,
      "learning_rate": 6.4054841464777696e-06,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8328775450587272,
      "num_tokens": 60482717.0,
      "step": 50320
    },
    {
      "entropy": 1.9675488024950027,
      "epoch": 0.1560185963751173,
      "grad_norm": 9.021512985229492,
      "learning_rate": 6.404847753706339e-06,
      "loss": 0.5867,
      "mean_token_accuracy": 0.8280714631080628,
      "num_tokens": 60493388.0,
      "step": 50330
    },
    {
      "entropy": 1.9901121139526368,
      "epoch": 0.156049595500167,
      "grad_norm": 8.092247009277344,
      "learning_rate": 6.404211550576453e-06,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8340759441256523,
      "num_tokens": 60504429.0,
      "step": 50340
    },
    {
      "entropy": 1.8783787608146667,
      "epoch": 0.1560805946252167,
      "grad_norm": 9.310537338256836,
      "learning_rate": 6.4035755369939425e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8348104074597359,
      "num_tokens": 60516387.0,
      "step": 50350
    },
    {
      "entropy": 1.8168802306056022,
      "epoch": 0.1561115937502664,
      "grad_norm": 3.784146785736084,
      "learning_rate": 6.4029397128647065e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8457161352038384,
      "num_tokens": 60529393.0,
      "step": 50360
    },
    {
      "entropy": 1.7647080093622207,
      "epoch": 0.1561425928753161,
      "grad_norm": 9.681499481201172,
      "learning_rate": 6.402304078094705e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8579833477735519,
      "num_tokens": 60543184.0,
      "step": 50370
    },
    {
      "entropy": 1.8571357518434524,
      "epoch": 0.1561735920003658,
      "grad_norm": 8.170286178588867,
      "learning_rate": 6.40166863258997e-06,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8396312475204468,
      "num_tokens": 60555841.0,
      "step": 50380
    },
    {
      "entropy": 1.9185105443000794,
      "epoch": 0.1562045911254155,
      "grad_norm": 7.637692451477051,
      "learning_rate": 6.401033376256593e-06,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8341576635837555,
      "num_tokens": 60567373.0,
      "step": 50390
    },
    {
      "entropy": 1.951904332637787,
      "epoch": 0.15623559025046518,
      "grad_norm": 7.780064105987549,
      "learning_rate": 6.40039830900073e-06,
      "loss": 0.5698,
      "mean_token_accuracy": 0.8260944783687592,
      "num_tokens": 60578359.0,
      "step": 50400
    },
    {
      "entropy": 1.937084037065506,
      "epoch": 0.15626658937551488,
      "grad_norm": 8.136171340942383,
      "learning_rate": 6.399763430728608e-06,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8284446790814399,
      "num_tokens": 60590095.0,
      "step": 50410
    },
    {
      "entropy": 1.8329094797372818,
      "epoch": 0.15629758850056458,
      "grad_norm": 7.345213890075684,
      "learning_rate": 6.399128741346514e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8410397842526436,
      "num_tokens": 60603043.0,
      "step": 50420
    },
    {
      "entropy": 1.8990901306271553,
      "epoch": 0.15632858762561427,
      "grad_norm": 8.570768356323242,
      "learning_rate": 6.398494240760803e-06,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8417127162218094,
      "num_tokens": 60614516.0,
      "step": 50430
    },
    {
      "entropy": 1.8423827588558197,
      "epoch": 0.15635958675066397,
      "grad_norm": 4.247179985046387,
      "learning_rate": 6.397859928877893e-06,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8417179524898529,
      "num_tokens": 60626871.0,
      "step": 50440
    },
    {
      "entropy": 2.0029925853013992,
      "epoch": 0.15639058587571367,
      "grad_norm": 8.819866180419922,
      "learning_rate": 6.3972258056042655e-06,
      "loss": 0.5682,
      "mean_token_accuracy": 0.829233068227768,
      "num_tokens": 60637756.0,
      "step": 50450
    },
    {
      "entropy": 1.9364984780550003,
      "epoch": 0.15642158500076336,
      "grad_norm": 9.162713050842285,
      "learning_rate": 6.396591870846475e-06,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8301840484142303,
      "num_tokens": 60648877.0,
      "step": 50460
    },
    {
      "entropy": 1.9391376033425332,
      "epoch": 0.15645258412581306,
      "grad_norm": 8.27598762512207,
      "learning_rate": 6.395958124511129e-06,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8327331587672233,
      "num_tokens": 60659919.0,
      "step": 50470
    },
    {
      "entropy": 1.9072021529078484,
      "epoch": 0.15648358325086276,
      "grad_norm": 9.113195419311523,
      "learning_rate": 6.395324566504908e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.831612104177475,
      "num_tokens": 60671627.0,
      "step": 50480
    },
    {
      "entropy": 1.9049764469265937,
      "epoch": 0.15651458237591245,
      "grad_norm": 4.894548416137695,
      "learning_rate": 6.394691196734555e-06,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8367709815502167,
      "num_tokens": 60683963.0,
      "step": 50490
    },
    {
      "entropy": 1.8629326492547988,
      "epoch": 0.15654558150096212,
      "grad_norm": 5.698431968688965,
      "learning_rate": 6.394058015106876e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.8405030608177185,
      "num_tokens": 60696888.0,
      "step": 50500
    },
    {
      "entropy": 1.9971718460321426,
      "epoch": 0.15657658062601182,
      "grad_norm": 7.774511337280273,
      "learning_rate": 6.393425021528746e-06,
      "loss": 0.5977,
      "mean_token_accuracy": 0.8220291540026665,
      "num_tokens": 60707257.0,
      "step": 50510
    },
    {
      "entropy": 1.9252537876367568,
      "epoch": 0.15660757975106152,
      "grad_norm": 4.02286434173584,
      "learning_rate": 6.392792215907099e-06,
      "loss": 0.51,
      "mean_token_accuracy": 0.8373642593622208,
      "num_tokens": 60719173.0,
      "step": 50520
    },
    {
      "entropy": 1.9036842539906502,
      "epoch": 0.1566385788761112,
      "grad_norm": 9.243278503417969,
      "learning_rate": 6.392159598148937e-06,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8288576260209084,
      "num_tokens": 60730887.0,
      "step": 50530
    },
    {
      "entropy": 1.950339911878109,
      "epoch": 0.1566695780011609,
      "grad_norm": 9.252776145935059,
      "learning_rate": 6.391527168161323e-06,
      "loss": 0.6262,
      "mean_token_accuracy": 0.8195764616131782,
      "num_tokens": 60742662.0,
      "step": 50540
    },
    {
      "entropy": 1.9394678846001625,
      "epoch": 0.1567005771262106,
      "grad_norm": 9.990178108215332,
      "learning_rate": 6.390894925851392e-06,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8261475265026093,
      "num_tokens": 60753958.0,
      "step": 50550
    },
    {
      "entropy": 1.8357860133051873,
      "epoch": 0.1567315762512603,
      "grad_norm": 6.805025100708008,
      "learning_rate": 6.390262871126333e-06,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8525337189435959,
      "num_tokens": 60766091.0,
      "step": 50560
    },
    {
      "entropy": 1.8700495898723601,
      "epoch": 0.15676257537631,
      "grad_norm": 8.728248596191406,
      "learning_rate": 6.3896310038934085e-06,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8417010590434074,
      "num_tokens": 60777610.0,
      "step": 50570
    },
    {
      "entropy": 1.8651820093393325,
      "epoch": 0.1567935745013597,
      "grad_norm": 9.503561019897461,
      "learning_rate": 6.388999324059937e-06,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8336687937378884,
      "num_tokens": 60789545.0,
      "step": 50580
    },
    {
      "entropy": 1.8762501239776612,
      "epoch": 0.1568245736264094,
      "grad_norm": 10.497973442077637,
      "learning_rate": 6.3883678315333085e-06,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8256103843450546,
      "num_tokens": 60801947.0,
      "step": 50590
    },
    {
      "entropy": 1.8612007051706314,
      "epoch": 0.1568555727514591,
      "grad_norm": 8.15131950378418,
      "learning_rate": 6.387736526220971e-06,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8382210031151771,
      "num_tokens": 60814283.0,
      "step": 50600
    },
    {
      "entropy": 1.7936832830309868,
      "epoch": 0.15688657187650878,
      "grad_norm": 8.772533416748047,
      "learning_rate": 6.387105408030442e-06,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8425097793340683,
      "num_tokens": 60827792.0,
      "step": 50610
    },
    {
      "entropy": 1.969840306043625,
      "epoch": 0.15691757100155848,
      "grad_norm": 8.165382385253906,
      "learning_rate": 6.386474476869298e-06,
      "loss": 0.5996,
      "mean_token_accuracy": 0.8277344390749931,
      "num_tokens": 60838984.0,
      "step": 50620
    },
    {
      "entropy": 1.8312163800001144,
      "epoch": 0.15694857012660818,
      "grad_norm": 3.5525267124176025,
      "learning_rate": 6.3858437326451805e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8387185871601105,
      "num_tokens": 60852094.0,
      "step": 50630
    },
    {
      "entropy": 1.8953583613038063,
      "epoch": 0.15697956925165787,
      "grad_norm": 9.432021141052246,
      "learning_rate": 6.3852131752658e-06,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8444772154092789,
      "num_tokens": 60863876.0,
      "step": 50640
    },
    {
      "entropy": 1.8836359083652496,
      "epoch": 0.15701056837670757,
      "grad_norm": 9.322160720825195,
      "learning_rate": 6.384582804638923e-06,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8311738058924675,
      "num_tokens": 60876801.0,
      "step": 50650
    },
    {
      "entropy": 1.92313295006752,
      "epoch": 0.15704156750175727,
      "grad_norm": 8.350409507751465,
      "learning_rate": 6.383952620672385e-06,
      "loss": 0.5855,
      "mean_token_accuracy": 0.8254999339580535,
      "num_tokens": 60887997.0,
      "step": 50660
    },
    {
      "entropy": 1.9507147312164306,
      "epoch": 0.15707256662680696,
      "grad_norm": 8.378862380981445,
      "learning_rate": 6.383322623274081e-06,
      "loss": 0.5762,
      "mean_token_accuracy": 0.8160288318991661,
      "num_tokens": 60899916.0,
      "step": 50670
    },
    {
      "entropy": 1.7835942827165128,
      "epoch": 0.15710356575185666,
      "grad_norm": 7.95631217956543,
      "learning_rate": 6.382692812351976e-06,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8510005518794059,
      "num_tokens": 60914587.0,
      "step": 50680
    },
    {
      "entropy": 1.8697744831442833,
      "epoch": 0.15713456487690636,
      "grad_norm": 4.098609924316406,
      "learning_rate": 6.382063187814093e-06,
      "loss": 0.4901,
      "mean_token_accuracy": 0.8406439542770385,
      "num_tokens": 60926953.0,
      "step": 50690
    },
    {
      "entropy": 1.8595498703420161,
      "epoch": 0.15716556400195605,
      "grad_norm": 8.888344764709473,
      "learning_rate": 6.381433749568522e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8456152617931366,
      "num_tokens": 60939063.0,
      "step": 50700
    },
    {
      "entropy": 1.8515797272324561,
      "epoch": 0.15719656312700575,
      "grad_norm": 6.374711513519287,
      "learning_rate": 6.380804497523409e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.844392117857933,
      "num_tokens": 60950485.0,
      "step": 50710
    },
    {
      "entropy": 1.8461545348167419,
      "epoch": 0.15722756225205545,
      "grad_norm": 9.635364532470703,
      "learning_rate": 6.380175431586977e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8438180416822434,
      "num_tokens": 60962994.0,
      "step": 50720
    },
    {
      "entropy": 1.9317549824714662,
      "epoch": 0.15725856137710514,
      "grad_norm": 9.940279960632324,
      "learning_rate": 6.379546551667498e-06,
      "loss": 0.5661,
      "mean_token_accuracy": 0.8308566614985466,
      "num_tokens": 60974445.0,
      "step": 50730
    },
    {
      "entropy": 1.919760164618492,
      "epoch": 0.15728956050215484,
      "grad_norm": 17.22686004638672,
      "learning_rate": 6.3789178576733166e-06,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8357348814606667,
      "num_tokens": 60985676.0,
      "step": 50740
    },
    {
      "entropy": 1.893989272415638,
      "epoch": 0.1573205596272045,
      "grad_norm": 8.417506217956543,
      "learning_rate": 6.378289349512838e-06,
      "loss": 0.5289,
      "mean_token_accuracy": 0.840349805355072,
      "num_tokens": 60997353.0,
      "step": 50750
    },
    {
      "entropy": 1.9457985952496528,
      "epoch": 0.1573515587522542,
      "grad_norm": 9.5839204788208,
      "learning_rate": 6.377661027094528e-06,
      "loss": 0.5644,
      "mean_token_accuracy": 0.8289625599980355,
      "num_tokens": 61008540.0,
      "step": 50760
    },
    {
      "entropy": 1.8832007065415381,
      "epoch": 0.1573825578773039,
      "grad_norm": 8.148798942565918,
      "learning_rate": 6.377032890326919e-06,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8396278038620949,
      "num_tokens": 61020653.0,
      "step": 50770
    },
    {
      "entropy": 1.8223789624869824,
      "epoch": 0.1574135570023536,
      "grad_norm": 8.812091827392578,
      "learning_rate": 6.376404939118606e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8405194833874703,
      "num_tokens": 61033783.0,
      "step": 50780
    },
    {
      "entropy": 1.908838202059269,
      "epoch": 0.1574445561274033,
      "grad_norm": 9.226709365844727,
      "learning_rate": 6.3757771733782435e-06,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8301179006695747,
      "num_tokens": 61045896.0,
      "step": 50790
    },
    {
      "entropy": 1.8385404348373413,
      "epoch": 0.157475555252453,
      "grad_norm": 3.6262776851654053,
      "learning_rate": 6.375149593014555e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.844711446762085,
      "num_tokens": 61059216.0,
      "step": 50800
    },
    {
      "entropy": 1.8200566530227662,
      "epoch": 0.1575065543775027,
      "grad_norm": 4.00075626373291,
      "learning_rate": 6.3745221979363226e-06,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8521258249878884,
      "num_tokens": 61072093.0,
      "step": 50810
    },
    {
      "entropy": 1.9493874102830886,
      "epoch": 0.15753755350255239,
      "grad_norm": 7.502165794372559,
      "learning_rate": 6.373894988052391e-06,
      "loss": 0.5919,
      "mean_token_accuracy": 0.8284584790468216,
      "num_tokens": 61083605.0,
      "step": 50820
    },
    {
      "entropy": 1.8854053810238838,
      "epoch": 0.15756855262760208,
      "grad_norm": 8.985443115234375,
      "learning_rate": 6.373267963271668e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8452239811420441,
      "num_tokens": 61095760.0,
      "step": 50830
    },
    {
      "entropy": 1.9016931489109994,
      "epoch": 0.15759955175265178,
      "grad_norm": 8.221643447875977,
      "learning_rate": 6.372641123503127e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8430635541677475,
      "num_tokens": 61108080.0,
      "step": 50840
    },
    {
      "entropy": 1.9211482793092727,
      "epoch": 0.15763055087770148,
      "grad_norm": 10.981581687927246,
      "learning_rate": 6.372014468655801e-06,
      "loss": 0.5592,
      "mean_token_accuracy": 0.825599467754364,
      "num_tokens": 61120063.0,
      "step": 50850
    },
    {
      "entropy": 1.948448945581913,
      "epoch": 0.15766155000275117,
      "grad_norm": 8.988844871520996,
      "learning_rate": 6.371387998638789e-06,
      "loss": 0.58,
      "mean_token_accuracy": 0.8233562901616096,
      "num_tokens": 61131416.0,
      "step": 50860
    },
    {
      "entropy": 1.923220631480217,
      "epoch": 0.15769254912780087,
      "grad_norm": 8.43247127532959,
      "learning_rate": 6.3707617133612456e-06,
      "loss": 0.5363,
      "mean_token_accuracy": 0.831841279566288,
      "num_tokens": 61143372.0,
      "step": 50870
    },
    {
      "entropy": 1.890692350268364,
      "epoch": 0.15772354825285057,
      "grad_norm": 4.421103477478027,
      "learning_rate": 6.370135612732394e-06,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8370485186576844,
      "num_tokens": 61155644.0,
      "step": 50880
    },
    {
      "entropy": 1.9700417637825012,
      "epoch": 0.15775454737790026,
      "grad_norm": 4.649320602416992,
      "learning_rate": 6.36950969666152e-06,
      "loss": 0.5609,
      "mean_token_accuracy": 0.828958623111248,
      "num_tokens": 61166750.0,
      "step": 50890
    },
    {
      "entropy": 1.8618547976017,
      "epoch": 0.15778554650294996,
      "grad_norm": 7.937415599822998,
      "learning_rate": 6.368883965057968e-06,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8421330273151397,
      "num_tokens": 61179716.0,
      "step": 50900
    },
    {
      "entropy": 1.9229750469326974,
      "epoch": 0.15781654562799965,
      "grad_norm": 7.491119861602783,
      "learning_rate": 6.368258417831149e-06,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8261383548378944,
      "num_tokens": 61191546.0,
      "step": 50910
    },
    {
      "entropy": 1.9325373709201812,
      "epoch": 0.15784754475304935,
      "grad_norm": 9.881427764892578,
      "learning_rate": 6.367633054890532e-06,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8323504284024239,
      "num_tokens": 61203143.0,
      "step": 50920
    },
    {
      "entropy": 1.8541507571935654,
      "epoch": 0.15787854387809905,
      "grad_norm": 4.566098690032959,
      "learning_rate": 6.367007876145651e-06,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8295344039797783,
      "num_tokens": 61216098.0,
      "step": 50930
    },
    {
      "entropy": 1.9541061252355576,
      "epoch": 0.15790954300314874,
      "grad_norm": 8.65689754486084,
      "learning_rate": 6.3663828815061e-06,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8173148021101951,
      "num_tokens": 61227492.0,
      "step": 50940
    },
    {
      "entropy": 1.9283887058496476,
      "epoch": 0.15794054212819844,
      "grad_norm": 8.396080017089844,
      "learning_rate": 6.36575807088154e-06,
      "loss": 0.513,
      "mean_token_accuracy": 0.8281655997037888,
      "num_tokens": 61239330.0,
      "step": 50950
    },
    {
      "entropy": 1.8624481573700904,
      "epoch": 0.15797154125324814,
      "grad_norm": 9.15777587890625,
      "learning_rate": 6.365133444181688e-06,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8356254741549491,
      "num_tokens": 61252054.0,
      "step": 50960
    },
    {
      "entropy": 1.9177945956587792,
      "epoch": 0.15800254037829783,
      "grad_norm": 9.632381439208984,
      "learning_rate": 6.364509001316326e-06,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8355124577879905,
      "num_tokens": 61263663.0,
      "step": 50970
    },
    {
      "entropy": 1.8173129588365555,
      "epoch": 0.15803353950334753,
      "grad_norm": 9.271222114562988,
      "learning_rate": 6.363884742195296e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8449875220656395,
      "num_tokens": 61277606.0,
      "step": 50980
    },
    {
      "entropy": 1.97777401804924,
      "epoch": 0.1580645386283972,
      "grad_norm": 11.567782402038574,
      "learning_rate": 6.363260666728507e-06,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8399723425507546,
      "num_tokens": 61287908.0,
      "step": 50990
    },
    {
      "entropy": 1.8491854876279832,
      "epoch": 0.1580955377534469,
      "grad_norm": 3.4037585258483887,
      "learning_rate": 6.362636774825923e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8377619609236717,
      "num_tokens": 61300804.0,
      "step": 51000
    },
    {
      "entropy": 1.9141796737909318,
      "epoch": 0.1581265368784966,
      "grad_norm": 4.402615547180176,
      "learning_rate": 6.362013066397575e-06,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8279877364635467,
      "num_tokens": 61312498.0,
      "step": 51010
    },
    {
      "entropy": 1.9528733968734742,
      "epoch": 0.1581575360035463,
      "grad_norm": 8.701175689697266,
      "learning_rate": 6.361389541353552e-06,
      "loss": 0.5713,
      "mean_token_accuracy": 0.836171668767929,
      "num_tokens": 61323367.0,
      "step": 51020
    },
    {
      "entropy": 1.8788547486066818,
      "epoch": 0.158188535128596,
      "grad_norm": 9.006681442260742,
      "learning_rate": 6.360766199604007e-06,
      "loss": 0.549,
      "mean_token_accuracy": 0.8283375024795532,
      "num_tokens": 61336492.0,
      "step": 51030
    },
    {
      "entropy": 1.856010665744543,
      "epoch": 0.15821953425364568,
      "grad_norm": 4.132041931152344,
      "learning_rate": 6.360143041059156e-06,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8421657636761666,
      "num_tokens": 61349803.0,
      "step": 51040
    },
    {
      "entropy": 1.8853138819336892,
      "epoch": 0.15825053337869538,
      "grad_norm": 4.42047643661499,
      "learning_rate": 6.359520065629272e-06,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8277313068509102,
      "num_tokens": 61361430.0,
      "step": 51050
    },
    {
      "entropy": 1.8170242205262184,
      "epoch": 0.15828153250374508,
      "grad_norm": 3.389427423477173,
      "learning_rate": 6.358897273224693e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8529497593641281,
      "num_tokens": 61374400.0,
      "step": 51060
    },
    {
      "entropy": 1.8062128841876983,
      "epoch": 0.15831253162879477,
      "grad_norm": 3.821852207183838,
      "learning_rate": 6.358274663755817e-06,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8493776023387909,
      "num_tokens": 61388410.0,
      "step": 51070
    },
    {
      "entropy": 1.904846543073654,
      "epoch": 0.15834353075384447,
      "grad_norm": 11.317872047424316,
      "learning_rate": 6.357652237133105e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.830742597579956,
      "num_tokens": 61399765.0,
      "step": 51080
    },
    {
      "entropy": 1.9227543294429779,
      "epoch": 0.15837452987889417,
      "grad_norm": 9.054058074951172,
      "learning_rate": 6.357029993267079e-06,
      "loss": 0.594,
      "mean_token_accuracy": 0.8182252869009972,
      "num_tokens": 61411205.0,
      "step": 51090
    },
    {
      "entropy": 1.7773767858743668,
      "epoch": 0.15840552900394386,
      "grad_norm": 3.7279295921325684,
      "learning_rate": 6.356407932068319e-06,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8401523649692535,
      "num_tokens": 61425116.0,
      "step": 51100
    },
    {
      "entropy": 1.932927194237709,
      "epoch": 0.15843652812899356,
      "grad_norm": 8.091573715209961,
      "learning_rate": 6.35578605344747e-06,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8386119574308395,
      "num_tokens": 61436160.0,
      "step": 51110
    },
    {
      "entropy": 1.9222755640745164,
      "epoch": 0.15846752725404326,
      "grad_norm": 9.271939277648926,
      "learning_rate": 6.355164357315238e-06,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8323445156216621,
      "num_tokens": 61447404.0,
      "step": 51120
    },
    {
      "entropy": 1.8717841893434524,
      "epoch": 0.15849852637909295,
      "grad_norm": 4.64661979675293,
      "learning_rate": 6.354542843582387e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8282085090875626,
      "num_tokens": 61459307.0,
      "step": 51130
    },
    {
      "entropy": 1.7520866304636002,
      "epoch": 0.15852952550414265,
      "grad_norm": 8.69471263885498,
      "learning_rate": 6.353921512159747e-06,
      "loss": 0.392,
      "mean_token_accuracy": 0.8573095768690109,
      "num_tokens": 61473555.0,
      "step": 51140
    },
    {
      "entropy": 1.933935022354126,
      "epoch": 0.15856052462919235,
      "grad_norm": 9.661917686462402,
      "learning_rate": 6.353300362958204e-06,
      "loss": 0.5802,
      "mean_token_accuracy": 0.8254079014062882,
      "num_tokens": 61485440.0,
      "step": 51150
    },
    {
      "entropy": 1.8875377878546715,
      "epoch": 0.15859152375424204,
      "grad_norm": 9.894033432006836,
      "learning_rate": 6.352679395888709e-06,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8381941422820092,
      "num_tokens": 61497948.0,
      "step": 51160
    },
    {
      "entropy": 1.954133327305317,
      "epoch": 0.15862252287929174,
      "grad_norm": 13.155253410339355,
      "learning_rate": 6.3520586108622695e-06,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8287410020828248,
      "num_tokens": 61509102.0,
      "step": 51170
    },
    {
      "entropy": 1.9011639848351478,
      "epoch": 0.15865352200434144,
      "grad_norm": 9.405858039855957,
      "learning_rate": 6.351438007789959e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.8395056575536728,
      "num_tokens": 61521308.0,
      "step": 51180
    },
    {
      "entropy": 1.9040748074650764,
      "epoch": 0.15868452112939113,
      "grad_norm": 8.241883277893066,
      "learning_rate": 6.350817586582909e-06,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8309203192591668,
      "num_tokens": 61533828.0,
      "step": 51190
    },
    {
      "entropy": 1.9427318632602693,
      "epoch": 0.15871552025444083,
      "grad_norm": 8.195354461669922,
      "learning_rate": 6.35019734715231e-06,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8373890981078148,
      "num_tokens": 61545278.0,
      "step": 51200
    },
    {
      "entropy": 1.904130506515503,
      "epoch": 0.15874651937949052,
      "grad_norm": 7.979343891143799,
      "learning_rate": 6.349577289409418e-06,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8309107288718224,
      "num_tokens": 61557125.0,
      "step": 51210
    },
    {
      "entropy": 1.9387200504541398,
      "epoch": 0.15877751850454022,
      "grad_norm": 8.9723539352417,
      "learning_rate": 6.348957413265544e-06,
      "loss": 0.5938,
      "mean_token_accuracy": 0.8150834292173386,
      "num_tokens": 61569250.0,
      "step": 51220
    },
    {
      "entropy": 1.8315011352300643,
      "epoch": 0.15880851762958992,
      "grad_norm": 4.4107136726379395,
      "learning_rate": 6.348337718632065e-06,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8472426652908325,
      "num_tokens": 61582476.0,
      "step": 51230
    },
    {
      "entropy": 1.8459465608000756,
      "epoch": 0.1588395167546396,
      "grad_norm": 8.931220054626465,
      "learning_rate": 6.347718205420413e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8469175025820732,
      "num_tokens": 61595011.0,
      "step": 51240
    },
    {
      "entropy": 1.9565140813589097,
      "epoch": 0.15887051587968928,
      "grad_norm": 8.753551483154297,
      "learning_rate": 6.347098873542088e-06,
      "loss": 0.5644,
      "mean_token_accuracy": 0.8283399358391762,
      "num_tokens": 61606389.0,
      "step": 51250
    },
    {
      "entropy": 1.9582069575786591,
      "epoch": 0.15890151500473898,
      "grad_norm": 7.78233003616333,
      "learning_rate": 6.346479722908642e-06,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8336837723851204,
      "num_tokens": 61617999.0,
      "step": 51260
    },
    {
      "entropy": 1.9196703389286995,
      "epoch": 0.15893251412978868,
      "grad_norm": 8.762555122375488,
      "learning_rate": 6.345860753431693e-06,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8311418369412422,
      "num_tokens": 61630045.0,
      "step": 51270
    },
    {
      "entropy": 1.9337300062179565,
      "epoch": 0.15896351325483837,
      "grad_norm": 9.353802680969238,
      "learning_rate": 6.345241965022917e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8350248277187348,
      "num_tokens": 61642241.0,
      "step": 51280
    },
    {
      "entropy": 1.903528119623661,
      "epoch": 0.15899451237988807,
      "grad_norm": 8.561066627502441,
      "learning_rate": 6.344623357594051e-06,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8348277062177658,
      "num_tokens": 61654226.0,
      "step": 51290
    },
    {
      "entropy": 1.9056920766830445,
      "epoch": 0.15902551150493777,
      "grad_norm": 4.238293170928955,
      "learning_rate": 6.344004931056894e-06,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8365932151675224,
      "num_tokens": 61666601.0,
      "step": 51300
    },
    {
      "entropy": 1.900420580804348,
      "epoch": 0.15905651062998746,
      "grad_norm": 8.474109649658203,
      "learning_rate": 6.343386685323301e-06,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8356752887368202,
      "num_tokens": 61678955.0,
      "step": 51310
    },
    {
      "entropy": 1.7979067623615266,
      "epoch": 0.15908750975503716,
      "grad_norm": 7.289546012878418,
      "learning_rate": 6.34276862030519e-06,
      "loss": 0.4091,
      "mean_token_accuracy": 0.8524860993027688,
      "num_tokens": 61692933.0,
      "step": 51320
    },
    {
      "entropy": 1.9506612807512282,
      "epoch": 0.15911850888008686,
      "grad_norm": 7.817605018615723,
      "learning_rate": 6.342150735914539e-06,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8349381595849991,
      "num_tokens": 61703549.0,
      "step": 51330
    },
    {
      "entropy": 1.9035085812211037,
      "epoch": 0.15914950800513655,
      "grad_norm": 8.978363037109375,
      "learning_rate": 6.341533032063384e-06,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8332462772727013,
      "num_tokens": 61715518.0,
      "step": 51340
    },
    {
      "entropy": 1.9238849043846131,
      "epoch": 0.15918050713018625,
      "grad_norm": 10.23372745513916,
      "learning_rate": 6.3409155086638244e-06,
      "loss": 0.5827,
      "mean_token_accuracy": 0.8245359167456627,
      "num_tokens": 61726468.0,
      "step": 51350
    },
    {
      "entropy": 1.912158764898777,
      "epoch": 0.15921150625523595,
      "grad_norm": 8.236334800720215,
      "learning_rate": 6.3402981656280174e-06,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8354252070188523,
      "num_tokens": 61738604.0,
      "step": 51360
    },
    {
      "entropy": 1.9760198339819908,
      "epoch": 0.15924250538028564,
      "grad_norm": 9.30893611907959,
      "learning_rate": 6.339681002868179e-06,
      "loss": 0.561,
      "mean_token_accuracy": 0.8204899609088898,
      "num_tokens": 61749706.0,
      "step": 51370
    },
    {
      "entropy": 1.9051903694868089,
      "epoch": 0.15927350450533534,
      "grad_norm": 9.52764892578125,
      "learning_rate": 6.3390640202965856e-06,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8380590423941612,
      "num_tokens": 61761965.0,
      "step": 51380
    },
    {
      "entropy": 1.9080309465527534,
      "epoch": 0.15930450363038504,
      "grad_norm": 3.284393072128296,
      "learning_rate": 6.338447217825577e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8422815844416618,
      "num_tokens": 61773196.0,
      "step": 51390
    },
    {
      "entropy": 1.848653295636177,
      "epoch": 0.15933550275543473,
      "grad_norm": 9.541685104370117,
      "learning_rate": 6.337830595367548e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.835855670273304,
      "num_tokens": 61785726.0,
      "step": 51400
    },
    {
      "entropy": 1.8705054201185702,
      "epoch": 0.15936650188048443,
      "grad_norm": 8.726213455200195,
      "learning_rate": 6.337214152834954e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.838579186797142,
      "num_tokens": 61798381.0,
      "step": 51410
    },
    {
      "entropy": 1.9128465965390204,
      "epoch": 0.15939750100553413,
      "grad_norm": 4.461379528045654,
      "learning_rate": 6.336597890140311e-06,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8308456495404244,
      "num_tokens": 61809810.0,
      "step": 51420
    },
    {
      "entropy": 1.8830090552568435,
      "epoch": 0.15942850013058382,
      "grad_norm": 7.6069135665893555,
      "learning_rate": 6.335981807196195e-06,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8455763354897499,
      "num_tokens": 61821564.0,
      "step": 51430
    },
    {
      "entropy": 1.879130232334137,
      "epoch": 0.15945949925563352,
      "grad_norm": 8.862757682800293,
      "learning_rate": 6.335365903915241e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.841680321097374,
      "num_tokens": 61833721.0,
      "step": 51440
    },
    {
      "entropy": 1.8420701980590821,
      "epoch": 0.15949049838068322,
      "grad_norm": 9.544261932373047,
      "learning_rate": 6.334750180210142e-06,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8434771358966827,
      "num_tokens": 61845911.0,
      "step": 51450
    },
    {
      "entropy": 1.8399369686841964,
      "epoch": 0.1595214975057329,
      "grad_norm": 8.646150588989258,
      "learning_rate": 6.334134635993651e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8362537592649459,
      "num_tokens": 61858452.0,
      "step": 51460
    },
    {
      "entropy": 1.8282588481903077,
      "epoch": 0.1595524966307826,
      "grad_norm": 8.844307899475098,
      "learning_rate": 6.333519271178583e-06,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8485338285565376,
      "num_tokens": 61871207.0,
      "step": 51470
    },
    {
      "entropy": 1.9945002496242523,
      "epoch": 0.1595834957558323,
      "grad_norm": 9.600845336914062,
      "learning_rate": 6.332904085677809e-06,
      "loss": 0.6231,
      "mean_token_accuracy": 0.8192642524838447,
      "num_tokens": 61881849.0,
      "step": 51480
    },
    {
      "entropy": 1.866794577240944,
      "epoch": 0.15961449488088197,
      "grad_norm": 8.577037811279297,
      "learning_rate": 6.33228907940426e-06,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8294419586658478,
      "num_tokens": 61894411.0,
      "step": 51490
    },
    {
      "entropy": 1.9821971401572227,
      "epoch": 0.15964549400593167,
      "grad_norm": 7.966568470001221,
      "learning_rate": 6.3316742522709295e-06,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8234156042337417,
      "num_tokens": 61905974.0,
      "step": 51500
    },
    {
      "entropy": 1.928943158686161,
      "epoch": 0.15967649313098137,
      "grad_norm": 8.466790199279785,
      "learning_rate": 6.331059604190863e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.83587566614151,
      "num_tokens": 61918230.0,
      "step": 51510
    },
    {
      "entropy": 1.8466455608606338,
      "epoch": 0.15970749225603106,
      "grad_norm": 8.579726219177246,
      "learning_rate": 6.330445135077171e-06,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8351440250873565,
      "num_tokens": 61930533.0,
      "step": 51520
    },
    {
      "entropy": 1.9407551258802413,
      "epoch": 0.15973849138108076,
      "grad_norm": 9.855902671813965,
      "learning_rate": 6.329830844843021e-06,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8276364892721176,
      "num_tokens": 61941838.0,
      "step": 51530
    },
    {
      "entropy": 1.8404507592320443,
      "epoch": 0.15976949050613046,
      "grad_norm": 7.8672261238098145,
      "learning_rate": 6.329216733401641e-06,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8361625537276268,
      "num_tokens": 61954790.0,
      "step": 51540
    },
    {
      "entropy": 1.8961133405566215,
      "epoch": 0.15980048963118015,
      "grad_norm": 9.657519340515137,
      "learning_rate": 6.328602800666316e-06,
      "loss": 0.538,
      "mean_token_accuracy": 0.8254346460103988,
      "num_tokens": 61967113.0,
      "step": 51550
    },
    {
      "entropy": 1.887896779179573,
      "epoch": 0.15983148875622985,
      "grad_norm": 7.927542209625244,
      "learning_rate": 6.32798904655039e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8375117152929306,
      "num_tokens": 61979299.0,
      "step": 51560
    },
    {
      "entropy": 1.8948641210794448,
      "epoch": 0.15986248788127955,
      "grad_norm": 7.149417400360107,
      "learning_rate": 6.327375470967267e-06,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8403451159596443,
      "num_tokens": 61990608.0,
      "step": 51570
    },
    {
      "entropy": 1.8972643151879311,
      "epoch": 0.15989348700632924,
      "grad_norm": 10.748580932617188,
      "learning_rate": 6.326762073830408e-06,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8346004873514176,
      "num_tokens": 62002236.0,
      "step": 51580
    },
    {
      "entropy": 1.9104709938168525,
      "epoch": 0.15992448613137894,
      "grad_norm": 9.303431510925293,
      "learning_rate": 6.326148855053335e-06,
      "loss": 0.542,
      "mean_token_accuracy": 0.8291504085063934,
      "num_tokens": 62013316.0,
      "step": 51590
    },
    {
      "entropy": 1.9200339168310165,
      "epoch": 0.15995548525642864,
      "grad_norm": 8.650131225585938,
      "learning_rate": 6.325535814549628e-06,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8344766482710838,
      "num_tokens": 62024979.0,
      "step": 51600
    },
    {
      "entropy": 1.9062774524092674,
      "epoch": 0.15998648438147833,
      "grad_norm": 9.20338249206543,
      "learning_rate": 6.324922952232924e-06,
      "loss": 0.5443,
      "mean_token_accuracy": 0.8257553368806839,
      "num_tokens": 62037183.0,
      "step": 51610
    },
    {
      "entropy": 1.8092211425304412,
      "epoch": 0.16001748350652803,
      "grad_norm": 8.952025413513184,
      "learning_rate": 6.32431026801692e-06,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8433082342147827,
      "num_tokens": 62050479.0,
      "step": 51620
    },
    {
      "entropy": 1.9418270736932755,
      "epoch": 0.16004848263157773,
      "grad_norm": 9.070456504821777,
      "learning_rate": 6.3236977618153725e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8335362508893013,
      "num_tokens": 62062292.0,
      "step": 51630
    },
    {
      "entropy": 1.8717383340001106,
      "epoch": 0.16007948175662742,
      "grad_norm": 4.403903961181641,
      "learning_rate": 6.323085433542092e-06,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8389642179012299,
      "num_tokens": 62074628.0,
      "step": 51640
    },
    {
      "entropy": 1.9106849119067193,
      "epoch": 0.16011048088167712,
      "grad_norm": 9.37865924835205,
      "learning_rate": 6.3224732831109535e-06,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8324133858084679,
      "num_tokens": 62086818.0,
      "step": 51650
    },
    {
      "entropy": 1.885637989640236,
      "epoch": 0.16014148000672682,
      "grad_norm": 4.711241722106934,
      "learning_rate": 6.321861310435887e-06,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8371865153312683,
      "num_tokens": 62098896.0,
      "step": 51660
    },
    {
      "entropy": 1.9405640929937362,
      "epoch": 0.1601724791317765,
      "grad_norm": 8.641385078430176,
      "learning_rate": 6.3212495154308804e-06,
      "loss": 0.5958,
      "mean_token_accuracy": 0.8278973504900933,
      "num_tokens": 62109979.0,
      "step": 51670
    },
    {
      "entropy": 1.934703852236271,
      "epoch": 0.1602034782568262,
      "grad_norm": 9.276317596435547,
      "learning_rate": 6.32063789800998e-06,
      "loss": 0.5185,
      "mean_token_accuracy": 0.8384066224098206,
      "num_tokens": 62121358.0,
      "step": 51680
    },
    {
      "entropy": 1.885528053343296,
      "epoch": 0.1602344773818759,
      "grad_norm": 8.702717781066895,
      "learning_rate": 6.320026458087292e-06,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8399693235754967,
      "num_tokens": 62132820.0,
      "step": 51690
    },
    {
      "entropy": 1.955546210706234,
      "epoch": 0.1602654765069256,
      "grad_norm": 9.13160228729248,
      "learning_rate": 6.319415195576981e-06,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8254757478833199,
      "num_tokens": 62143735.0,
      "step": 51700
    },
    {
      "entropy": 1.9146698236465454,
      "epoch": 0.1602964756319753,
      "grad_norm": 8.36129379272461,
      "learning_rate": 6.318804110393267e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8527568742632866,
      "num_tokens": 62155133.0,
      "step": 51710
    },
    {
      "entropy": 1.859630098938942,
      "epoch": 0.160327474757025,
      "grad_norm": 4.852166175842285,
      "learning_rate": 6.318193202450428e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8412252008914948,
      "num_tokens": 62167946.0,
      "step": 51720
    },
    {
      "entropy": 1.9424054473638535,
      "epoch": 0.16035847388207466,
      "grad_norm": 10.02141284942627,
      "learning_rate": 6.317582471662803e-06,
      "loss": 0.5314,
      "mean_token_accuracy": 0.839588788151741,
      "num_tokens": 62179109.0,
      "step": 51730
    },
    {
      "entropy": 1.9220136627554893,
      "epoch": 0.16038947300712436,
      "grad_norm": 12.059248924255371,
      "learning_rate": 6.3169719179447885e-06,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8367831781506538,
      "num_tokens": 62190468.0,
      "step": 51740
    },
    {
      "entropy": 1.85854529440403,
      "epoch": 0.16042047213217406,
      "grad_norm": 9.116106986999512,
      "learning_rate": 6.316361541210837e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8309743300080299,
      "num_tokens": 62202630.0,
      "step": 51750
    },
    {
      "entropy": 1.9260679721832275,
      "epoch": 0.16045147125722375,
      "grad_norm": 9.218046188354492,
      "learning_rate": 6.315751341375458e-06,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8390769004821778,
      "num_tokens": 62214408.0,
      "step": 51760
    },
    {
      "entropy": 1.9168349742889403,
      "epoch": 0.16048247038227345,
      "grad_norm": 8.362269401550293,
      "learning_rate": 6.315141318353224e-06,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8344539046287537,
      "num_tokens": 62226824.0,
      "step": 51770
    },
    {
      "entropy": 1.847986987233162,
      "epoch": 0.16051346950732315,
      "grad_norm": 8.942157745361328,
      "learning_rate": 6.314531472058758e-06,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8420057892799377,
      "num_tokens": 62239194.0,
      "step": 51780
    },
    {
      "entropy": 1.890682005882263,
      "epoch": 0.16054446863237284,
      "grad_norm": 8.703591346740723,
      "learning_rate": 6.313921802406747e-06,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8423530831933022,
      "num_tokens": 62251150.0,
      "step": 51790
    },
    {
      "entropy": 1.8866778627038001,
      "epoch": 0.16057546775742254,
      "grad_norm": 9.444439888000488,
      "learning_rate": 6.313312309311932e-06,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8364355772733688,
      "num_tokens": 62262711.0,
      "step": 51800
    },
    {
      "entropy": 1.8455047011375427,
      "epoch": 0.16060646688247224,
      "grad_norm": 8.43193244934082,
      "learning_rate": 6.312702992689113e-06,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8457055032253266,
      "num_tokens": 62274824.0,
      "step": 51810
    },
    {
      "entropy": 1.850141017138958,
      "epoch": 0.16063746600752193,
      "grad_norm": 10.177520751953125,
      "learning_rate": 6.312093852453148e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.832466046512127,
      "num_tokens": 62286953.0,
      "step": 51820
    },
    {
      "entropy": 1.8169025957584382,
      "epoch": 0.16066846513257163,
      "grad_norm": 2.502871036529541,
      "learning_rate": 6.31148488851895e-06,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8574206903576851,
      "num_tokens": 62300292.0,
      "step": 51830
    },
    {
      "entropy": 1.8187232732772827,
      "epoch": 0.16069946425762133,
      "grad_norm": 7.720560550689697,
      "learning_rate": 6.3108761008014915e-06,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8429305925965309,
      "num_tokens": 62312236.0,
      "step": 51840
    },
    {
      "entropy": 1.879136349260807,
      "epoch": 0.16073046338267102,
      "grad_norm": 7.850799083709717,
      "learning_rate": 6.310267489215804e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8317036911845207,
      "num_tokens": 62324178.0,
      "step": 51850
    },
    {
      "entropy": 1.7745792135596274,
      "epoch": 0.16076146250772072,
      "grad_norm": 4.761654853820801,
      "learning_rate": 6.309659053676972e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8426973283290863,
      "num_tokens": 62338322.0,
      "step": 51860
    },
    {
      "entropy": 1.8939239963889123,
      "epoch": 0.16079246163277042,
      "grad_norm": 3.76540470123291,
      "learning_rate": 6.309050794100141e-06,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8374624997377396,
      "num_tokens": 62349756.0,
      "step": 51870
    },
    {
      "entropy": 1.9361850887537002,
      "epoch": 0.1608234607578201,
      "grad_norm": 7.414719581604004,
      "learning_rate": 6.308442710400513e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8424745500087738,
      "num_tokens": 62360166.0,
      "step": 51880
    },
    {
      "entropy": 1.816766545176506,
      "epoch": 0.1608544598828698,
      "grad_norm": 9.440898895263672,
      "learning_rate": 6.3078348024933465e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8390632942318916,
      "num_tokens": 62373447.0,
      "step": 51890
    },
    {
      "entropy": 1.9127420842647553,
      "epoch": 0.1608854590079195,
      "grad_norm": 8.341848373413086,
      "learning_rate": 6.307227070293956e-06,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8356220990419387,
      "num_tokens": 62384460.0,
      "step": 51900
    },
    {
      "entropy": 1.8038682281970977,
      "epoch": 0.1609164581329692,
      "grad_norm": 3.9414637088775635,
      "learning_rate": 6.3066195137177146e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.841181357204914,
      "num_tokens": 62397300.0,
      "step": 51910
    },
    {
      "entropy": 1.9235049843788148,
      "epoch": 0.1609474572580189,
      "grad_norm": 4.492739677429199,
      "learning_rate": 6.306012132680054e-06,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8201109856367111,
      "num_tokens": 62409356.0,
      "step": 51920
    },
    {
      "entropy": 1.8290775090456008,
      "epoch": 0.1609784563830686,
      "grad_norm": 3.958709955215454,
      "learning_rate": 6.3054049270964605e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.83324736058712,
      "num_tokens": 62423295.0,
      "step": 51930
    },
    {
      "entropy": 1.933160249888897,
      "epoch": 0.1610094555081183,
      "grad_norm": 7.670217037200928,
      "learning_rate": 6.304797896882477e-06,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8253769144415856,
      "num_tokens": 62434328.0,
      "step": 51940
    },
    {
      "entropy": 1.8668658897280692,
      "epoch": 0.161040454633168,
      "grad_norm": 4.041083812713623,
      "learning_rate": 6.3041910419537055e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8454795464873314,
      "num_tokens": 62446093.0,
      "step": 51950
    },
    {
      "entropy": 1.9969589874148368,
      "epoch": 0.1610714537582177,
      "grad_norm": 10.19937801361084,
      "learning_rate": 6.3035843622258045e-06,
      "loss": 0.5644,
      "mean_token_accuracy": 0.8266312971711158,
      "num_tokens": 62457402.0,
      "step": 51960
    },
    {
      "entropy": 1.911833395063877,
      "epoch": 0.16110245288326738,
      "grad_norm": 9.627182006835938,
      "learning_rate": 6.302977857614485e-06,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8356106966733933,
      "num_tokens": 62470041.0,
      "step": 51970
    },
    {
      "entropy": 1.9067854672670363,
      "epoch": 0.16113345200831705,
      "grad_norm": 3.0671586990356445,
      "learning_rate": 6.302371528035522e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8428362473845482,
      "num_tokens": 62482431.0,
      "step": 51980
    },
    {
      "entropy": 1.9820441797375679,
      "epoch": 0.16116445113336675,
      "grad_norm": 11.670134544372559,
      "learning_rate": 6.301765373404741e-06,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8310918644070625,
      "num_tokens": 62494113.0,
      "step": 51990
    },
    {
      "entropy": 1.9489408016204834,
      "epoch": 0.16119545025841645,
      "grad_norm": 9.345502853393555,
      "learning_rate": 6.301159393638029e-06,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8365255728363991,
      "num_tokens": 62505895.0,
      "step": 52000
    },
    {
      "entropy": 1.9901140004396438,
      "epoch": 0.16122644938346614,
      "grad_norm": 8.158758163452148,
      "learning_rate": 6.300553588651326e-06,
      "loss": 0.5582,
      "mean_token_accuracy": 0.8245610594749451,
      "num_tokens": 62516556.0,
      "step": 52010
    },
    {
      "entropy": 1.8812553852796554,
      "epoch": 0.16125744850851584,
      "grad_norm": 4.510929584503174,
      "learning_rate": 6.2999479583606295e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8425406232476235,
      "num_tokens": 62529781.0,
      "step": 52020
    },
    {
      "entropy": 1.9245860621333122,
      "epoch": 0.16128844763356553,
      "grad_norm": 5.81223726272583,
      "learning_rate": 6.299342502681993e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8385741889476777,
      "num_tokens": 62541343.0,
      "step": 52030
    },
    {
      "entropy": 1.8486530616879464,
      "epoch": 0.16131944675861523,
      "grad_norm": 9.269059181213379,
      "learning_rate": 6.298737221531529e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8453450709581375,
      "num_tokens": 62554053.0,
      "step": 52040
    },
    {
      "entropy": 1.9573129430413245,
      "epoch": 0.16135044588366493,
      "grad_norm": 4.898494720458984,
      "learning_rate": 6.298132114825405e-06,
      "loss": 0.5759,
      "mean_token_accuracy": 0.8290790885686874,
      "num_tokens": 62565444.0,
      "step": 52050
    },
    {
      "entropy": 1.9822146385908126,
      "epoch": 0.16138144500871462,
      "grad_norm": 9.721142768859863,
      "learning_rate": 6.2975271824798425e-06,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8361834734678268,
      "num_tokens": 62576680.0,
      "step": 52060
    },
    {
      "entropy": 1.9362813726067543,
      "epoch": 0.16141244413376432,
      "grad_norm": 9.152141571044922,
      "learning_rate": 6.2969224244111225e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.8337451457977295,
      "num_tokens": 62588501.0,
      "step": 52070
    },
    {
      "entropy": 1.9137926280498505,
      "epoch": 0.16144344325881402,
      "grad_norm": 7.9217047691345215,
      "learning_rate": 6.296317840535582e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8420775726437568,
      "num_tokens": 62601151.0,
      "step": 52080
    },
    {
      "entropy": 2.0264596685767176,
      "epoch": 0.16147444238386371,
      "grad_norm": 9.242731094360352,
      "learning_rate": 6.295713430769611e-06,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8302224919199943,
      "num_tokens": 62612685.0,
      "step": 52090
    },
    {
      "entropy": 1.9726679027080536,
      "epoch": 0.1615054415089134,
      "grad_norm": 3.5792272090911865,
      "learning_rate": 6.29510919502966e-06,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8441024616360664,
      "num_tokens": 62624794.0,
      "step": 52100
    },
    {
      "entropy": 2.029518485069275,
      "epoch": 0.1615364406339631,
      "grad_norm": 9.071422576904297,
      "learning_rate": 6.294505133232234e-06,
      "loss": 0.6116,
      "mean_token_accuracy": 0.815070490539074,
      "num_tokens": 62636376.0,
      "step": 52110
    },
    {
      "entropy": 1.9669608414173125,
      "epoch": 0.1615674397590128,
      "grad_norm": 8.340578079223633,
      "learning_rate": 6.293901245293893e-06,
      "loss": 0.5875,
      "mean_token_accuracy": 0.827280244231224,
      "num_tokens": 62648432.0,
      "step": 52120
    },
    {
      "entropy": 1.9860535800457,
      "epoch": 0.1615984388840625,
      "grad_norm": 8.199106216430664,
      "learning_rate": 6.293297531131253e-06,
      "loss": 0.5733,
      "mean_token_accuracy": 0.83171975761652,
      "num_tokens": 62660966.0,
      "step": 52130
    },
    {
      "entropy": 2.0739176899194716,
      "epoch": 0.1616294380091122,
      "grad_norm": 10.330364227294922,
      "learning_rate": 6.292693990660986e-06,
      "loss": 0.5954,
      "mean_token_accuracy": 0.8177470460534095,
      "num_tokens": 62671756.0,
      "step": 52140
    },
    {
      "entropy": 1.9170552849769593,
      "epoch": 0.1616604371341619,
      "grad_norm": 8.208378791809082,
      "learning_rate": 6.292090623799823e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8416977405548096,
      "num_tokens": 62683638.0,
      "step": 52150
    },
    {
      "entropy": 1.8894746892154217,
      "epoch": 0.1616914362592116,
      "grad_norm": 6.65895414352417,
      "learning_rate": 6.291487430464548e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.845618699491024,
      "num_tokens": 62696586.0,
      "step": 52160
    },
    {
      "entropy": 1.8414909780025481,
      "epoch": 0.1617224353842613,
      "grad_norm": 9.130861282348633,
      "learning_rate": 6.290884410572e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8525190591812134,
      "num_tokens": 62709495.0,
      "step": 52170
    },
    {
      "entropy": 1.9621912389993668,
      "epoch": 0.16175343450931098,
      "grad_norm": 9.641398429870605,
      "learning_rate": 6.290281564039078e-06,
      "loss": 0.551,
      "mean_token_accuracy": 0.8189822494983673,
      "num_tokens": 62721137.0,
      "step": 52180
    },
    {
      "entropy": 1.869258552789688,
      "epoch": 0.16178443363436068,
      "grad_norm": 5.695123672485352,
      "learning_rate": 6.28967889078273e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8320344671607017,
      "num_tokens": 62734333.0,
      "step": 52190
    },
    {
      "entropy": 1.860531361401081,
      "epoch": 0.16181543275941038,
      "grad_norm": 7.216394424438477,
      "learning_rate": 6.289076390719966e-06,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8472234353423118,
      "num_tokens": 62746055.0,
      "step": 52200
    },
    {
      "entropy": 1.9746491819620133,
      "epoch": 0.16184643188446007,
      "grad_norm": 9.281171798706055,
      "learning_rate": 6.2884740637678486e-06,
      "loss": 0.5562,
      "mean_token_accuracy": 0.8372055932879447,
      "num_tokens": 62757053.0,
      "step": 52210
    },
    {
      "entropy": 1.9339981719851493,
      "epoch": 0.16187743100950977,
      "grad_norm": 10.073769569396973,
      "learning_rate": 6.2878719098434975e-06,
      "loss": 0.577,
      "mean_token_accuracy": 0.8267990037798881,
      "num_tokens": 62768818.0,
      "step": 52220
    },
    {
      "entropy": 1.75873264670372,
      "epoch": 0.16190843013455944,
      "grad_norm": 3.918210744857788,
      "learning_rate": 6.287269928864085e-06,
      "loss": 0.4269,
      "mean_token_accuracy": 0.8522544384002686,
      "num_tokens": 62782705.0,
      "step": 52230
    },
    {
      "entropy": 1.9759118229150772,
      "epoch": 0.16193942925960914,
      "grad_norm": 8.20750904083252,
      "learning_rate": 6.286668120746842e-06,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8221010401844978,
      "num_tokens": 62794524.0,
      "step": 52240
    },
    {
      "entropy": 1.8256248250603675,
      "epoch": 0.16197042838465883,
      "grad_norm": 3.0523853302001953,
      "learning_rate": 6.286066485409056e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.849830961227417,
      "num_tokens": 62807370.0,
      "step": 52250
    },
    {
      "entropy": 1.864981435239315,
      "epoch": 0.16200142750970853,
      "grad_norm": 8.436219215393066,
      "learning_rate": 6.285465022768064e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.8273715361952781,
      "num_tokens": 62820130.0,
      "step": 52260
    },
    {
      "entropy": 1.9281801611185074,
      "epoch": 0.16203242663475823,
      "grad_norm": 8.644983291625977,
      "learning_rate": 6.284863732741263e-06,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8462273836135864,
      "num_tokens": 62832096.0,
      "step": 52270
    },
    {
      "entropy": 1.9527894288301468,
      "epoch": 0.16206342575980792,
      "grad_norm": 9.219314575195312,
      "learning_rate": 6.284262615246107e-06,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8327635273337364,
      "num_tokens": 62843243.0,
      "step": 52280
    },
    {
      "entropy": 1.823768149316311,
      "epoch": 0.16209442488485762,
      "grad_norm": 3.6631810665130615,
      "learning_rate": 6.283661670200099e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8509679839015007,
      "num_tokens": 62856797.0,
      "step": 52290
    },
    {
      "entropy": 1.9193030267953872,
      "epoch": 0.16212542400990732,
      "grad_norm": 19.060598373413086,
      "learning_rate": 6.283060897520804e-06,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8333917066454888,
      "num_tokens": 62869635.0,
      "step": 52300
    },
    {
      "entropy": 1.9955947354435921,
      "epoch": 0.162156423134957,
      "grad_norm": 4.73981237411499,
      "learning_rate": 6.282460297125835e-06,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8443202391266823,
      "num_tokens": 62881179.0,
      "step": 52310
    },
    {
      "entropy": 1.9485764712095262,
      "epoch": 0.1621874222600067,
      "grad_norm": 8.990504264831543,
      "learning_rate": 6.281859868932869e-06,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8265508517622948,
      "num_tokens": 62893144.0,
      "step": 52320
    },
    {
      "entropy": 1.9622452184557915,
      "epoch": 0.1622184213850564,
      "grad_norm": 8.875653266906738,
      "learning_rate": 6.281259612859629e-06,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8386439487338067,
      "num_tokens": 62904857.0,
      "step": 52330
    },
    {
      "entropy": 1.889234210550785,
      "epoch": 0.1622494205101061,
      "grad_norm": 10.868480682373047,
      "learning_rate": 6.2806595288239e-06,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8372628569602967,
      "num_tokens": 62917145.0,
      "step": 52340
    },
    {
      "entropy": 1.8264701277017594,
      "epoch": 0.1622804196351558,
      "grad_norm": 9.613685607910156,
      "learning_rate": 6.2800596167435165e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.850235678255558,
      "num_tokens": 62930488.0,
      "step": 52350
    },
    {
      "entropy": 1.854723860323429,
      "epoch": 0.1623114187602055,
      "grad_norm": 9.742049217224121,
      "learning_rate": 6.279459876536374e-06,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8460089027881622,
      "num_tokens": 62942658.0,
      "step": 52360
    },
    {
      "entropy": 1.9286055207252502,
      "epoch": 0.1623424178852552,
      "grad_norm": 4.1667304039001465,
      "learning_rate": 6.278860308120416e-06,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8262563273310661,
      "num_tokens": 62953922.0,
      "step": 52370
    },
    {
      "entropy": 1.7510716319084167,
      "epoch": 0.1623734170103049,
      "grad_norm": 8.469949722290039,
      "learning_rate": 6.278260911413646e-06,
      "loss": 0.415,
      "mean_token_accuracy": 0.8467779159545898,
      "num_tokens": 62967483.0,
      "step": 52380
    },
    {
      "entropy": 1.9073357090353966,
      "epoch": 0.16240441613535458,
      "grad_norm": 8.063495635986328,
      "learning_rate": 6.27766168633412e-06,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8321097016334533,
      "num_tokens": 62979720.0,
      "step": 52390
    },
    {
      "entropy": 1.9408669352531434,
      "epoch": 0.16243541526040428,
      "grad_norm": 10.30274772644043,
      "learning_rate": 6.277062632799949e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.8371255546808243,
      "num_tokens": 62991264.0,
      "step": 52400
    },
    {
      "entropy": 1.9547599002718925,
      "epoch": 0.16246641438545398,
      "grad_norm": 8.85878849029541,
      "learning_rate": 6.276463750729301e-06,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8315024048089981,
      "num_tokens": 63003018.0,
      "step": 52410
    },
    {
      "entropy": 1.9163083508610725,
      "epoch": 0.16249741351050367,
      "grad_norm": 9.163949966430664,
      "learning_rate": 6.2758650400403964e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8458400592207909,
      "num_tokens": 63014741.0,
      "step": 52420
    },
    {
      "entropy": 1.9846520096063613,
      "epoch": 0.16252841263555337,
      "grad_norm": 7.502941131591797,
      "learning_rate": 6.275266500651508e-06,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8262620836496353,
      "num_tokens": 63025808.0,
      "step": 52430
    },
    {
      "entropy": 1.8606478095054626,
      "epoch": 0.16255941176060307,
      "grad_norm": 4.25178861618042,
      "learning_rate": 6.274668132480967e-06,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8410593420267105,
      "num_tokens": 63038607.0,
      "step": 52440
    },
    {
      "entropy": 1.8719539806246757,
      "epoch": 0.16259041088565276,
      "grad_norm": 4.510512828826904,
      "learning_rate": 6.274069935447157e-06,
      "loss": 0.491,
      "mean_token_accuracy": 0.8395673424005509,
      "num_tokens": 63050950.0,
      "step": 52450
    },
    {
      "entropy": 1.8893551722168922,
      "epoch": 0.16262141001070246,
      "grad_norm": 3.8029367923736572,
      "learning_rate": 6.273471909468518e-06,
      "loss": 0.487,
      "mean_token_accuracy": 0.8418749555945396,
      "num_tokens": 63063475.0,
      "step": 52460
    },
    {
      "entropy": 1.8528440594673157,
      "epoch": 0.16265240913575213,
      "grad_norm": 8.126982688903809,
      "learning_rate": 6.272874054463543e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8420227542519569,
      "num_tokens": 63076738.0,
      "step": 52470
    },
    {
      "entropy": 1.9618181601166724,
      "epoch": 0.16268340826080183,
      "grad_norm": 11.542217254638672,
      "learning_rate": 6.272276370350776e-06,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8369494050741195,
      "num_tokens": 63087700.0,
      "step": 52480
    },
    {
      "entropy": 1.9238998390734197,
      "epoch": 0.16271440738585152,
      "grad_norm": 7.640765190124512,
      "learning_rate": 6.271678857048824e-06,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8411808148026466,
      "num_tokens": 63099944.0,
      "step": 52490
    },
    {
      "entropy": 1.888495209813118,
      "epoch": 0.16274540651090122,
      "grad_norm": 7.534304141998291,
      "learning_rate": 6.271081514476341e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8466553881764411,
      "num_tokens": 63112697.0,
      "step": 52500
    },
    {
      "entropy": 1.9347805023193358,
      "epoch": 0.16277640563595092,
      "grad_norm": 7.860371112823486,
      "learning_rate": 6.270484342552038e-06,
      "loss": 0.535,
      "mean_token_accuracy": 0.8298186644911766,
      "num_tokens": 63124515.0,
      "step": 52510
    },
    {
      "entropy": 1.8718351736664771,
      "epoch": 0.1628074047610006,
      "grad_norm": 11.632791519165039,
      "learning_rate": 6.269887341194678e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8416498139500618,
      "num_tokens": 63137607.0,
      "step": 52520
    },
    {
      "entropy": 1.8613640531897544,
      "epoch": 0.1628384038860503,
      "grad_norm": 5.130488872528076,
      "learning_rate": 6.269290510323079e-06,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8313415363430977,
      "num_tokens": 63150710.0,
      "step": 52530
    },
    {
      "entropy": 1.8332793086767196,
      "epoch": 0.1628694030111,
      "grad_norm": 8.796858787536621,
      "learning_rate": 6.2686938498561155e-06,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8430872440338135,
      "num_tokens": 63163425.0,
      "step": 52540
    },
    {
      "entropy": 1.8761805295944214,
      "epoch": 0.1629004021361497,
      "grad_norm": 4.210749626159668,
      "learning_rate": 6.268097359712715e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8424938842654228,
      "num_tokens": 63175765.0,
      "step": 52550
    },
    {
      "entropy": 1.9241129815578462,
      "epoch": 0.1629314012611994,
      "grad_norm": 7.9560465812683105,
      "learning_rate": 6.267501039811856e-06,
      "loss": 0.5341,
      "mean_token_accuracy": 0.836355759203434,
      "num_tokens": 63187161.0,
      "step": 52560
    },
    {
      "entropy": 1.9873510360717774,
      "epoch": 0.1629624003862491,
      "grad_norm": 8.924347877502441,
      "learning_rate": 6.2669048900725745e-06,
      "loss": 0.6066,
      "mean_token_accuracy": 0.8214996859431267,
      "num_tokens": 63198468.0,
      "step": 52570
    },
    {
      "entropy": 1.9314937353134156,
      "epoch": 0.1629933995112988,
      "grad_norm": 5.659384727478027,
      "learning_rate": 6.266308910413959e-06,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8333692610263824,
      "num_tokens": 63210134.0,
      "step": 52580
    },
    {
      "entropy": 2.028110182285309,
      "epoch": 0.1630243986363485,
      "grad_norm": 8.73976993560791,
      "learning_rate": 6.2657131007551516e-06,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8240557476878166,
      "num_tokens": 63221318.0,
      "step": 52590
    },
    {
      "entropy": 1.8795430675148963,
      "epoch": 0.16305539776139819,
      "grad_norm": 7.525078296661377,
      "learning_rate": 6.265117461015348e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8401163816452026,
      "num_tokens": 63234281.0,
      "step": 52600
    },
    {
      "entropy": 1.9891944900155067,
      "epoch": 0.16308639688644788,
      "grad_norm": 8.702574729919434,
      "learning_rate": 6.2645219911138e-06,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8414581835269928,
      "num_tokens": 63245528.0,
      "step": 52610
    },
    {
      "entropy": 1.9499772146344185,
      "epoch": 0.16311739601149758,
      "grad_norm": 8.397680282592773,
      "learning_rate": 6.263926690969809e-06,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8338237330317497,
      "num_tokens": 63257825.0,
      "step": 52620
    },
    {
      "entropy": 1.9680059745907783,
      "epoch": 0.16314839513654728,
      "grad_norm": 8.269797325134277,
      "learning_rate": 6.263331560502734e-06,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8288321003317833,
      "num_tokens": 63269246.0,
      "step": 52630
    },
    {
      "entropy": 1.9180641785264014,
      "epoch": 0.16317939426159697,
      "grad_norm": 7.361063003540039,
      "learning_rate": 6.262736599631985e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8423007354140282,
      "num_tokens": 63281078.0,
      "step": 52640
    },
    {
      "entropy": 1.9701405078172685,
      "epoch": 0.16321039338664667,
      "grad_norm": 9.259462356567383,
      "learning_rate": 6.262141808277028e-06,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8284604266285897,
      "num_tokens": 63292492.0,
      "step": 52650
    },
    {
      "entropy": 1.8732146829366685,
      "epoch": 0.16324139251169636,
      "grad_norm": 4.93553352355957,
      "learning_rate": 6.261547186357378e-06,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8330485403537751,
      "num_tokens": 63305365.0,
      "step": 52660
    },
    {
      "entropy": 1.8874418213963509,
      "epoch": 0.16327239163674606,
      "grad_norm": 12.51389217376709,
      "learning_rate": 6.260952733792611e-06,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8417252108454705,
      "num_tokens": 63317381.0,
      "step": 52670
    },
    {
      "entropy": 1.9065624982118607,
      "epoch": 0.16330339076179576,
      "grad_norm": 4.349983215332031,
      "learning_rate": 6.26035845050235e-06,
      "loss": 0.4946,
      "mean_token_accuracy": 0.83489740639925,
      "num_tokens": 63329554.0,
      "step": 52680
    },
    {
      "entropy": 1.9560609757900238,
      "epoch": 0.16333438988684545,
      "grad_norm": 8.065526008605957,
      "learning_rate": 6.259764336406272e-06,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8307097449898719,
      "num_tokens": 63341290.0,
      "step": 52690
    },
    {
      "entropy": 1.9741875648498535,
      "epoch": 0.16336538901189515,
      "grad_norm": 4.957176685333252,
      "learning_rate": 6.259170391424109e-06,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8235329478979111,
      "num_tokens": 63352756.0,
      "step": 52700
    },
    {
      "entropy": 1.897114197909832,
      "epoch": 0.16339638813694485,
      "grad_norm": 7.078670024871826,
      "learning_rate": 6.25857661547565e-06,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8442688778042793,
      "num_tokens": 63364598.0,
      "step": 52710
    },
    {
      "entropy": 1.9124154165387153,
      "epoch": 0.16342738726199452,
      "grad_norm": 8.284811973571777,
      "learning_rate": 6.257983008480728e-06,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8364136472344399,
      "num_tokens": 63376703.0,
      "step": 52720
    },
    {
      "entropy": 1.9177636936306954,
      "epoch": 0.1634583863870442,
      "grad_norm": 6.819045066833496,
      "learning_rate": 6.257389570359238e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.846918910741806,
      "num_tokens": 63389432.0,
      "step": 52730
    },
    {
      "entropy": 1.9418374314904212,
      "epoch": 0.1634893855120939,
      "grad_norm": 9.605525970458984,
      "learning_rate": 6.256796301031124e-06,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8311137303709983,
      "num_tokens": 63401131.0,
      "step": 52740
    },
    {
      "entropy": 1.8737993687391281,
      "epoch": 0.1635203846371436,
      "grad_norm": 8.54281234741211,
      "learning_rate": 6.256203200416383e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8313992112874985,
      "num_tokens": 63413232.0,
      "step": 52750
    },
    {
      "entropy": 1.8898098900914193,
      "epoch": 0.1635513837621933,
      "grad_norm": 8.377137184143066,
      "learning_rate": 6.255610268435066e-06,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8360078081488609,
      "num_tokens": 63425498.0,
      "step": 52760
    },
    {
      "entropy": 1.9383550241589547,
      "epoch": 0.163582382887243,
      "grad_norm": 8.232552528381348,
      "learning_rate": 6.255017505007278e-06,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8423770934343338,
      "num_tokens": 63436970.0,
      "step": 52770
    },
    {
      "entropy": 1.9232452899217605,
      "epoch": 0.1636133820122927,
      "grad_norm": 7.204452991485596,
      "learning_rate": 6.254424910053175e-06,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8377453580498695,
      "num_tokens": 63448237.0,
      "step": 52780
    },
    {
      "entropy": 1.9503621995449065,
      "epoch": 0.1636443811373424,
      "grad_norm": 7.743916034698486,
      "learning_rate": 6.253832483492968e-06,
      "loss": 0.604,
      "mean_token_accuracy": 0.818188302218914,
      "num_tokens": 63459542.0,
      "step": 52790
    },
    {
      "entropy": 1.9227148860692977,
      "epoch": 0.1636753802623921,
      "grad_norm": 8.97805404663086,
      "learning_rate": 6.253240225246917e-06,
      "loss": 0.5535,
      "mean_token_accuracy": 0.835084454715252,
      "num_tokens": 63470819.0,
      "step": 52800
    },
    {
      "entropy": 1.8931695908308028,
      "epoch": 0.16370637938744179,
      "grad_norm": 8.5215482711792,
      "learning_rate": 6.25264813523534e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.841637410223484,
      "num_tokens": 63483021.0,
      "step": 52810
    },
    {
      "entropy": 1.9346808150410653,
      "epoch": 0.16373737851249148,
      "grad_norm": 3.57350754737854,
      "learning_rate": 6.252056213378607e-06,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8383998513221741,
      "num_tokens": 63495186.0,
      "step": 52820
    },
    {
      "entropy": 1.8804344907402992,
      "epoch": 0.16376837763754118,
      "grad_norm": 8.589835166931152,
      "learning_rate": 6.251464459597134e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8427335307002067,
      "num_tokens": 63507869.0,
      "step": 52830
    },
    {
      "entropy": 1.8359813764691353,
      "epoch": 0.16379937676259088,
      "grad_norm": 9.559521675109863,
      "learning_rate": 6.2508728738114e-06,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8516914382576942,
      "num_tokens": 63519894.0,
      "step": 52840
    },
    {
      "entropy": 1.9041066259145736,
      "epoch": 0.16383037588764057,
      "grad_norm": 4.24696683883667,
      "learning_rate": 6.250281455941929e-06,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8393527343869209,
      "num_tokens": 63532566.0,
      "step": 52850
    },
    {
      "entropy": 1.857696245610714,
      "epoch": 0.16386137501269027,
      "grad_norm": 13.18340015411377,
      "learning_rate": 6.249690205909301e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8314845860004425,
      "num_tokens": 63545311.0,
      "step": 52860
    },
    {
      "entropy": 1.9533265694975852,
      "epoch": 0.16389237413773997,
      "grad_norm": 8.314949035644531,
      "learning_rate": 6.249099123634147e-06,
      "loss": 0.5793,
      "mean_token_accuracy": 0.8232409819960594,
      "num_tokens": 63556591.0,
      "step": 52870
    },
    {
      "entropy": 1.9042376592755317,
      "epoch": 0.16392337326278966,
      "grad_norm": 7.390749454498291,
      "learning_rate": 6.248508209037151e-06,
      "loss": 0.5547,
      "mean_token_accuracy": 0.827096089720726,
      "num_tokens": 63568839.0,
      "step": 52880
    },
    {
      "entropy": 1.8934383913874626,
      "epoch": 0.16395437238783936,
      "grad_norm": 8.171467781066895,
      "learning_rate": 6.2479174620390516e-06,
      "loss": 0.5367,
      "mean_token_accuracy": 0.8317389547824859,
      "num_tokens": 63580636.0,
      "step": 52890
    },
    {
      "entropy": 1.8028965070843697,
      "epoch": 0.16398537151288906,
      "grad_norm": 9.49335765838623,
      "learning_rate": 6.247326882560637e-06,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8371429294347763,
      "num_tokens": 63593629.0,
      "step": 52900
    },
    {
      "entropy": 1.848221817612648,
      "epoch": 0.16401637063793875,
      "grad_norm": 8.622106552124023,
      "learning_rate": 6.246736470522748e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8468390449881553,
      "num_tokens": 63606139.0,
      "step": 52910
    },
    {
      "entropy": 1.8176067188382148,
      "epoch": 0.16404736976298845,
      "grad_norm": 9.547903060913086,
      "learning_rate": 6.2461462258462804e-06,
      "loss": 0.4316,
      "mean_token_accuracy": 0.854963929951191,
      "num_tokens": 63619469.0,
      "step": 52920
    },
    {
      "entropy": 1.9598242774605752,
      "epoch": 0.16407836888803815,
      "grad_norm": 8.696121215820312,
      "learning_rate": 6.245556148452177e-06,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8299388602375984,
      "num_tokens": 63630770.0,
      "step": 52930
    },
    {
      "entropy": 1.9861217468976975,
      "epoch": 0.16410936801308784,
      "grad_norm": 7.65268611907959,
      "learning_rate": 6.244966238261442e-06,
      "loss": 0.5992,
      "mean_token_accuracy": 0.8203268766403198,
      "num_tokens": 63641633.0,
      "step": 52940
    },
    {
      "entropy": 1.933871328830719,
      "epoch": 0.16414036713813754,
      "grad_norm": 4.0689263343811035,
      "learning_rate": 6.2443764951951215e-06,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8379737511277199,
      "num_tokens": 63653263.0,
      "step": 52950
    },
    {
      "entropy": 1.924587444961071,
      "epoch": 0.16417136626318724,
      "grad_norm": 9.592199325561523,
      "learning_rate": 6.24378691917432e-06,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8331802487373352,
      "num_tokens": 63664170.0,
      "step": 52960
    },
    {
      "entropy": 1.922967004776001,
      "epoch": 0.1642023653882369,
      "grad_norm": 3.9042298793792725,
      "learning_rate": 6.2431975101201926e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8360741794109344,
      "num_tokens": 63675465.0,
      "step": 52970
    },
    {
      "entropy": 1.8445443019270897,
      "epoch": 0.1642333645132866,
      "grad_norm": 7.786768913269043,
      "learning_rate": 6.242608267953947e-06,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8440708309412003,
      "num_tokens": 63688438.0,
      "step": 52980
    },
    {
      "entropy": 1.9318128436803819,
      "epoch": 0.1642643636383363,
      "grad_norm": 7.483189582824707,
      "learning_rate": 6.242019192596842e-06,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8481845185160637,
      "num_tokens": 63700364.0,
      "step": 52990
    },
    {
      "entropy": 1.9713671594858169,
      "epoch": 0.164295362763386,
      "grad_norm": 7.716215133666992,
      "learning_rate": 6.241430283970189e-06,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8340044632554054,
      "num_tokens": 63711655.0,
      "step": 53000
    },
    {
      "entropy": 1.916495531797409,
      "epoch": 0.1643263618884357,
      "grad_norm": 7.8675971031188965,
      "learning_rate": 6.24084154199535e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8242413744330406,
      "num_tokens": 63723415.0,
      "step": 53010
    },
    {
      "entropy": 1.8746812611818313,
      "epoch": 0.1643573610134854,
      "grad_norm": 4.13045072555542,
      "learning_rate": 6.240252966593741e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8390706807374955,
      "num_tokens": 63735381.0,
      "step": 53020
    },
    {
      "entropy": 1.9430684581398965,
      "epoch": 0.16438836013853508,
      "grad_norm": 8.874407768249512,
      "learning_rate": 6.23966455768683e-06,
      "loss": 0.5869,
      "mean_token_accuracy": 0.8245584949851036,
      "num_tokens": 63746701.0,
      "step": 53030
    },
    {
      "entropy": 1.9074441373348237,
      "epoch": 0.16441935926358478,
      "grad_norm": 3.734499454498291,
      "learning_rate": 6.239076315196135e-06,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8352471068501472,
      "num_tokens": 63758681.0,
      "step": 53040
    },
    {
      "entropy": 1.9198872432112695,
      "epoch": 0.16445035838863448,
      "grad_norm": 7.779430866241455,
      "learning_rate": 6.2384882390432265e-06,
      "loss": 0.5311,
      "mean_token_accuracy": 0.8286612689495086,
      "num_tokens": 63770142.0,
      "step": 53050
    },
    {
      "entropy": 1.9856867283582686,
      "epoch": 0.16448135751368417,
      "grad_norm": 8.502123832702637,
      "learning_rate": 6.2379003291497265e-06,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8293860018253326,
      "num_tokens": 63781495.0,
      "step": 53060
    },
    {
      "entropy": 1.8787162870168685,
      "epoch": 0.16451235663873387,
      "grad_norm": 3.1227447986602783,
      "learning_rate": 6.237312585437309e-06,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8432197883725167,
      "num_tokens": 63793956.0,
      "step": 53070
    },
    {
      "entropy": 1.8861326739192008,
      "epoch": 0.16454335576378357,
      "grad_norm": 9.353623390197754,
      "learning_rate": 6.236725007827702e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8440233051776886,
      "num_tokens": 63806818.0,
      "step": 53080
    },
    {
      "entropy": 1.7985064759850502,
      "epoch": 0.16457435488883326,
      "grad_norm": 10.218011856079102,
      "learning_rate": 6.23613759624268e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8478423848748207,
      "num_tokens": 63820679.0,
      "step": 53090
    },
    {
      "entropy": 1.8236029118299484,
      "epoch": 0.16460535401388296,
      "grad_norm": 4.523934841156006,
      "learning_rate": 6.235550350604071e-06,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8463995724916458,
      "num_tokens": 63833871.0,
      "step": 53100
    },
    {
      "entropy": 1.8776141807436943,
      "epoch": 0.16463635313893266,
      "grad_norm": 4.4371161460876465,
      "learning_rate": 6.234963270833758e-06,
      "loss": 0.474,
      "mean_token_accuracy": 0.8449958503246308,
      "num_tokens": 63845568.0,
      "step": 53110
    },
    {
      "entropy": 1.8257340744137764,
      "epoch": 0.16466735226398235,
      "grad_norm": 9.007696151733398,
      "learning_rate": 6.234376356853673e-06,
      "loss": 0.431,
      "mean_token_accuracy": 0.8533379226922989,
      "num_tokens": 63858077.0,
      "step": 53120
    },
    {
      "entropy": 1.8883861318230628,
      "epoch": 0.16469835138903205,
      "grad_norm": 8.40960693359375,
      "learning_rate": 6.233789608585796e-06,
      "loss": 0.5674,
      "mean_token_accuracy": 0.8298670023679733,
      "num_tokens": 63870245.0,
      "step": 53130
    },
    {
      "entropy": 1.993936476111412,
      "epoch": 0.16472935051408175,
      "grad_norm": 9.627812385559082,
      "learning_rate": 6.233203025952166e-06,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8264122769236565,
      "num_tokens": 63881175.0,
      "step": 53140
    },
    {
      "entropy": 1.8338624104857444,
      "epoch": 0.16476034963913144,
      "grad_norm": 10.100017547607422,
      "learning_rate": 6.232616608874865e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8356154605746269,
      "num_tokens": 63894063.0,
      "step": 53150
    },
    {
      "entropy": 1.8793131604790687,
      "epoch": 0.16479134876418114,
      "grad_norm": 8.098730087280273,
      "learning_rate": 6.232030357276034e-06,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8461975052952766,
      "num_tokens": 63905919.0,
      "step": 53160
    },
    {
      "entropy": 1.9642018005251884,
      "epoch": 0.16482234788923084,
      "grad_norm": 7.914994716644287,
      "learning_rate": 6.231444271077859e-06,
      "loss": 0.555,
      "mean_token_accuracy": 0.829823549091816,
      "num_tokens": 63916952.0,
      "step": 53170
    },
    {
      "entropy": 1.8831199273467063,
      "epoch": 0.16485334701428053,
      "grad_norm": 10.928045272827148,
      "learning_rate": 6.23085835020258e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8363705515861511,
      "num_tokens": 63928969.0,
      "step": 53180
    },
    {
      "entropy": 1.9851090759038925,
      "epoch": 0.16488434613933023,
      "grad_norm": 9.08737564086914,
      "learning_rate": 6.230272594572488e-06,
      "loss": 0.5901,
      "mean_token_accuracy": 0.8261380925774574,
      "num_tokens": 63939337.0,
      "step": 53190
    },
    {
      "entropy": 1.908937330543995,
      "epoch": 0.16491534526437993,
      "grad_norm": 11.058307647705078,
      "learning_rate": 6.229687004109927e-06,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8325806826353073,
      "num_tokens": 63951300.0,
      "step": 53200
    },
    {
      "entropy": 1.9652216613292695,
      "epoch": 0.16494634438942962,
      "grad_norm": 9.384000778198242,
      "learning_rate": 6.229101578737288e-06,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8316969543695449,
      "num_tokens": 63961965.0,
      "step": 53210
    },
    {
      "entropy": 1.9054602891206742,
      "epoch": 0.1649773435144793,
      "grad_norm": 4.693671226501465,
      "learning_rate": 6.228516318377016e-06,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8300312593579292,
      "num_tokens": 63974012.0,
      "step": 53220
    },
    {
      "entropy": 1.87213394343853,
      "epoch": 0.165008342639529,
      "grad_norm": 8.082072257995605,
      "learning_rate": 6.227931222951605e-06,
      "loss": 0.522,
      "mean_token_accuracy": 0.836493344604969,
      "num_tokens": 63985663.0,
      "step": 53230
    },
    {
      "entropy": 1.9317936196923255,
      "epoch": 0.16503934176457868,
      "grad_norm": 8.324505805969238,
      "learning_rate": 6.227346292383604e-06,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8266930311918259,
      "num_tokens": 63996929.0,
      "step": 53240
    },
    {
      "entropy": 1.8709998071193694,
      "epoch": 0.16507034088962838,
      "grad_norm": 8.387805938720703,
      "learning_rate": 6.226761526595607e-06,
      "loss": 0.557,
      "mean_token_accuracy": 0.8292562425136566,
      "num_tokens": 64008847.0,
      "step": 53250
    },
    {
      "entropy": 1.8826777443289757,
      "epoch": 0.16510134001467808,
      "grad_norm": 7.2748541831970215,
      "learning_rate": 6.2261769255102635e-06,
      "loss": 0.513,
      "mean_token_accuracy": 0.8443334132432938,
      "num_tokens": 64020608.0,
      "step": 53260
    },
    {
      "entropy": 1.8901048377156258,
      "epoch": 0.16513233913972777,
      "grad_norm": 4.1918625831604,
      "learning_rate": 6.225592489050273e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8356076464056968,
      "num_tokens": 64032582.0,
      "step": 53270
    },
    {
      "entropy": 1.8856966137886046,
      "epoch": 0.16516333826477747,
      "grad_norm": 9.814857482910156,
      "learning_rate": 6.225008217138383e-06,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8268331870436668,
      "num_tokens": 64044568.0,
      "step": 53280
    },
    {
      "entropy": 1.9235829666256905,
      "epoch": 0.16519433738982717,
      "grad_norm": 8.297536849975586,
      "learning_rate": 6.224424109697395e-06,
      "loss": 0.509,
      "mean_token_accuracy": 0.8338031157851219,
      "num_tokens": 64055682.0,
      "step": 53290
    },
    {
      "entropy": 1.8362517848610878,
      "epoch": 0.16522533651487686,
      "grad_norm": 4.987396717071533,
      "learning_rate": 6.2238401666501594e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.842101874947548,
      "num_tokens": 64068519.0,
      "step": 53300
    },
    {
      "entropy": 1.8466239094734191,
      "epoch": 0.16525633563992656,
      "grad_norm": 7.831912517547607,
      "learning_rate": 6.2232563879195784e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8441867157816887,
      "num_tokens": 64081113.0,
      "step": 53310
    },
    {
      "entropy": 1.901265199482441,
      "epoch": 0.16528733476497626,
      "grad_norm": 9.943422317504883,
      "learning_rate": 6.222672773428604e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8303415447473526,
      "num_tokens": 64093471.0,
      "step": 53320
    },
    {
      "entropy": 1.8734540060162543,
      "epoch": 0.16531833389002595,
      "grad_norm": 9.80716609954834,
      "learning_rate": 6.222089323100241e-06,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8340218141674995,
      "num_tokens": 64106551.0,
      "step": 53330
    },
    {
      "entropy": 1.8805134430527688,
      "epoch": 0.16534933301507565,
      "grad_norm": 8.605008125305176,
      "learning_rate": 6.221506036857539e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8431431293487549,
      "num_tokens": 64119113.0,
      "step": 53340
    },
    {
      "entropy": 1.8234007403254509,
      "epoch": 0.16538033214012535,
      "grad_norm": 4.009090423583984,
      "learning_rate": 6.220922914623604e-06,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8466459035873413,
      "num_tokens": 64132616.0,
      "step": 53350
    },
    {
      "entropy": 1.8845936864614488,
      "epoch": 0.16541133126517504,
      "grad_norm": 10.148909568786621,
      "learning_rate": 6.2203399563215905e-06,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8377220943570137,
      "num_tokens": 64145357.0,
      "step": 53360
    },
    {
      "entropy": 1.959778368473053,
      "epoch": 0.16544233039022474,
      "grad_norm": 8.213615417480469,
      "learning_rate": 6.219757161874702e-06,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8356760248541832,
      "num_tokens": 64156979.0,
      "step": 53370
    },
    {
      "entropy": 1.9542056560516357,
      "epoch": 0.16547332951527444,
      "grad_norm": 7.512001991271973,
      "learning_rate": 6.219174531206195e-06,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8426282212138176,
      "num_tokens": 64167936.0,
      "step": 53380
    },
    {
      "entropy": 1.8867290809750557,
      "epoch": 0.16550432864032413,
      "grad_norm": 11.55746078491211,
      "learning_rate": 6.2185920642393724e-06,
      "loss": 0.538,
      "mean_token_accuracy": 0.8335390016436577,
      "num_tokens": 64178732.0,
      "step": 53390
    },
    {
      "entropy": 1.9342323437333107,
      "epoch": 0.16553532776537383,
      "grad_norm": 8.260815620422363,
      "learning_rate": 6.218009760897592e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8423149228096009,
      "num_tokens": 64189523.0,
      "step": 53400
    },
    {
      "entropy": 1.8842595234513282,
      "epoch": 0.16556632689042353,
      "grad_norm": 8.171215057373047,
      "learning_rate": 6.21742762110426e-06,
      "loss": 0.4739,
      "mean_token_accuracy": 0.852245531976223,
      "num_tokens": 64200851.0,
      "step": 53410
    },
    {
      "entropy": 1.9514251738786696,
      "epoch": 0.16559732601547322,
      "grad_norm": 11.585762977600098,
      "learning_rate": 6.216845644782831e-06,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8217004343867302,
      "num_tokens": 64212043.0,
      "step": 53420
    },
    {
      "entropy": 1.9424443751573564,
      "epoch": 0.16562832514052292,
      "grad_norm": 7.723189353942871,
      "learning_rate": 6.216263831856811e-06,
      "loss": 0.5982,
      "mean_token_accuracy": 0.830877748131752,
      "num_tokens": 64223504.0,
      "step": 53430
    },
    {
      "entropy": 1.9373986154794693,
      "epoch": 0.16565932426557262,
      "grad_norm": 8.541019439697266,
      "learning_rate": 6.215682182249758e-06,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8272872015833854,
      "num_tokens": 64235303.0,
      "step": 53440
    },
    {
      "entropy": 1.926292023062706,
      "epoch": 0.1656903233906223,
      "grad_norm": 9.712891578674316,
      "learning_rate": 6.215100695885277e-06,
      "loss": 0.5625,
      "mean_token_accuracy": 0.8317703247070313,
      "num_tokens": 64246713.0,
      "step": 53450
    },
    {
      "entropy": 1.8299464777112007,
      "epoch": 0.16572132251567198,
      "grad_norm": 8.690601348876953,
      "learning_rate": 6.214519372687023e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8461026042699814,
      "num_tokens": 64259907.0,
      "step": 53460
    },
    {
      "entropy": 1.9531344637274741,
      "epoch": 0.16575232164072168,
      "grad_norm": 8.977042198181152,
      "learning_rate": 6.2139382125787065e-06,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8346742272377015,
      "num_tokens": 64271538.0,
      "step": 53470
    },
    {
      "entropy": 1.870873971283436,
      "epoch": 0.16578332076577137,
      "grad_norm": 10.686627388000488,
      "learning_rate": 6.213357215484079e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8391006410121917,
      "num_tokens": 64283477.0,
      "step": 53480
    },
    {
      "entropy": 1.884416152536869,
      "epoch": 0.16581431989082107,
      "grad_norm": 9.072136878967285,
      "learning_rate": 6.2127763813269494e-06,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8330440178513527,
      "num_tokens": 64295950.0,
      "step": 53490
    },
    {
      "entropy": 1.8264600470662118,
      "epoch": 0.16584531901587077,
      "grad_norm": 3.7949202060699463,
      "learning_rate": 6.212195710031174e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8554823577404023,
      "num_tokens": 64308519.0,
      "step": 53500
    },
    {
      "entropy": 1.8256630197167396,
      "epoch": 0.16587631814092046,
      "grad_norm": 8.502163887023926,
      "learning_rate": 6.211615201520656e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8437755182385445,
      "num_tokens": 64321187.0,
      "step": 53510
    },
    {
      "entropy": 1.7614564910531043,
      "epoch": 0.16590731726597016,
      "grad_norm": 9.997268676757812,
      "learning_rate": 6.211034855719356e-06,
      "loss": 0.466,
      "mean_token_accuracy": 0.8389209792017936,
      "num_tokens": 64334994.0,
      "step": 53520
    },
    {
      "entropy": 1.934888207912445,
      "epoch": 0.16593831639101986,
      "grad_norm": 7.5674519538879395,
      "learning_rate": 6.210454672551274e-06,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8264615863561631,
      "num_tokens": 64345597.0,
      "step": 53530
    },
    {
      "entropy": 1.8772856667637825,
      "epoch": 0.16596931551606955,
      "grad_norm": 10.031050682067871,
      "learning_rate": 6.209874651940466e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8403290301561356,
      "num_tokens": 64357564.0,
      "step": 53540
    },
    {
      "entropy": 1.8533400312066077,
      "epoch": 0.16600031464111925,
      "grad_norm": 10.334144592285156,
      "learning_rate": 6.209294793811038e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8353090777993202,
      "num_tokens": 64369611.0,
      "step": 53550
    },
    {
      "entropy": 1.940661406517029,
      "epoch": 0.16603131376616895,
      "grad_norm": 11.576028823852539,
      "learning_rate": 6.208715098087144e-06,
      "loss": 0.5575,
      "mean_token_accuracy": 0.8241597592830658,
      "num_tokens": 64380934.0,
      "step": 53560
    },
    {
      "entropy": 1.9238759770989418,
      "epoch": 0.16606231289121864,
      "grad_norm": 6.117959022521973,
      "learning_rate": 6.208135564692989e-06,
      "loss": 0.581,
      "mean_token_accuracy": 0.824966461956501,
      "num_tokens": 64392121.0,
      "step": 53570
    },
    {
      "entropy": 1.9680738180875779,
      "epoch": 0.16609331201626834,
      "grad_norm": 6.631012916564941,
      "learning_rate": 6.207556193552824e-06,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8322376728057861,
      "num_tokens": 64403048.0,
      "step": 53580
    },
    {
      "entropy": 1.895513379573822,
      "epoch": 0.16612431114131804,
      "grad_norm": 9.87430191040039,
      "learning_rate": 6.206976984590952e-06,
      "loss": 0.6129,
      "mean_token_accuracy": 0.8248256593942642,
      "num_tokens": 64414999.0,
      "step": 53590
    },
    {
      "entropy": 1.8739793948829173,
      "epoch": 0.16615531026636773,
      "grad_norm": 10.234819412231445,
      "learning_rate": 6.206397937731726e-06,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8436161160469056,
      "num_tokens": 64427895.0,
      "step": 53600
    },
    {
      "entropy": 1.885368101298809,
      "epoch": 0.16618630939141743,
      "grad_norm": 7.840350151062012,
      "learning_rate": 6.205819052899549e-06,
      "loss": 0.5118,
      "mean_token_accuracy": 0.8369289055466652,
      "num_tokens": 64440003.0,
      "step": 53610
    },
    {
      "entropy": 1.9216457679867744,
      "epoch": 0.16621730851646713,
      "grad_norm": 9.213172912597656,
      "learning_rate": 6.205240330018869e-06,
      "loss": 0.5225,
      "mean_token_accuracy": 0.840261846780777,
      "num_tokens": 64451450.0,
      "step": 53620
    },
    {
      "entropy": 1.9085844457149506,
      "epoch": 0.16624830764151682,
      "grad_norm": 8.908021926879883,
      "learning_rate": 6.2046617690141876e-06,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8355207294225693,
      "num_tokens": 64463262.0,
      "step": 53630
    },
    {
      "entropy": 1.932268613576889,
      "epoch": 0.16627930676656652,
      "grad_norm": 8.519464492797852,
      "learning_rate": 6.204083369810055e-06,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8326381966471672,
      "num_tokens": 64474641.0,
      "step": 53640
    },
    {
      "entropy": 1.8569614686071874,
      "epoch": 0.16631030589161622,
      "grad_norm": 4.804983139038086,
      "learning_rate": 6.203505132331069e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8352816164493561,
      "num_tokens": 64487602.0,
      "step": 53650
    },
    {
      "entropy": 1.858240360021591,
      "epoch": 0.1663413050166659,
      "grad_norm": 10.345724105834961,
      "learning_rate": 6.202927056501878e-06,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8289370179176331,
      "num_tokens": 64500395.0,
      "step": 53660
    },
    {
      "entropy": 1.807939064502716,
      "epoch": 0.1663723041417156,
      "grad_norm": 7.798347473144531,
      "learning_rate": 6.202349142247179e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8403982803225517,
      "num_tokens": 64513854.0,
      "step": 53670
    },
    {
      "entropy": 1.972070437669754,
      "epoch": 0.1664033032667653,
      "grad_norm": 7.832572937011719,
      "learning_rate": 6.201771389491718e-06,
      "loss": 0.5794,
      "mean_token_accuracy": 0.8308521166443825,
      "num_tokens": 64524527.0,
      "step": 53680
    },
    {
      "entropy": 1.8211177855730056,
      "epoch": 0.166434302391815,
      "grad_norm": 7.727509498596191,
      "learning_rate": 6.20119379816029e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8501279547810554,
      "num_tokens": 64536806.0,
      "step": 53690
    },
    {
      "entropy": 1.9093163907527924,
      "epoch": 0.1664653015168647,
      "grad_norm": 8.725934982299805,
      "learning_rate": 6.20061636817774e-06,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8283161029219628,
      "num_tokens": 64547580.0,
      "step": 53700
    },
    {
      "entropy": 1.8534967467188834,
      "epoch": 0.16649630064191437,
      "grad_norm": 9.739397048950195,
      "learning_rate": 6.200039099468959e-06,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8432888746261596,
      "num_tokens": 64559322.0,
      "step": 53710
    },
    {
      "entropy": 1.8829628586769105,
      "epoch": 0.16652729976696407,
      "grad_norm": 9.243392944335938,
      "learning_rate": 6.1994619919588925e-06,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8410766318440437,
      "num_tokens": 64571411.0,
      "step": 53720
    },
    {
      "entropy": 1.850360631942749,
      "epoch": 0.16655829889201376,
      "grad_norm": 8.87844467163086,
      "learning_rate": 6.19888504557253e-06,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8418926939368248,
      "num_tokens": 64583534.0,
      "step": 53730
    },
    {
      "entropy": 1.7772665143013,
      "epoch": 0.16658929801706346,
      "grad_norm": 10.492433547973633,
      "learning_rate": 6.198308260234912e-06,
      "loss": 0.4102,
      "mean_token_accuracy": 0.8469656750559806,
      "num_tokens": 64596994.0,
      "step": 53740
    },
    {
      "entropy": 1.8638063728809358,
      "epoch": 0.16662029714211316,
      "grad_norm": 4.035283088684082,
      "learning_rate": 6.197731635871126e-06,
      "loss": 0.5047,
      "mean_token_accuracy": 0.8372615680098534,
      "num_tokens": 64608787.0,
      "step": 53750
    },
    {
      "entropy": 1.8116811349987985,
      "epoch": 0.16665129626716285,
      "grad_norm": 3.7760744094848633,
      "learning_rate": 6.197155172406311e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8445352002978325,
      "num_tokens": 64620766.0,
      "step": 53760
    },
    {
      "entropy": 1.764721181988716,
      "epoch": 0.16668229539221255,
      "grad_norm": 9.610373497009277,
      "learning_rate": 6.19657886976565e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8372749388217926,
      "num_tokens": 64634866.0,
      "step": 53770
    },
    {
      "entropy": 1.9135527536273003,
      "epoch": 0.16671329451726224,
      "grad_norm": 7.860201358795166,
      "learning_rate": 6.196002727874382e-06,
      "loss": 0.528,
      "mean_token_accuracy": 0.8299287378787994,
      "num_tokens": 64646883.0,
      "step": 53780
    },
    {
      "entropy": 1.9007146388292313,
      "epoch": 0.16674429364231194,
      "grad_norm": 3.8875370025634766,
      "learning_rate": 6.195426746657789e-06,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8316178724169732,
      "num_tokens": 64658973.0,
      "step": 53790
    },
    {
      "entropy": 1.826033054292202,
      "epoch": 0.16677529276736164,
      "grad_norm": 8.549474716186523,
      "learning_rate": 6.194850926041201e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.848523524403572,
      "num_tokens": 64671112.0,
      "step": 53800
    },
    {
      "entropy": 1.9412501826882362,
      "epoch": 0.16680629189241133,
      "grad_norm": 8.226346015930176,
      "learning_rate": 6.194275265950003e-06,
      "loss": 0.5854,
      "mean_token_accuracy": 0.817940565943718,
      "num_tokens": 64682585.0,
      "step": 53810
    },
    {
      "entropy": 1.9826242715120315,
      "epoch": 0.16683729101746103,
      "grad_norm": 7.8780622482299805,
      "learning_rate": 6.193699766309622e-06,
      "loss": 0.5808,
      "mean_token_accuracy": 0.8211043074727058,
      "num_tokens": 64693614.0,
      "step": 53820
    },
    {
      "entropy": 1.9336616814136505,
      "epoch": 0.16686829014251073,
      "grad_norm": 10.512943267822266,
      "learning_rate": 6.193124427045535e-06,
      "loss": 0.576,
      "mean_token_accuracy": 0.8250611051917076,
      "num_tokens": 64705081.0,
      "step": 53830
    },
    {
      "entropy": 1.8315666154026986,
      "epoch": 0.16689928926756042,
      "grad_norm": 9.55489730834961,
      "learning_rate": 6.1925492480832705e-06,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8315412878990174,
      "num_tokens": 64717880.0,
      "step": 53840
    },
    {
      "entropy": 1.878394363820553,
      "epoch": 0.16693028839261012,
      "grad_norm": 7.533677101135254,
      "learning_rate": 6.1919742293484e-06,
      "loss": 0.5741,
      "mean_token_accuracy": 0.8287183776497841,
      "num_tokens": 64728991.0,
      "step": 53850
    },
    {
      "entropy": 1.9339822947978973,
      "epoch": 0.16696128751765982,
      "grad_norm": 4.745457172393799,
      "learning_rate": 6.1913993707665485e-06,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8289982140064239,
      "num_tokens": 64740494.0,
      "step": 53860
    },
    {
      "entropy": 1.8400120854377746,
      "epoch": 0.16699228664270951,
      "grad_norm": 8.681490898132324,
      "learning_rate": 6.190824672263388e-06,
      "loss": 0.4626,
      "mean_token_accuracy": 0.8391780480742455,
      "num_tokens": 64753203.0,
      "step": 53870
    },
    {
      "entropy": 1.9312397927045821,
      "epoch": 0.1670232857677592,
      "grad_norm": 9.676896095275879,
      "learning_rate": 6.190250133764637e-06,
      "loss": 0.5628,
      "mean_token_accuracy": 0.832855261862278,
      "num_tokens": 64764277.0,
      "step": 53880
    },
    {
      "entropy": 1.8353455513715744,
      "epoch": 0.1670542848928089,
      "grad_norm": 7.824866771697998,
      "learning_rate": 6.189675755196064e-06,
      "loss": 0.4281,
      "mean_token_accuracy": 0.854548205435276,
      "num_tokens": 64776606.0,
      "step": 53890
    },
    {
      "entropy": 1.9409961223602294,
      "epoch": 0.1670852840178586,
      "grad_norm": 9.087993621826172,
      "learning_rate": 6.189101536483484e-06,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8235799849033356,
      "num_tokens": 64788237.0,
      "step": 53900
    },
    {
      "entropy": 1.8919391304254531,
      "epoch": 0.1671162831429083,
      "grad_norm": 7.690661430358887,
      "learning_rate": 6.188527477552761e-06,
      "loss": 0.564,
      "mean_token_accuracy": 0.832922050356865,
      "num_tokens": 64800456.0,
      "step": 53910
    },
    {
      "entropy": 1.8101703599095345,
      "epoch": 0.167147282267958,
      "grad_norm": 8.9353609085083,
      "learning_rate": 6.18795357832981e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8428716316819191,
      "num_tokens": 64813777.0,
      "step": 53920
    },
    {
      "entropy": 1.8997833237051964,
      "epoch": 0.1671782813930077,
      "grad_norm": 5.414516448974609,
      "learning_rate": 6.187379838740587e-06,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8327810063958168,
      "num_tokens": 64826025.0,
      "step": 53930
    },
    {
      "entropy": 1.9362305417656898,
      "epoch": 0.1672092805180574,
      "grad_norm": 7.484461784362793,
      "learning_rate": 6.186806258711105e-06,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8236838817596436,
      "num_tokens": 64837897.0,
      "step": 53940
    },
    {
      "entropy": 1.7711832091212272,
      "epoch": 0.1672402796431071,
      "grad_norm": 3.344655990600586,
      "learning_rate": 6.186232838167419e-06,
      "loss": 0.3899,
      "mean_token_accuracy": 0.847083292901516,
      "num_tokens": 64852441.0,
      "step": 53950
    },
    {
      "entropy": 1.8284263715147973,
      "epoch": 0.16727127876815676,
      "grad_norm": 7.023416042327881,
      "learning_rate": 6.185659577035632e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8410525426268578,
      "num_tokens": 64865349.0,
      "step": 53960
    },
    {
      "entropy": 1.9532892093062402,
      "epoch": 0.16730227789320645,
      "grad_norm": 7.449609756469727,
      "learning_rate": 6.185086475241898e-06,
      "loss": 0.5544,
      "mean_token_accuracy": 0.8271322563290596,
      "num_tokens": 64876306.0,
      "step": 53970
    },
    {
      "entropy": 1.943205328285694,
      "epoch": 0.16733327701825615,
      "grad_norm": 9.61139965057373,
      "learning_rate": 6.184513532712416e-06,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8366443976759911,
      "num_tokens": 64888292.0,
      "step": 53980
    },
    {
      "entropy": 1.9256486117839813,
      "epoch": 0.16736427614330585,
      "grad_norm": 5.424153804779053,
      "learning_rate": 6.183940749373436e-06,
      "loss": 0.5322,
      "mean_token_accuracy": 0.836666439473629,
      "num_tokens": 64899609.0,
      "step": 53990
    },
    {
      "entropy": 1.9205065920948983,
      "epoch": 0.16739527526835554,
      "grad_norm": 8.4097900390625,
      "learning_rate": 6.1833681251512516e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8369838654994964,
      "num_tokens": 64911133.0,
      "step": 54000
    },
    {
      "entropy": 1.837955741584301,
      "epoch": 0.16742627439340524,
      "grad_norm": 4.225302696228027,
      "learning_rate": 6.182795659972208e-06,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8357640966773033,
      "num_tokens": 64923539.0,
      "step": 54010
    },
    {
      "entropy": 1.87107597514987,
      "epoch": 0.16745727351845494,
      "grad_norm": 9.220149993896484,
      "learning_rate": 6.182223353762697e-06,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8416076585650444,
      "num_tokens": 64935263.0,
      "step": 54020
    },
    {
      "entropy": 1.8615114361047744,
      "epoch": 0.16748827264350463,
      "grad_norm": 8.91309928894043,
      "learning_rate": 6.181651206449155e-06,
      "loss": 0.4344,
      "mean_token_accuracy": 0.8517223030328751,
      "num_tokens": 64947376.0,
      "step": 54030
    },
    {
      "entropy": 1.8877154156565665,
      "epoch": 0.16751927176855433,
      "grad_norm": 4.553346633911133,
      "learning_rate": 6.181079217958073e-06,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8457445085048676,
      "num_tokens": 64959541.0,
      "step": 54040
    },
    {
      "entropy": 1.8997020781040193,
      "epoch": 0.16755027089360403,
      "grad_norm": 9.368854522705078,
      "learning_rate": 6.180507388215983e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8364048600196838,
      "num_tokens": 64972259.0,
      "step": 54050
    },
    {
      "entropy": 1.8942409604787827,
      "epoch": 0.16758127001865372,
      "grad_norm": 8.198491096496582,
      "learning_rate": 6.1799357171494655e-06,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8378923922777176,
      "num_tokens": 64983781.0,
      "step": 54060
    },
    {
      "entropy": 1.8540250420570374,
      "epoch": 0.16761226914370342,
      "grad_norm": 10.47086238861084,
      "learning_rate": 6.179364204685151e-06,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8316616043448448,
      "num_tokens": 64996187.0,
      "step": 54070
    },
    {
      "entropy": 1.9425106346607208,
      "epoch": 0.16764326826875312,
      "grad_norm": 9.815423965454102,
      "learning_rate": 6.1787928507497175e-06,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8328223183751107,
      "num_tokens": 65007642.0,
      "step": 54080
    },
    {
      "entropy": 1.7625815600156785,
      "epoch": 0.1676742673938028,
      "grad_norm": 4.722161769866943,
      "learning_rate": 6.178221655269889e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.851328319311142,
      "num_tokens": 65021921.0,
      "step": 54090
    },
    {
      "entropy": 1.877830518782139,
      "epoch": 0.1677052665188525,
      "grad_norm": 9.134954452514648,
      "learning_rate": 6.177650618172437e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8438004121184349,
      "num_tokens": 65033368.0,
      "step": 54100
    },
    {
      "entropy": 1.8599343091249465,
      "epoch": 0.1677362656439022,
      "grad_norm": 10.030871391296387,
      "learning_rate": 6.177079739384181e-06,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8409288361668587,
      "num_tokens": 65045878.0,
      "step": 54110
    },
    {
      "entropy": 1.7290209725499153,
      "epoch": 0.1677672647689519,
      "grad_norm": 4.038937568664551,
      "learning_rate": 6.176509018831986e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8499686598777771,
      "num_tokens": 65060347.0,
      "step": 54120
    },
    {
      "entropy": 1.8978177532553673,
      "epoch": 0.1677982638940016,
      "grad_norm": 8.696316719055176,
      "learning_rate": 6.175938456442767e-06,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8306406125426292,
      "num_tokens": 65071304.0,
      "step": 54130
    },
    {
      "entropy": 1.8237817287445068,
      "epoch": 0.1678292630190513,
      "grad_norm": 5.508317947387695,
      "learning_rate": 6.175368052143486e-06,
      "loss": 0.4792,
      "mean_token_accuracy": 0.84134771078825,
      "num_tokens": 65083895.0,
      "step": 54140
    },
    {
      "entropy": 1.8195566333830357,
      "epoch": 0.167860262144101,
      "grad_norm": 3.8358309268951416,
      "learning_rate": 6.174797805861148e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8476643204689026,
      "num_tokens": 65096233.0,
      "step": 54150
    },
    {
      "entropy": 1.8712603464722632,
      "epoch": 0.1678912612691507,
      "grad_norm": 8.841750144958496,
      "learning_rate": 6.174227717522813e-06,
      "loss": 0.5239,
      "mean_token_accuracy": 0.834831403195858,
      "num_tokens": 65108051.0,
      "step": 54160
    },
    {
      "entropy": 1.8733776569366456,
      "epoch": 0.16792226039420038,
      "grad_norm": 7.8258280754089355,
      "learning_rate": 6.173657787055579e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8338091254234314,
      "num_tokens": 65120148.0,
      "step": 54170
    },
    {
      "entropy": 1.7442950963974,
      "epoch": 0.16795325951925008,
      "grad_norm": 9.016060829162598,
      "learning_rate": 6.173088014386599e-06,
      "loss": 0.3985,
      "mean_token_accuracy": 0.85818080753088,
      "num_tokens": 65133990.0,
      "step": 54180
    },
    {
      "entropy": 1.8510777726769447,
      "epoch": 0.16798425864429978,
      "grad_norm": 4.038900852203369,
      "learning_rate": 6.1725183994430695e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8405930832028389,
      "num_tokens": 65145866.0,
      "step": 54190
    },
    {
      "entropy": 1.874160850048065,
      "epoch": 0.16801525776934945,
      "grad_norm": 9.50784969329834,
      "learning_rate": 6.1719489421522305e-06,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8350860238075256,
      "num_tokens": 65157550.0,
      "step": 54200
    },
    {
      "entropy": 1.8412716314196587,
      "epoch": 0.16804625689439914,
      "grad_norm": 4.003512859344482,
      "learning_rate": 6.1713796424413765e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8513872250914574,
      "num_tokens": 65169597.0,
      "step": 54210
    },
    {
      "entropy": 1.9005790546536445,
      "epoch": 0.16807725601944884,
      "grad_norm": 9.995362281799316,
      "learning_rate": 6.170810500237844e-06,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8309820532798767,
      "num_tokens": 65180854.0,
      "step": 54220
    },
    {
      "entropy": 1.8263572067022324,
      "epoch": 0.16810825514449854,
      "grad_norm": 7.676513195037842,
      "learning_rate": 6.170241515469018e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8502917841076851,
      "num_tokens": 65192720.0,
      "step": 54230
    },
    {
      "entropy": 1.832256343960762,
      "epoch": 0.16813925426954823,
      "grad_norm": 4.75320291519165,
      "learning_rate": 6.1696726880623285e-06,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8385294854640961,
      "num_tokens": 65204655.0,
      "step": 54240
    },
    {
      "entropy": 1.8810152530670166,
      "epoch": 0.16817025339459793,
      "grad_norm": 9.239448547363281,
      "learning_rate": 6.1691040179452545e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.832660236954689,
      "num_tokens": 65216115.0,
      "step": 54250
    },
    {
      "entropy": 1.7937642320990563,
      "epoch": 0.16820125251964763,
      "grad_norm": 8.329822540283203,
      "learning_rate": 6.168535505045321e-06,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8271018177270889,
      "num_tokens": 65229542.0,
      "step": 54260
    },
    {
      "entropy": 1.8577524468302726,
      "epoch": 0.16823225164469732,
      "grad_norm": 3.8464255332946777,
      "learning_rate": 6.1679671492901005e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8379319593310356,
      "num_tokens": 65241000.0,
      "step": 54270
    },
    {
      "entropy": 1.8444037914276123,
      "epoch": 0.16826325076974702,
      "grad_norm": 9.02126407623291,
      "learning_rate": 6.167398950607211e-06,
      "loss": 0.4539,
      "mean_token_accuracy": 0.8464455410838128,
      "num_tokens": 65253095.0,
      "step": 54280
    },
    {
      "entropy": 1.8382619485259055,
      "epoch": 0.16829424989479672,
      "grad_norm": 8.693649291992188,
      "learning_rate": 6.166830908924317e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8406226679682731,
      "num_tokens": 65265949.0,
      "step": 54290
    },
    {
      "entropy": 1.9115242898464202,
      "epoch": 0.1683252490198464,
      "grad_norm": 3.702934741973877,
      "learning_rate": 6.16626302416913e-06,
      "loss": 0.5333,
      "mean_token_accuracy": 0.833556392788887,
      "num_tokens": 65277257.0,
      "step": 54300
    },
    {
      "entropy": 1.8722372770309448,
      "epoch": 0.1683562481448961,
      "grad_norm": 9.893916130065918,
      "learning_rate": 6.16569529626941e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.8460499599575997,
      "num_tokens": 65288847.0,
      "step": 54310
    },
    {
      "entropy": 1.8180507212877273,
      "epoch": 0.1683872472699458,
      "grad_norm": 9.089763641357422,
      "learning_rate": 6.165127725152958e-06,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8375856161117554,
      "num_tokens": 65301059.0,
      "step": 54320
    },
    {
      "entropy": 1.8814466312527656,
      "epoch": 0.1684182463949955,
      "grad_norm": 8.537693977355957,
      "learning_rate": 6.1645603107476316e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8440789937973022,
      "num_tokens": 65313029.0,
      "step": 54330
    },
    {
      "entropy": 1.8159365728497505,
      "epoch": 0.1684492455200452,
      "grad_norm": 3.978240489959717,
      "learning_rate": 6.163993052981323e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8402403250336647,
      "num_tokens": 65325995.0,
      "step": 54340
    },
    {
      "entropy": 1.8512892082333565,
      "epoch": 0.1684802446450949,
      "grad_norm": 3.7935879230499268,
      "learning_rate": 6.163425951781979e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8360736206173897,
      "num_tokens": 65338118.0,
      "step": 54350
    },
    {
      "entropy": 1.802795946598053,
      "epoch": 0.1685112437701446,
      "grad_norm": 8.774079322814941,
      "learning_rate": 6.162859007077591e-06,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8323176607489586,
      "num_tokens": 65350124.0,
      "step": 54360
    },
    {
      "entropy": 1.9113412827253342,
      "epoch": 0.1685422428951943,
      "grad_norm": 9.643198013305664,
      "learning_rate": 6.162292218796194e-06,
      "loss": 0.6128,
      "mean_token_accuracy": 0.821797750890255,
      "num_tokens": 65361216.0,
      "step": 54370
    },
    {
      "entropy": 1.880385261774063,
      "epoch": 0.16857324202024399,
      "grad_norm": 4.427012920379639,
      "learning_rate": 6.161725586865874e-06,
      "loss": 0.514,
      "mean_token_accuracy": 0.8347649067640305,
      "num_tokens": 65372951.0,
      "step": 54380
    },
    {
      "entropy": 1.8761203557252883,
      "epoch": 0.16860424114529368,
      "grad_norm": 7.965682506561279,
      "learning_rate": 6.1611591112147576e-06,
      "loss": 0.5154,
      "mean_token_accuracy": 0.8341512382030487,
      "num_tokens": 65384596.0,
      "step": 54390
    },
    {
      "entropy": 1.9145608723163605,
      "epoch": 0.16863524027034338,
      "grad_norm": 9.253949165344238,
      "learning_rate": 6.160592791771022e-06,
      "loss": 0.5824,
      "mean_token_accuracy": 0.8162755355238914,
      "num_tokens": 65396034.0,
      "step": 54400
    },
    {
      "entropy": 1.831618282198906,
      "epoch": 0.16866623939539307,
      "grad_norm": 8.794281005859375,
      "learning_rate": 6.16002662846289e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8415708675980568,
      "num_tokens": 65408333.0,
      "step": 54410
    },
    {
      "entropy": 1.8698890820145606,
      "epoch": 0.16869723852044277,
      "grad_norm": 9.203094482421875,
      "learning_rate": 6.159460621218628e-06,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8259828209877014,
      "num_tokens": 65420329.0,
      "step": 54420
    },
    {
      "entropy": 1.8370443254709243,
      "epoch": 0.16872823764549247,
      "grad_norm": 4.42335319519043,
      "learning_rate": 6.158894769966554e-06,
      "loss": 0.51,
      "mean_token_accuracy": 0.8294874534010888,
      "num_tokens": 65432554.0,
      "step": 54430
    },
    {
      "entropy": 1.929424050450325,
      "epoch": 0.16875923677054216,
      "grad_norm": 8.103418350219727,
      "learning_rate": 6.158329074635024e-06,
      "loss": 0.6025,
      "mean_token_accuracy": 0.8246657729148865,
      "num_tokens": 65443420.0,
      "step": 54440
    },
    {
      "entropy": 1.9033832669258117,
      "epoch": 0.16879023589559183,
      "grad_norm": 11.029186248779297,
      "learning_rate": 6.157763535152448e-06,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8389760866761208,
      "num_tokens": 65455561.0,
      "step": 54450
    },
    {
      "entropy": 1.8955723971128464,
      "epoch": 0.16882123502064153,
      "grad_norm": 10.687283515930176,
      "learning_rate": 6.1571981514472765e-06,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8430651158094407,
      "num_tokens": 65467045.0,
      "step": 54460
    },
    {
      "entropy": 1.9213042184710503,
      "epoch": 0.16885223414569123,
      "grad_norm": 8.316679000854492,
      "learning_rate": 6.156632923448008e-06,
      "loss": 0.5787,
      "mean_token_accuracy": 0.8268252417445183,
      "num_tokens": 65479119.0,
      "step": 54470
    },
    {
      "entropy": 1.8560591742396355,
      "epoch": 0.16888323327074092,
      "grad_norm": 4.520045757293701,
      "learning_rate": 6.156067851083189e-06,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8373791992664337,
      "num_tokens": 65491055.0,
      "step": 54480
    },
    {
      "entropy": 1.8120755463838578,
      "epoch": 0.16891423239579062,
      "grad_norm": 9.259931564331055,
      "learning_rate": 6.1555029342814085e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8388746708631516,
      "num_tokens": 65503547.0,
      "step": 54490
    },
    {
      "entropy": 1.8446748018264771,
      "epoch": 0.16894523152084032,
      "grad_norm": 5.609341144561768,
      "learning_rate": 6.154938172971303e-06,
      "loss": 0.517,
      "mean_token_accuracy": 0.8386319100856781,
      "num_tokens": 65516410.0,
      "step": 54500
    },
    {
      "entropy": 1.7979810684919357,
      "epoch": 0.16897623064589,
      "grad_norm": 8.738170623779297,
      "learning_rate": 6.154373567081555e-06,
      "loss": 0.4665,
      "mean_token_accuracy": 0.8370955526828766,
      "num_tokens": 65529389.0,
      "step": 54510
    },
    {
      "entropy": 1.9188962578773499,
      "epoch": 0.1690072297709397,
      "grad_norm": 8.479401588439941,
      "learning_rate": 6.15380911654089e-06,
      "loss": 0.536,
      "mean_token_accuracy": 0.8272537529468537,
      "num_tokens": 65541159.0,
      "step": 54520
    },
    {
      "entropy": 1.8414244174957275,
      "epoch": 0.1690382288959894,
      "grad_norm": 9.82351016998291,
      "learning_rate": 6.153244821278084e-06,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8392342895269393,
      "num_tokens": 65553128.0,
      "step": 54530
    },
    {
      "entropy": 1.9202020570635796,
      "epoch": 0.1690692280210391,
      "grad_norm": 7.951665878295898,
      "learning_rate": 6.152680681221957e-06,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8311819225549698,
      "num_tokens": 65564486.0,
      "step": 54540
    },
    {
      "entropy": 1.7439219117164613,
      "epoch": 0.1691002271460888,
      "grad_norm": 7.696526527404785,
      "learning_rate": 6.1521166963013704e-06,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8509344890713691,
      "num_tokens": 65578178.0,
      "step": 54550
    },
    {
      "entropy": 1.8539911389350892,
      "epoch": 0.1691312262711385,
      "grad_norm": 8.005914688110352,
      "learning_rate": 6.151552866445237e-06,
      "loss": 0.487,
      "mean_token_accuracy": 0.8328571185469628,
      "num_tokens": 65590436.0,
      "step": 54560
    },
    {
      "entropy": 1.7372994154691697,
      "epoch": 0.1691622253961882,
      "grad_norm": 8.57450008392334,
      "learning_rate": 6.150989191582515e-06,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8435987919569016,
      "num_tokens": 65604247.0,
      "step": 54570
    },
    {
      "entropy": 1.8149765402078628,
      "epoch": 0.1691932245212379,
      "grad_norm": 9.524360656738281,
      "learning_rate": 6.150425671642202e-06,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8407787919044495,
      "num_tokens": 65616338.0,
      "step": 54580
    },
    {
      "entropy": 1.7159005388617516,
      "epoch": 0.16922422364628759,
      "grad_norm": 8.246034622192383,
      "learning_rate": 6.1498623065533485e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8674893081188202,
      "num_tokens": 65629499.0,
      "step": 54590
    },
    {
      "entropy": 1.7766979977488517,
      "epoch": 0.16925522277133728,
      "grad_norm": 3.937098741531372,
      "learning_rate": 6.1492990962450465e-06,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8527488321065902,
      "num_tokens": 65641287.0,
      "step": 54600
    },
    {
      "entropy": 1.8135869204998016,
      "epoch": 0.16928622189638698,
      "grad_norm": 9.094282150268555,
      "learning_rate": 6.148736040646432e-06,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8407742142677307,
      "num_tokens": 65653614.0,
      "step": 54610
    },
    {
      "entropy": 1.7666838884353637,
      "epoch": 0.16931722102143668,
      "grad_norm": 8.157979011535645,
      "learning_rate": 6.148173139686692e-06,
      "loss": 0.4819,
      "mean_token_accuracy": 0.845091213285923,
      "num_tokens": 65666332.0,
      "step": 54620
    },
    {
      "entropy": 1.7945738911628724,
      "epoch": 0.16934822014648637,
      "grad_norm": 4.526710510253906,
      "learning_rate": 6.147610393295055e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8315363183617592,
      "num_tokens": 65679667.0,
      "step": 54630
    },
    {
      "entropy": 1.8847228914499283,
      "epoch": 0.16937921927153607,
      "grad_norm": 8.596298217773438,
      "learning_rate": 6.147047801400793e-06,
      "loss": 0.569,
      "mean_token_accuracy": 0.8253517493605613,
      "num_tokens": 65690829.0,
      "step": 54640
    },
    {
      "entropy": 1.9206166476011277,
      "epoch": 0.16941021839658577,
      "grad_norm": 7.514388561248779,
      "learning_rate": 6.146485363933227e-06,
      "loss": 0.5221,
      "mean_token_accuracy": 0.8408697739243507,
      "num_tokens": 65702308.0,
      "step": 54650
    },
    {
      "entropy": 1.8065496653318405,
      "epoch": 0.16944121752163546,
      "grad_norm": 8.206125259399414,
      "learning_rate": 6.145923080821722e-06,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8288014903664589,
      "num_tokens": 65715353.0,
      "step": 54660
    },
    {
      "entropy": 1.8215434283018113,
      "epoch": 0.16947221664668516,
      "grad_norm": 10.296819686889648,
      "learning_rate": 6.145360951995688e-06,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8371994882822037,
      "num_tokens": 65727578.0,
      "step": 54670
    },
    {
      "entropy": 1.8798654437065125,
      "epoch": 0.16950321577173486,
      "grad_norm": 8.519831657409668,
      "learning_rate": 6.144798977384581e-06,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8545126497745514,
      "num_tokens": 65739521.0,
      "step": 54680
    },
    {
      "entropy": 1.9068218991160393,
      "epoch": 0.16953421489678455,
      "grad_norm": 7.611881256103516,
      "learning_rate": 6.144237156917899e-06,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8298189043998718,
      "num_tokens": 65751188.0,
      "step": 54690
    },
    {
      "entropy": 1.8502942398190498,
      "epoch": 0.16956521402183422,
      "grad_norm": 8.908893585205078,
      "learning_rate": 6.143675490525191e-06,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8421571254730225,
      "num_tokens": 65763519.0,
      "step": 54700
    },
    {
      "entropy": 1.926678617298603,
      "epoch": 0.16959621314688392,
      "grad_norm": 8.78310489654541,
      "learning_rate": 6.143113978136046e-06,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8322173491120338,
      "num_tokens": 65775010.0,
      "step": 54710
    },
    {
      "entropy": 1.9700055122375488,
      "epoch": 0.16962721227193361,
      "grad_norm": 9.067830085754395,
      "learning_rate": 6.1425526196801e-06,
      "loss": 0.5819,
      "mean_token_accuracy": 0.824844454228878,
      "num_tokens": 65785503.0,
      "step": 54720
    },
    {
      "entropy": 1.8520893216133119,
      "epoch": 0.1696582113969833,
      "grad_norm": 4.218345642089844,
      "learning_rate": 6.1419914150870315e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8238402858376503,
      "num_tokens": 65799211.0,
      "step": 54730
    },
    {
      "entropy": 1.9024053156375884,
      "epoch": 0.169689210522033,
      "grad_norm": 5.589046001434326,
      "learning_rate": 6.141430364286569e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8389298126101494,
      "num_tokens": 65811181.0,
      "step": 54740
    },
    {
      "entropy": 1.8264511436223985,
      "epoch": 0.1697202096470827,
      "grad_norm": 9.439501762390137,
      "learning_rate": 6.140869467208483e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.839516569674015,
      "num_tokens": 65824036.0,
      "step": 54750
    },
    {
      "entropy": 1.9636918157339096,
      "epoch": 0.1697512087721324,
      "grad_norm": 8.578853607177734,
      "learning_rate": 6.140308723782587e-06,
      "loss": 0.5827,
      "mean_token_accuracy": 0.8268523201346397,
      "num_tokens": 65834646.0,
      "step": 54760
    },
    {
      "entropy": 1.8637161239981652,
      "epoch": 0.1697822078971821,
      "grad_norm": 8.782843589782715,
      "learning_rate": 6.1397481339387444e-06,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8350477933883667,
      "num_tokens": 65846256.0,
      "step": 54770
    },
    {
      "entropy": 1.8629076793789863,
      "epoch": 0.1698132070222318,
      "grad_norm": 8.186110496520996,
      "learning_rate": 6.139187697606855e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.84441829174757,
      "num_tokens": 65858192.0,
      "step": 54780
    },
    {
      "entropy": 1.8901616916060449,
      "epoch": 0.1698442061472815,
      "grad_norm": 9.074200630187988,
      "learning_rate": 6.138627414716874e-06,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8320909798145294,
      "num_tokens": 65869497.0,
      "step": 54790
    },
    {
      "entropy": 1.8830202847719193,
      "epoch": 0.1698752052723312,
      "grad_norm": 8.882548332214355,
      "learning_rate": 6.138067285198796e-06,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8349261358380318,
      "num_tokens": 65881627.0,
      "step": 54800
    },
    {
      "entropy": 1.8568723633885384,
      "epoch": 0.16990620439738088,
      "grad_norm": 8.651144981384277,
      "learning_rate": 6.1375073089826556e-06,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8417845577001571,
      "num_tokens": 65894013.0,
      "step": 54810
    },
    {
      "entropy": 1.6487878575921058,
      "epoch": 0.16993720352243058,
      "grad_norm": 2.7366909980773926,
      "learning_rate": 6.13694748599854e-06,
      "loss": 0.3365,
      "mean_token_accuracy": 0.862517063319683,
      "num_tokens": 65909433.0,
      "step": 54820
    },
    {
      "entropy": 1.9101332992315292,
      "epoch": 0.16996820264748028,
      "grad_norm": 9.56188678741455,
      "learning_rate": 6.136387816176578e-06,
      "loss": 0.596,
      "mean_token_accuracy": 0.8215406507253646,
      "num_tokens": 65921625.0,
      "step": 54830
    },
    {
      "entropy": 1.7827743321657181,
      "epoch": 0.16999920177252997,
      "grad_norm": 8.454669952392578,
      "learning_rate": 6.135828299446942e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8514051124453544,
      "num_tokens": 65934681.0,
      "step": 54840
    },
    {
      "entropy": 1.8705067232251167,
      "epoch": 0.17003020089757967,
      "grad_norm": 8.229545593261719,
      "learning_rate": 6.135268935739851e-06,
      "loss": 0.4988,
      "mean_token_accuracy": 0.847279503941536,
      "num_tokens": 65946087.0,
      "step": 54850
    },
    {
      "entropy": 1.7427136436104775,
      "epoch": 0.17006120002262937,
      "grad_norm": 10.23863697052002,
      "learning_rate": 6.134709724985567e-06,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8420625299215316,
      "num_tokens": 65959753.0,
      "step": 54860
    },
    {
      "entropy": 1.8604324340820313,
      "epoch": 0.17009219914767906,
      "grad_norm": 9.759727478027344,
      "learning_rate": 6.134150667114395e-06,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8294590935111046,
      "num_tokens": 65971696.0,
      "step": 54870
    },
    {
      "entropy": 1.8827021196484566,
      "epoch": 0.17012319827272876,
      "grad_norm": 8.212512969970703,
      "learning_rate": 6.133591762056689e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8433489888906479,
      "num_tokens": 65983076.0,
      "step": 54880
    },
    {
      "entropy": 1.8750552967190743,
      "epoch": 0.17015419739777846,
      "grad_norm": 9.474791526794434,
      "learning_rate": 6.133033009742842e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8393171012401581,
      "num_tokens": 65994772.0,
      "step": 54890
    },
    {
      "entropy": 1.7940944850444793,
      "epoch": 0.17018519652282815,
      "grad_norm": 4.888198375701904,
      "learning_rate": 6.132474410103298e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8466850072145462,
      "num_tokens": 66007645.0,
      "step": 54900
    },
    {
      "entropy": 1.8010751977562904,
      "epoch": 0.17021619564787785,
      "grad_norm": 8.661337852478027,
      "learning_rate": 6.131915963068537e-06,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8513104304671287,
      "num_tokens": 66020326.0,
      "step": 54910
    },
    {
      "entropy": 1.7909870207309724,
      "epoch": 0.17024719477292755,
      "grad_norm": 7.634841442108154,
      "learning_rate": 6.13135766856909e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8531160518527031,
      "num_tokens": 66033075.0,
      "step": 54920
    },
    {
      "entropy": 1.8581923857331275,
      "epoch": 0.17027819389797724,
      "grad_norm": 8.779280662536621,
      "learning_rate": 6.130799526535529e-06,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8206439360976219,
      "num_tokens": 66046620.0,
      "step": 54930
    },
    {
      "entropy": 1.8611151441931724,
      "epoch": 0.1703091930230269,
      "grad_norm": 4.603653430938721,
      "learning_rate": 6.1302415368984725e-06,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8282294601202012,
      "num_tokens": 66059390.0,
      "step": 54940
    },
    {
      "entropy": 1.9167177721858024,
      "epoch": 0.1703401921480766,
      "grad_norm": 3.782064199447632,
      "learning_rate": 6.129683699588581e-06,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8340412959456444,
      "num_tokens": 66071324.0,
      "step": 54950
    },
    {
      "entropy": 1.7558236002922059,
      "epoch": 0.1703711912731263,
      "grad_norm": 3.8991594314575195,
      "learning_rate": 6.129126014536561e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.8545200228691101,
      "num_tokens": 66084375.0,
      "step": 54960
    },
    {
      "entropy": 1.9062550529837607,
      "epoch": 0.170402190398176,
      "grad_norm": 3.855738639831543,
      "learning_rate": 6.12856848167316e-06,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8293624669313431,
      "num_tokens": 66096293.0,
      "step": 54970
    },
    {
      "entropy": 1.9391276210546493,
      "epoch": 0.1704331895232257,
      "grad_norm": 8.4966402053833,
      "learning_rate": 6.1280111009291744e-06,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8367387875914574,
      "num_tokens": 66106920.0,
      "step": 54980
    },
    {
      "entropy": 1.8991558268666267,
      "epoch": 0.1704641886482754,
      "grad_norm": 5.249369144439697,
      "learning_rate": 6.127453872235442e-06,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8212674275040627,
      "num_tokens": 66118337.0,
      "step": 54990
    },
    {
      "entropy": 1.8485107704997064,
      "epoch": 0.1704951877733251,
      "grad_norm": 9.159661293029785,
      "learning_rate": 6.1268967955228405e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8363189935684204,
      "num_tokens": 66130596.0,
      "step": 55000
    },
    {
      "entropy": 1.851409375667572,
      "epoch": 0.1705261868983748,
      "grad_norm": 8.441533088684082,
      "learning_rate": 6.126339870722301e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8494085937738418,
      "num_tokens": 66142188.0,
      "step": 55010
    },
    {
      "entropy": 1.9319586962461472,
      "epoch": 0.17055718602342448,
      "grad_norm": 9.650931358337402,
      "learning_rate": 6.12578309776479e-06,
      "loss": 0.5737,
      "mean_token_accuracy": 0.8261142611503601,
      "num_tokens": 66153013.0,
      "step": 55020
    },
    {
      "entropy": 1.8354326650500297,
      "epoch": 0.17058818514847418,
      "grad_norm": 3.9305970668792725,
      "learning_rate": 6.125226476581324e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8330473214387893,
      "num_tokens": 66165874.0,
      "step": 55030
    },
    {
      "entropy": 1.9324169605970383,
      "epoch": 0.17061918427352388,
      "grad_norm": 4.871352672576904,
      "learning_rate": 6.124670007102958e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8405327409505844,
      "num_tokens": 66176898.0,
      "step": 55040
    },
    {
      "entropy": 1.8468404799699782,
      "epoch": 0.17065018339857357,
      "grad_norm": 9.20655632019043,
      "learning_rate": 6.124113689260793e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8357664123177528,
      "num_tokens": 66188431.0,
      "step": 55050
    },
    {
      "entropy": 1.7879943639039992,
      "epoch": 0.17068118252362327,
      "grad_norm": 3.959418296813965,
      "learning_rate": 6.123557522985977e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8417293280363083,
      "num_tokens": 66201593.0,
      "step": 55060
    },
    {
      "entropy": 1.924604320526123,
      "epoch": 0.17071218164867297,
      "grad_norm": 8.743783950805664,
      "learning_rate": 6.123001508209696e-06,
      "loss": 0.6023,
      "mean_token_accuracy": 0.8270179316401481,
      "num_tokens": 66212878.0,
      "step": 55070
    },
    {
      "entropy": 1.9165123641490935,
      "epoch": 0.17074318077372266,
      "grad_norm": 7.5204949378967285,
      "learning_rate": 6.122445644863187e-06,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8256003141403199,
      "num_tokens": 66224720.0,
      "step": 55080
    },
    {
      "entropy": 1.889313419163227,
      "epoch": 0.17077417989877236,
      "grad_norm": 8.590926170349121,
      "learning_rate": 6.121889932877719e-06,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8257635533809662,
      "num_tokens": 66236384.0,
      "step": 55090
    },
    {
      "entropy": 1.870141127705574,
      "epoch": 0.17080517902382206,
      "grad_norm": 3.6843600273132324,
      "learning_rate": 6.121334372184618e-06,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8345419481396675,
      "num_tokens": 66248946.0,
      "step": 55100
    },
    {
      "entropy": 1.8741691306233406,
      "epoch": 0.17083617814887175,
      "grad_norm": 9.643596649169922,
      "learning_rate": 6.120778962715248e-06,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8437133207917213,
      "num_tokens": 66260837.0,
      "step": 55110
    },
    {
      "entropy": 1.8679534777998925,
      "epoch": 0.17086717727392145,
      "grad_norm": 4.9696149826049805,
      "learning_rate": 6.120223704401012e-06,
      "loss": 0.4552,
      "mean_token_accuracy": 0.8462657079100608,
      "num_tokens": 66272827.0,
      "step": 55120
    },
    {
      "entropy": 1.8148210868239403,
      "epoch": 0.17089817639897115,
      "grad_norm": 8.890388488769531,
      "learning_rate": 6.119668597173365e-06,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8454957276582717,
      "num_tokens": 66285297.0,
      "step": 55130
    },
    {
      "entropy": 1.8707555949687957,
      "epoch": 0.17092917552402084,
      "grad_norm": 8.708013534545898,
      "learning_rate": 6.119113640963797e-06,
      "loss": 0.5601,
      "mean_token_accuracy": 0.8221472725272179,
      "num_tokens": 66296618.0,
      "step": 55140
    },
    {
      "entropy": 1.8355353966355323,
      "epoch": 0.17096017464907054,
      "grad_norm": 9.65738296508789,
      "learning_rate": 6.11855883570385e-06,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8310455739498138,
      "num_tokens": 66309122.0,
      "step": 55150
    },
    {
      "entropy": 1.8585526466369628,
      "epoch": 0.17099117377412024,
      "grad_norm": 9.257162094116211,
      "learning_rate": 6.118004181325103e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8375745385885238,
      "num_tokens": 66321433.0,
      "step": 55160
    },
    {
      "entropy": 1.8798829302191735,
      "epoch": 0.17102217289916993,
      "grad_norm": 8.34717845916748,
      "learning_rate": 6.117449677759181e-06,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8309325784444809,
      "num_tokens": 66333576.0,
      "step": 55170
    },
    {
      "entropy": 1.9052462548017501,
      "epoch": 0.17105317202421963,
      "grad_norm": 9.19088363647461,
      "learning_rate": 6.11689532493775e-06,
      "loss": 0.5508,
      "mean_token_accuracy": 0.8245049402117729,
      "num_tokens": 66345424.0,
      "step": 55180
    },
    {
      "entropy": 1.8513251379132272,
      "epoch": 0.1710841711492693,
      "grad_norm": 4.772027969360352,
      "learning_rate": 6.1163411227925265e-06,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8387829244136811,
      "num_tokens": 66357556.0,
      "step": 55190
    },
    {
      "entropy": 1.8194545328617096,
      "epoch": 0.171115170274319,
      "grad_norm": 8.34814453125,
      "learning_rate": 6.11578707125526e-06,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8422711014747619,
      "num_tokens": 66370273.0,
      "step": 55200
    },
    {
      "entropy": 1.8819702237844467,
      "epoch": 0.1711461693993687,
      "grad_norm": 8.056477546691895,
      "learning_rate": 6.1152331702577514e-06,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8342429891228675,
      "num_tokens": 66382365.0,
      "step": 55210
    },
    {
      "entropy": 1.898891557753086,
      "epoch": 0.1711771685244184,
      "grad_norm": 4.703845977783203,
      "learning_rate": 6.114679419731841e-06,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8460740655660629,
      "num_tokens": 66393344.0,
      "step": 55220
    },
    {
      "entropy": 1.957570144534111,
      "epoch": 0.17120816764946808,
      "grad_norm": 7.900186061859131,
      "learning_rate": 6.114125819609411e-06,
      "loss": 0.5556,
      "mean_token_accuracy": 0.8310376718640328,
      "num_tokens": 66404439.0,
      "step": 55230
    },
    {
      "entropy": 1.8322395712137223,
      "epoch": 0.17123916677451778,
      "grad_norm": 3.6799228191375732,
      "learning_rate": 6.113572369822391e-06,
      "loss": 0.5098,
      "mean_token_accuracy": 0.8300682485103608,
      "num_tokens": 66417136.0,
      "step": 55240
    },
    {
      "entropy": 1.8394572094082833,
      "epoch": 0.17127016589956748,
      "grad_norm": 4.595379829406738,
      "learning_rate": 6.113019070302754e-06,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8420197859406471,
      "num_tokens": 66429039.0,
      "step": 55250
    },
    {
      "entropy": 1.8247755616903305,
      "epoch": 0.17130116502461717,
      "grad_norm": 7.622194290161133,
      "learning_rate": 6.11246592098251e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8401867300271988,
      "num_tokens": 66441004.0,
      "step": 55260
    },
    {
      "entropy": 1.858515314757824,
      "epoch": 0.17133216414966687,
      "grad_norm": 9.903068542480469,
      "learning_rate": 6.111912921793715e-06,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8486079826951027,
      "num_tokens": 66452827.0,
      "step": 55270
    },
    {
      "entropy": 1.8595426589250565,
      "epoch": 0.17136316327471657,
      "grad_norm": 6.444674968719482,
      "learning_rate": 6.111360072668473e-06,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8341106563806534,
      "num_tokens": 66464688.0,
      "step": 55280
    },
    {
      "entropy": 1.7985023483633995,
      "epoch": 0.17139416239976626,
      "grad_norm": 9.139451026916504,
      "learning_rate": 6.110807373538924e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8438055410981178,
      "num_tokens": 66478770.0,
      "step": 55290
    },
    {
      "entropy": 1.8379919603466988,
      "epoch": 0.17142516152481596,
      "grad_norm": 2.0652432441711426,
      "learning_rate": 6.110254824337254e-06,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8349084377288818,
      "num_tokens": 66491604.0,
      "step": 55300
    },
    {
      "entropy": 1.8246984764933587,
      "epoch": 0.17145616064986566,
      "grad_norm": 4.141010284423828,
      "learning_rate": 6.109702424995692e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.8502539679408073,
      "num_tokens": 66503887.0,
      "step": 55310
    },
    {
      "entropy": 1.7455498218536376,
      "epoch": 0.17148715977491535,
      "grad_norm": 7.454645156860352,
      "learning_rate": 6.1091501754465084e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8398867219686508,
      "num_tokens": 66516981.0,
      "step": 55320
    },
    {
      "entropy": 1.8381555840373038,
      "epoch": 0.17151815889996505,
      "grad_norm": 8.063959121704102,
      "learning_rate": 6.10859807562202e-06,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8499990120530129,
      "num_tokens": 66528982.0,
      "step": 55330
    },
    {
      "entropy": 1.9049826204776763,
      "epoch": 0.17154915802501475,
      "grad_norm": 8.288491249084473,
      "learning_rate": 6.108046125454582e-06,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8419345527887344,
      "num_tokens": 66540768.0,
      "step": 55340
    },
    {
      "entropy": 1.8578782469034194,
      "epoch": 0.17158015715006444,
      "grad_norm": 7.979318618774414,
      "learning_rate": 6.107494324876594e-06,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8317851856350899,
      "num_tokens": 66552308.0,
      "step": 55350
    },
    {
      "entropy": 1.8647135615348815,
      "epoch": 0.17161115627511414,
      "grad_norm": 4.911669731140137,
      "learning_rate": 6.1069426738205e-06,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8424048766493797,
      "num_tokens": 66564209.0,
      "step": 55360
    },
    {
      "entropy": 1.9092617228627204,
      "epoch": 0.17164215540016384,
      "grad_norm": 9.604966163635254,
      "learning_rate": 6.106391172218784e-06,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8417677640914917,
      "num_tokens": 66575424.0,
      "step": 55370
    },
    {
      "entropy": 1.8798686414957047,
      "epoch": 0.17167315452521353,
      "grad_norm": 9.60474681854248,
      "learning_rate": 6.105839820003976e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8326731339097023,
      "num_tokens": 66588461.0,
      "step": 55380
    },
    {
      "entropy": 1.787381762266159,
      "epoch": 0.17170415365026323,
      "grad_norm": 8.320876121520996,
      "learning_rate": 6.105288617108646e-06,
      "loss": 0.4899,
      "mean_token_accuracy": 0.8443598464131356,
      "num_tokens": 66601689.0,
      "step": 55390
    },
    {
      "entropy": 1.8882825881242753,
      "epoch": 0.17173515277531293,
      "grad_norm": 8.86194896697998,
      "learning_rate": 6.104737563465406e-06,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8265114605426789,
      "num_tokens": 66613156.0,
      "step": 55400
    },
    {
      "entropy": 1.9447656571865082,
      "epoch": 0.17176615190036262,
      "grad_norm": 10.87563419342041,
      "learning_rate": 6.104186659006913e-06,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8329192861914635,
      "num_tokens": 66623750.0,
      "step": 55410
    },
    {
      "entropy": 1.862581080198288,
      "epoch": 0.17179715102541232,
      "grad_norm": 9.020846366882324,
      "learning_rate": 6.103635903665865e-06,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8274956732988358,
      "num_tokens": 66636578.0,
      "step": 55420
    },
    {
      "entropy": 1.8903913646936417,
      "epoch": 0.17182815015046202,
      "grad_norm": 8.09145736694336,
      "learning_rate": 6.103085297375004e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8357395201921463,
      "num_tokens": 66648330.0,
      "step": 55430
    },
    {
      "entropy": 1.9172933578491211,
      "epoch": 0.17185914927551169,
      "grad_norm": 4.03477144241333,
      "learning_rate": 6.102534840067114e-06,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8339840278029442,
      "num_tokens": 66660056.0,
      "step": 55440
    },
    {
      "entropy": 1.9062219202518462,
      "epoch": 0.17189014840056138,
      "grad_norm": 9.107460021972656,
      "learning_rate": 6.101984531675016e-06,
      "loss": 0.565,
      "mean_token_accuracy": 0.833404429256916,
      "num_tokens": 66670678.0,
      "step": 55450
    },
    {
      "entropy": 1.9296542406082153,
      "epoch": 0.17192114752561108,
      "grad_norm": 8.541874885559082,
      "learning_rate": 6.1014343721315835e-06,
      "loss": 0.517,
      "mean_token_accuracy": 0.8412581771612168,
      "num_tokens": 66680863.0,
      "step": 55460
    },
    {
      "entropy": 1.9104670375585555,
      "epoch": 0.17195214665066078,
      "grad_norm": 7.40362548828125,
      "learning_rate": 6.1008843613697255e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8393293395638466,
      "num_tokens": 66692889.0,
      "step": 55470
    },
    {
      "entropy": 1.8890238150954246,
      "epoch": 0.17198314577571047,
      "grad_norm": 9.017800331115723,
      "learning_rate": 6.100334499322393e-06,
      "loss": 0.5257,
      "mean_token_accuracy": 0.834419809281826,
      "num_tokens": 66705358.0,
      "step": 55480
    },
    {
      "entropy": 1.8500746071338654,
      "epoch": 0.17201414490076017,
      "grad_norm": 8.356595993041992,
      "learning_rate": 6.099784785922585e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8483300238847733,
      "num_tokens": 66717381.0,
      "step": 55490
    },
    {
      "entropy": 1.9085285305976867,
      "epoch": 0.17204514402580987,
      "grad_norm": 9.017643928527832,
      "learning_rate": 6.0992352211033335e-06,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8379713967442513,
      "num_tokens": 66729410.0,
      "step": 55500
    },
    {
      "entropy": 1.914148934185505,
      "epoch": 0.17207614315085956,
      "grad_norm": 8.142106056213379,
      "learning_rate": 6.098685804797724e-06,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8404997661709785,
      "num_tokens": 66740551.0,
      "step": 55510
    },
    {
      "entropy": 1.9194943860173226,
      "epoch": 0.17210714227590926,
      "grad_norm": 8.375265121459961,
      "learning_rate": 6.098136536938873e-06,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8302345380187035,
      "num_tokens": 66751861.0,
      "step": 55520
    },
    {
      "entropy": 1.8494600921869278,
      "epoch": 0.17213814140095895,
      "grad_norm": 7.672314643859863,
      "learning_rate": 6.097587417459949e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8483859553933144,
      "num_tokens": 66764180.0,
      "step": 55530
    },
    {
      "entropy": 1.876507543027401,
      "epoch": 0.17216914052600865,
      "grad_norm": 8.333982467651367,
      "learning_rate": 6.097038446294156e-06,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8330820754170418,
      "num_tokens": 66776554.0,
      "step": 55540
    },
    {
      "entropy": 1.8608839854598045,
      "epoch": 0.17220013965105835,
      "grad_norm": 8.950138092041016,
      "learning_rate": 6.096489623374742e-06,
      "loss": 0.5448,
      "mean_token_accuracy": 0.8301695078611374,
      "num_tokens": 66789000.0,
      "step": 55550
    },
    {
      "entropy": 1.871629747748375,
      "epoch": 0.17223113877610804,
      "grad_norm": 4.220654487609863,
      "learning_rate": 6.095940948634997e-06,
      "loss": 0.578,
      "mean_token_accuracy": 0.8214164420962333,
      "num_tokens": 66802167.0,
      "step": 55560
    },
    {
      "entropy": 1.8488945305347442,
      "epoch": 0.17226213790115774,
      "grad_norm": 4.729830265045166,
      "learning_rate": 6.095392422008255e-06,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8359545990824699,
      "num_tokens": 66813727.0,
      "step": 55570
    },
    {
      "entropy": 1.8354024216532707,
      "epoch": 0.17229313702620744,
      "grad_norm": 8.174845695495605,
      "learning_rate": 6.094844043427889e-06,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8510124906897545,
      "num_tokens": 66825754.0,
      "step": 55580
    },
    {
      "entropy": 1.8678600177168847,
      "epoch": 0.17232413615125713,
      "grad_norm": 11.34805679321289,
      "learning_rate": 6.094295812827316e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8442059218883514,
      "num_tokens": 66838613.0,
      "step": 55590
    },
    {
      "entropy": 1.8801333606243134,
      "epoch": 0.17235513527630683,
      "grad_norm": 8.798630714416504,
      "learning_rate": 6.0937477301399924e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8383255407214165,
      "num_tokens": 66850523.0,
      "step": 55600
    },
    {
      "entropy": 1.863646823167801,
      "epoch": 0.17238613440135653,
      "grad_norm": 8.633447647094727,
      "learning_rate": 6.093199795299421e-06,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8378429308533668,
      "num_tokens": 66863104.0,
      "step": 55610
    },
    {
      "entropy": 1.9260516792535782,
      "epoch": 0.17241713352640622,
      "grad_norm": 6.970504283905029,
      "learning_rate": 6.092652008239141e-06,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8314559891819954,
      "num_tokens": 66874671.0,
      "step": 55620
    },
    {
      "entropy": 1.9149534597992897,
      "epoch": 0.17244813265145592,
      "grad_norm": 5.021518230438232,
      "learning_rate": 6.0921043688927366e-06,
      "loss": 0.53,
      "mean_token_accuracy": 0.8325030341744423,
      "num_tokens": 66886294.0,
      "step": 55630
    },
    {
      "entropy": 1.9480685591697693,
      "epoch": 0.17247913177650562,
      "grad_norm": 8.716386795043945,
      "learning_rate": 6.091556877193834e-06,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8272656366229058,
      "num_tokens": 66897341.0,
      "step": 55640
    },
    {
      "entropy": 1.8431043431162835,
      "epoch": 0.17251013090155531,
      "grad_norm": 8.646214485168457,
      "learning_rate": 6.091009533076101e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8376592069864273,
      "num_tokens": 66909553.0,
      "step": 55650
    },
    {
      "entropy": 1.9310444086790084,
      "epoch": 0.172541130026605,
      "grad_norm": 7.217070579528809,
      "learning_rate": 6.090462336473245e-06,
      "loss": 0.6076,
      "mean_token_accuracy": 0.8188234835863113,
      "num_tokens": 66920698.0,
      "step": 55660
    },
    {
      "entropy": 1.7853051990270614,
      "epoch": 0.1725721291516547,
      "grad_norm": 11.338160514831543,
      "learning_rate": 6.089915287319018e-06,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8323128372430801,
      "num_tokens": 66933403.0,
      "step": 55670
    },
    {
      "entropy": 1.8359568476676942,
      "epoch": 0.17260312827670438,
      "grad_norm": 4.960665702819824,
      "learning_rate": 6.089368385547212e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.837063156068325,
      "num_tokens": 66945928.0,
      "step": 55680
    },
    {
      "entropy": 1.7752636238932609,
      "epoch": 0.17263412740175407,
      "grad_norm": 9.67094612121582,
      "learning_rate": 6.088821631091659e-06,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8366348952054977,
      "num_tokens": 66959462.0,
      "step": 55690
    },
    {
      "entropy": 1.7888485848903657,
      "epoch": 0.17266512652680377,
      "grad_norm": 9.076850891113281,
      "learning_rate": 6.088275023886237e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8395483061671257,
      "num_tokens": 66972656.0,
      "step": 55700
    },
    {
      "entropy": 1.8987372756004333,
      "epoch": 0.17269612565185347,
      "grad_norm": 7.72656774520874,
      "learning_rate": 6.087728563864862e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8445681780576706,
      "num_tokens": 66983938.0,
      "step": 55710
    },
    {
      "entropy": 1.894448073208332,
      "epoch": 0.17272712477690316,
      "grad_norm": 8.844822883605957,
      "learning_rate": 6.087182250961492e-06,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8229103952646255,
      "num_tokens": 66996286.0,
      "step": 55720
    },
    {
      "entropy": 1.7768873780965806,
      "epoch": 0.17275812390195286,
      "grad_norm": 4.438620567321777,
      "learning_rate": 6.086636085110128e-06,
      "loss": 0.394,
      "mean_token_accuracy": 0.8516925528645516,
      "num_tokens": 67010213.0,
      "step": 55730
    },
    {
      "entropy": 1.8323894530534743,
      "epoch": 0.17278912302700256,
      "grad_norm": 9.754355430603027,
      "learning_rate": 6.08609006624481e-06,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8407779783010483,
      "num_tokens": 67022907.0,
      "step": 55740
    },
    {
      "entropy": 1.8603756889700889,
      "epoch": 0.17282012215205225,
      "grad_norm": 10.162839889526367,
      "learning_rate": 6.085544194299622e-06,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8378032267093658,
      "num_tokens": 67035045.0,
      "step": 55750
    },
    {
      "entropy": 1.7585888996720314,
      "epoch": 0.17285112127710195,
      "grad_norm": 9.135080337524414,
      "learning_rate": 6.084998469208687e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8512668639421463,
      "num_tokens": 67048645.0,
      "step": 55760
    },
    {
      "entropy": 1.8613043814897536,
      "epoch": 0.17288212040215165,
      "grad_norm": 9.348379135131836,
      "learning_rate": 6.084452890906173e-06,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8385450929403305,
      "num_tokens": 67060973.0,
      "step": 55770
    },
    {
      "entropy": 1.9134966135025024,
      "epoch": 0.17291311952720134,
      "grad_norm": 9.092744827270508,
      "learning_rate": 6.083907459326285e-06,
      "loss": 0.5414,
      "mean_token_accuracy": 0.838169914484024,
      "num_tokens": 67072327.0,
      "step": 55780
    },
    {
      "entropy": 1.8521252527832985,
      "epoch": 0.17294411865225104,
      "grad_norm": 8.718194007873535,
      "learning_rate": 6.08336217440327e-06,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8312934651970864,
      "num_tokens": 67084319.0,
      "step": 55790
    },
    {
      "entropy": 1.8963706970214844,
      "epoch": 0.17297511777730074,
      "grad_norm": 8.831055641174316,
      "learning_rate": 6.08281703607142e-06,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8281921342015266,
      "num_tokens": 67096177.0,
      "step": 55800
    },
    {
      "entropy": 1.8676214978098868,
      "epoch": 0.17300611690235043,
      "grad_norm": 9.61502742767334,
      "learning_rate": 6.082272044265064e-06,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8491611614823341,
      "num_tokens": 67108106.0,
      "step": 55810
    },
    {
      "entropy": 1.8385474801063537,
      "epoch": 0.17303711602740013,
      "grad_norm": 3.662489414215088,
      "learning_rate": 6.0817271989185745e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8533884555101394,
      "num_tokens": 67120164.0,
      "step": 55820
    },
    {
      "entropy": 1.7729324698448181,
      "epoch": 0.17306811515244983,
      "grad_norm": 4.637348175048828,
      "learning_rate": 6.081182499966365e-06,
      "loss": 0.4626,
      "mean_token_accuracy": 0.8501248374581337,
      "num_tokens": 67134497.0,
      "step": 55830
    },
    {
      "entropy": 1.9585619807243346,
      "epoch": 0.17309911427749952,
      "grad_norm": 7.857696056365967,
      "learning_rate": 6.080637947342887e-06,
      "loss": 0.5646,
      "mean_token_accuracy": 0.8382192403078079,
      "num_tokens": 67145397.0,
      "step": 55840
    },
    {
      "entropy": 1.809555734694004,
      "epoch": 0.17313011340254922,
      "grad_norm": 5.004009246826172,
      "learning_rate": 6.080093540982638e-06,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8505311787128449,
      "num_tokens": 67157963.0,
      "step": 55850
    },
    {
      "entropy": 1.9263376638293266,
      "epoch": 0.17316111252759891,
      "grad_norm": 7.850517272949219,
      "learning_rate": 6.079549280820153e-06,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8317690268158913,
      "num_tokens": 67169537.0,
      "step": 55860
    },
    {
      "entropy": 1.9509922355413436,
      "epoch": 0.1731921116526486,
      "grad_norm": 9.430130958557129,
      "learning_rate": 6.079005166790011e-06,
      "loss": 0.56,
      "mean_token_accuracy": 0.8258754447102546,
      "num_tokens": 67180681.0,
      "step": 55870
    },
    {
      "entropy": 1.9066618397831916,
      "epoch": 0.1732231107776983,
      "grad_norm": 8.74619197845459,
      "learning_rate": 6.078461198826828e-06,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8300882831215859,
      "num_tokens": 67191836.0,
      "step": 55880
    },
    {
      "entropy": 1.7900028765201568,
      "epoch": 0.173254109902748,
      "grad_norm": 6.856855869293213,
      "learning_rate": 6.077917376865262e-06,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8394763827323913,
      "num_tokens": 67206034.0,
      "step": 55890
    },
    {
      "entropy": 1.9227029278874397,
      "epoch": 0.1732851090277977,
      "grad_norm": 8.958833694458008,
      "learning_rate": 6.077373700840018e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8387436643242836,
      "num_tokens": 67217234.0,
      "step": 55900
    },
    {
      "entropy": 1.9202365294098853,
      "epoch": 0.1733161081528474,
      "grad_norm": 4.27255392074585,
      "learning_rate": 6.076830170685832e-06,
      "loss": 0.5532,
      "mean_token_accuracy": 0.827874468266964,
      "num_tokens": 67228530.0,
      "step": 55910
    },
    {
      "entropy": 1.908749097585678,
      "epoch": 0.1733471072778971,
      "grad_norm": 8.770615577697754,
      "learning_rate": 6.07628678633749e-06,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8317655339837075,
      "num_tokens": 67240085.0,
      "step": 55920
    },
    {
      "entropy": 1.8192409992218017,
      "epoch": 0.17337810640294676,
      "grad_norm": 8.269194602966309,
      "learning_rate": 6.0757435477298085e-06,
      "loss": 0.501,
      "mean_token_accuracy": 0.8376537606120109,
      "num_tokens": 67252721.0,
      "step": 55930
    },
    {
      "entropy": 1.835218572616577,
      "epoch": 0.17340910552799646,
      "grad_norm": 8.810011863708496,
      "learning_rate": 6.075200454797657e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8403037443757058,
      "num_tokens": 67265339.0,
      "step": 55940
    },
    {
      "entropy": 1.9089889451861382,
      "epoch": 0.17344010465304616,
      "grad_norm": 5.602406978607178,
      "learning_rate": 6.0746575074759365e-06,
      "loss": 0.5341,
      "mean_token_accuracy": 0.8314460068941116,
      "num_tokens": 67278228.0,
      "step": 55950
    },
    {
      "entropy": 1.8503844618797303,
      "epoch": 0.17347110377809585,
      "grad_norm": 4.604806423187256,
      "learning_rate": 6.074114705699592e-06,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8360475450754166,
      "num_tokens": 67290208.0,
      "step": 55960
    },
    {
      "entropy": 1.8351913020014763,
      "epoch": 0.17350210290314555,
      "grad_norm": 10.080077171325684,
      "learning_rate": 6.073572049403609e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8317245200276375,
      "num_tokens": 67303011.0,
      "step": 55970
    },
    {
      "entropy": 1.7987764164805413,
      "epoch": 0.17353310202819525,
      "grad_norm": 8.698798179626465,
      "learning_rate": 6.073029538523015e-06,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8366831094026566,
      "num_tokens": 67316802.0,
      "step": 55980
    },
    {
      "entropy": 1.8339210391044616,
      "epoch": 0.17356410115324494,
      "grad_norm": 3.881253957748413,
      "learning_rate": 6.072487172992875e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8414662271738053,
      "num_tokens": 67329858.0,
      "step": 55990
    },
    {
      "entropy": 1.8059268981218337,
      "epoch": 0.17359510027829464,
      "grad_norm": 4.120025157928467,
      "learning_rate": 6.0719449527482976e-06,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8513453081250191,
      "num_tokens": 67342567.0,
      "step": 56000
    },
    {
      "entropy": 1.9046938508749007,
      "epoch": 0.17362609940334434,
      "grad_norm": 4.1900787353515625,
      "learning_rate": 6.07140287772443e-06,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8171088561415673,
      "num_tokens": 67354240.0,
      "step": 56010
    },
    {
      "entropy": 1.8154946908354759,
      "epoch": 0.17365709852839403,
      "grad_norm": 7.720127582550049,
      "learning_rate": 6.070860947856461e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.841141340136528,
      "num_tokens": 67366512.0,
      "step": 56020
    },
    {
      "entropy": 1.7826613813638688,
      "epoch": 0.17368809765344373,
      "grad_norm": 8.429262161254883,
      "learning_rate": 6.07031916307962e-06,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8355521574616432,
      "num_tokens": 67379705.0,
      "step": 56030
    },
    {
      "entropy": 1.8282306790351868,
      "epoch": 0.17371909677849343,
      "grad_norm": 8.058755874633789,
      "learning_rate": 6.0697775233291746e-06,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8397795766592026,
      "num_tokens": 67392350.0,
      "step": 56040
    },
    {
      "entropy": 1.7306653708219528,
      "epoch": 0.17375009590354312,
      "grad_norm": 8.275370597839355,
      "learning_rate": 6.069236028540436e-06,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8475747913122177,
      "num_tokens": 67406980.0,
      "step": 56050
    },
    {
      "entropy": 1.8713230773806573,
      "epoch": 0.17378109502859282,
      "grad_norm": 8.6502103805542,
      "learning_rate": 6.068694678648755e-06,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8327876642346382,
      "num_tokens": 67419032.0,
      "step": 56060
    },
    {
      "entropy": 1.9657189399003983,
      "epoch": 0.17381209415364252,
      "grad_norm": 8.886541366577148,
      "learning_rate": 6.068153473589519e-06,
      "loss": 0.5763,
      "mean_token_accuracy": 0.8254697665572166,
      "num_tokens": 67429412.0,
      "step": 56070
    },
    {
      "entropy": 1.9012456104159354,
      "epoch": 0.1738430932786922,
      "grad_norm": 7.720726013183594,
      "learning_rate": 6.0676124132981626e-06,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8330578565597534,
      "num_tokens": 67441415.0,
      "step": 56080
    },
    {
      "entropy": 1.8742031693458556,
      "epoch": 0.1738740924037419,
      "grad_norm": 3.867934465408325,
      "learning_rate": 6.067071497710155e-06,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8421854302287102,
      "num_tokens": 67453389.0,
      "step": 56090
    },
    {
      "entropy": 1.9605601608753205,
      "epoch": 0.1739050915287916,
      "grad_norm": 8.339366912841797,
      "learning_rate": 6.066530726761009e-06,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8267245456576348,
      "num_tokens": 67464579.0,
      "step": 56100
    },
    {
      "entropy": 1.9092912346124649,
      "epoch": 0.1739360906538413,
      "grad_norm": 8.375153541564941,
      "learning_rate": 6.065990100386274e-06,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8393413156270981,
      "num_tokens": 67476222.0,
      "step": 56110
    },
    {
      "entropy": 1.805423805117607,
      "epoch": 0.173967089778891,
      "grad_norm": 7.936673641204834,
      "learning_rate": 6.065449618521544e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8524986699223518,
      "num_tokens": 67488739.0,
      "step": 56120
    },
    {
      "entropy": 1.8608886793255806,
      "epoch": 0.1739980889039407,
      "grad_norm": 9.917237281799316,
      "learning_rate": 6.0649092811024514e-06,
      "loss": 0.51,
      "mean_token_accuracy": 0.8281596288084984,
      "num_tokens": 67500883.0,
      "step": 56130
    },
    {
      "entropy": 1.9484786361455917,
      "epoch": 0.1740290880289904,
      "grad_norm": 11.2413969039917,
      "learning_rate": 6.064369088064665e-06,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8397915035486221,
      "num_tokens": 67512036.0,
      "step": 56140
    },
    {
      "entropy": 1.896202652156353,
      "epoch": 0.1740600871540401,
      "grad_norm": 10.009648323059082,
      "learning_rate": 6.063829039343899e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8400693356990814,
      "num_tokens": 67523562.0,
      "step": 56150
    },
    {
      "entropy": 1.8820205710828304,
      "epoch": 0.17409108627908979,
      "grad_norm": 4.450737953186035,
      "learning_rate": 6.063289134875907e-06,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8278883457183838,
      "num_tokens": 67536330.0,
      "step": 56160
    },
    {
      "entropy": 1.8365131139755249,
      "epoch": 0.17412208540413948,
      "grad_norm": 4.729415416717529,
      "learning_rate": 6.062749374596479e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8442412883043289,
      "num_tokens": 67549775.0,
      "step": 56170
    },
    {
      "entropy": 1.9157144859433175,
      "epoch": 0.17415308452918915,
      "grad_norm": 8.306681632995605,
      "learning_rate": 6.062209758441451e-06,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8310726657509804,
      "num_tokens": 67561670.0,
      "step": 56180
    },
    {
      "entropy": 1.8954879730939864,
      "epoch": 0.17418408365423885,
      "grad_norm": 9.215730667114258,
      "learning_rate": 6.0616702863466905e-06,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8425735384225845,
      "num_tokens": 67573051.0,
      "step": 56190
    },
    {
      "entropy": 1.9187992975115775,
      "epoch": 0.17421508277928854,
      "grad_norm": 7.861081600189209,
      "learning_rate": 6.061130958248112e-06,
      "loss": 0.601,
      "mean_token_accuracy": 0.8312497869133949,
      "num_tokens": 67585205.0,
      "step": 56200
    },
    {
      "entropy": 1.954244077205658,
      "epoch": 0.17424608190433824,
      "grad_norm": 9.027844429016113,
      "learning_rate": 6.060591774081669e-06,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8323284685611725,
      "num_tokens": 67596130.0,
      "step": 56210
    },
    {
      "entropy": 1.917762076854706,
      "epoch": 0.17427708102938794,
      "grad_norm": 8.658714294433594,
      "learning_rate": 6.060052733783352e-06,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8378539264202118,
      "num_tokens": 67607764.0,
      "step": 56220
    },
    {
      "entropy": 1.9130154103040695,
      "epoch": 0.17430808015443763,
      "grad_norm": 10.506196022033691,
      "learning_rate": 6.0595138372891934e-06,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8319619536399842,
      "num_tokens": 67619152.0,
      "step": 56230
    },
    {
      "entropy": 1.8991611540317535,
      "epoch": 0.17433907927948733,
      "grad_norm": 7.824307918548584,
      "learning_rate": 6.0589750845352644e-06,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8250974491238594,
      "num_tokens": 67630997.0,
      "step": 56240
    },
    {
      "entropy": 1.8883090645074845,
      "epoch": 0.17437007840453703,
      "grad_norm": 8.599298477172852,
      "learning_rate": 6.058436475457677e-06,
      "loss": 0.5425,
      "mean_token_accuracy": 0.831093080341816,
      "num_tokens": 67642786.0,
      "step": 56250
    },
    {
      "entropy": 1.8896732196211814,
      "epoch": 0.17440107752958672,
      "grad_norm": 4.400026798248291,
      "learning_rate": 6.057898009992582e-06,
      "loss": 0.53,
      "mean_token_accuracy": 0.8287911862134933,
      "num_tokens": 67654432.0,
      "step": 56260
    },
    {
      "entropy": 1.941761639714241,
      "epoch": 0.17443207665463642,
      "grad_norm": 7.403299808502197,
      "learning_rate": 6.057359688076171e-06,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8296359524130821,
      "num_tokens": 67665128.0,
      "step": 56270
    },
    {
      "entropy": 1.8876193895936013,
      "epoch": 0.17446307577968612,
      "grad_norm": 8.507538795471191,
      "learning_rate": 6.0568215096446736e-06,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8257350817322731,
      "num_tokens": 67676670.0,
      "step": 56280
    },
    {
      "entropy": 1.9427810519933701,
      "epoch": 0.1744940749047358,
      "grad_norm": 8.400871276855469,
      "learning_rate": 6.0562834746343615e-06,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8306556642055511,
      "num_tokens": 67688070.0,
      "step": 56290
    },
    {
      "entropy": 1.8497335493564606,
      "epoch": 0.1745250740297855,
      "grad_norm": 8.613706588745117,
      "learning_rate": 6.0557455829815425e-06,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8401633113622665,
      "num_tokens": 67699880.0,
      "step": 56300
    },
    {
      "entropy": 1.8545069240033627,
      "epoch": 0.1745560731548352,
      "grad_norm": 8.968703269958496,
      "learning_rate": 6.055207834622569e-06,
      "loss": 0.4838,
      "mean_token_accuracy": 0.83926263153553,
      "num_tokens": 67712700.0,
      "step": 56310
    },
    {
      "entropy": 1.878834429383278,
      "epoch": 0.1745870722798849,
      "grad_norm": 3.72434663772583,
      "learning_rate": 6.054670229493826e-06,
      "loss": 0.5328,
      "mean_token_accuracy": 0.837174066901207,
      "num_tokens": 67724809.0,
      "step": 56320
    },
    {
      "entropy": 1.891117848455906,
      "epoch": 0.1746180714049346,
      "grad_norm": 7.3367180824279785,
      "learning_rate": 6.054132767531746e-06,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8528507500886917,
      "num_tokens": 67735990.0,
      "step": 56330
    },
    {
      "entropy": 1.818806654214859,
      "epoch": 0.1746490705299843,
      "grad_norm": 4.021301746368408,
      "learning_rate": 6.053595448672795e-06,
      "loss": 0.4479,
      "mean_token_accuracy": 0.848919802904129,
      "num_tokens": 67748381.0,
      "step": 56340
    },
    {
      "entropy": 1.9311149969697,
      "epoch": 0.174680069655034,
      "grad_norm": 9.249166488647461,
      "learning_rate": 6.053058272853482e-06,
      "loss": 0.5285,
      "mean_token_accuracy": 0.8428058341145516,
      "num_tokens": 67759883.0,
      "step": 56350
    },
    {
      "entropy": 1.8628065332770347,
      "epoch": 0.1747110687800837,
      "grad_norm": 5.258411884307861,
      "learning_rate": 6.0525212400103525e-06,
      "loss": 0.6164,
      "mean_token_accuracy": 0.8357118725776672,
      "num_tokens": 67772424.0,
      "step": 56360
    },
    {
      "entropy": 1.892784410715103,
      "epoch": 0.17474206790513339,
      "grad_norm": 3.5523910522460938,
      "learning_rate": 6.051984350079994e-06,
      "loss": 0.5414,
      "mean_token_accuracy": 0.83593248128891,
      "num_tokens": 67783804.0,
      "step": 56370
    },
    {
      "entropy": 1.8566182047128676,
      "epoch": 0.17477306703018308,
      "grad_norm": 10.872590065002441,
      "learning_rate": 6.051447602999031e-06,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8340441033244133,
      "num_tokens": 67795593.0,
      "step": 56380
    },
    {
      "entropy": 1.8216275811195373,
      "epoch": 0.17480406615523278,
      "grad_norm": 8.305917739868164,
      "learning_rate": 6.050910998704129e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8441584646701813,
      "num_tokens": 67807848.0,
      "step": 56390
    },
    {
      "entropy": 1.8245621785521506,
      "epoch": 0.17483506528028248,
      "grad_norm": 4.366250514984131,
      "learning_rate": 6.050374537131993e-06,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8390408426523208,
      "num_tokens": 67820873.0,
      "step": 56400
    },
    {
      "entropy": 1.8166996166110039,
      "epoch": 0.17486606440533217,
      "grad_norm": 8.542943954467773,
      "learning_rate": 6.049838218219366e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8468366071581841,
      "num_tokens": 67833571.0,
      "step": 56410
    },
    {
      "entropy": 1.858149343729019,
      "epoch": 0.17489706353038187,
      "grad_norm": 8.90046501159668,
      "learning_rate": 6.049302041903031e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8406156003475189,
      "num_tokens": 67845941.0,
      "step": 56420
    },
    {
      "entropy": 1.965251961350441,
      "epoch": 0.17492806265543154,
      "grad_norm": 10.929553985595703,
      "learning_rate": 6.048766008119811e-06,
      "loss": 0.5884,
      "mean_token_accuracy": 0.8183304697275162,
      "num_tokens": 67856898.0,
      "step": 56430
    },
    {
      "entropy": 1.9665222853422164,
      "epoch": 0.17495906178048123,
      "grad_norm": 9.303450584411621,
      "learning_rate": 6.048230116806566e-06,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8342076927423477,
      "num_tokens": 67868059.0,
      "step": 56440
    },
    {
      "entropy": 1.916182427108288,
      "epoch": 0.17499006090553093,
      "grad_norm": 7.816822052001953,
      "learning_rate": 6.047694367900196e-06,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8353784546256066,
      "num_tokens": 67880386.0,
      "step": 56450
    },
    {
      "entropy": 1.9688440665602684,
      "epoch": 0.17502106003058063,
      "grad_norm": 8.755724906921387,
      "learning_rate": 6.047158761337643e-06,
      "loss": 0.5154,
      "mean_token_accuracy": 0.8388349682092666,
      "num_tokens": 67892083.0,
      "step": 56460
    },
    {
      "entropy": 1.8669242531061172,
      "epoch": 0.17505205915563032,
      "grad_norm": 7.993595600128174,
      "learning_rate": 6.046623297055885e-06,
      "loss": 0.5223,
      "mean_token_accuracy": 0.8375104799866676,
      "num_tokens": 67904144.0,
      "step": 56470
    },
    {
      "entropy": 1.816992525756359,
      "epoch": 0.17508305828068002,
      "grad_norm": 4.964598655700684,
      "learning_rate": 6.046087974991937e-06,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8406194254755974,
      "num_tokens": 67917537.0,
      "step": 56480
    },
    {
      "entropy": 1.8135570034384727,
      "epoch": 0.17511405740572972,
      "grad_norm": 8.609065055847168,
      "learning_rate": 6.045552795082859e-06,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8388777539134026,
      "num_tokens": 67931322.0,
      "step": 56490
    },
    {
      "entropy": 1.9724404200911523,
      "epoch": 0.1751450565307794,
      "grad_norm": 9.910813331604004,
      "learning_rate": 6.0450177572657435e-06,
      "loss": 0.5724,
      "mean_token_accuracy": 0.829445195198059,
      "num_tokens": 67942971.0,
      "step": 56500
    },
    {
      "entropy": 1.8769863620400429,
      "epoch": 0.1751760556558291,
      "grad_norm": 10.879551887512207,
      "learning_rate": 6.044482861477728e-06,
      "loss": 0.4882,
      "mean_token_accuracy": 0.84668777436018,
      "num_tokens": 67954835.0,
      "step": 56510
    },
    {
      "entropy": 1.895396353304386,
      "epoch": 0.1752070547808788,
      "grad_norm": 7.5075178146362305,
      "learning_rate": 6.043948107655985e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.859335508942604,
      "num_tokens": 67966585.0,
      "step": 56520
    },
    {
      "entropy": 1.845558349788189,
      "epoch": 0.1752380539059285,
      "grad_norm": 5.098723411560059,
      "learning_rate": 6.0434134957377275e-06,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8512143403291702,
      "num_tokens": 67978885.0,
      "step": 56530
    },
    {
      "entropy": 1.8857697710394858,
      "epoch": 0.1752690530309782,
      "grad_norm": 9.507451057434082,
      "learning_rate": 6.042879025660207e-06,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8429853811860084,
      "num_tokens": 67990999.0,
      "step": 56540
    },
    {
      "entropy": 1.8623712986707688,
      "epoch": 0.1753000521560279,
      "grad_norm": 9.963186264038086,
      "learning_rate": 6.042344697360713e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8374731317162514,
      "num_tokens": 68003103.0,
      "step": 56550
    },
    {
      "entropy": 1.8676358297467233,
      "epoch": 0.1753310512810776,
      "grad_norm": 7.353335857391357,
      "learning_rate": 6.041810510776573e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8497109815478325,
      "num_tokens": 68015029.0,
      "step": 56560
    },
    {
      "entropy": 1.9030646055936813,
      "epoch": 0.1753620504061273,
      "grad_norm": 4.22896671295166,
      "learning_rate": 6.041276465845158e-06,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8298840820789337,
      "num_tokens": 68026783.0,
      "step": 56570
    },
    {
      "entropy": 1.9653840601444243,
      "epoch": 0.175393049531177,
      "grad_norm": 7.451119899749756,
      "learning_rate": 6.040742562503874e-06,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8305002480745316,
      "num_tokens": 68037670.0,
      "step": 56580
    },
    {
      "entropy": 1.9089856892824173,
      "epoch": 0.17542404865622668,
      "grad_norm": 9.136287689208984,
      "learning_rate": 6.040208800690164e-06,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8452465951442718,
      "num_tokens": 68049712.0,
      "step": 56590
    },
    {
      "entropy": 1.9428860023617744,
      "epoch": 0.17545504778127638,
      "grad_norm": 4.60407829284668,
      "learning_rate": 6.039675180341514e-06,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8250961035490036,
      "num_tokens": 68061772.0,
      "step": 56600
    },
    {
      "entropy": 1.8673185929656029,
      "epoch": 0.17548604690632608,
      "grad_norm": 8.91592788696289,
      "learning_rate": 6.039141701395445e-06,
      "loss": 0.5159,
      "mean_token_accuracy": 0.831261222064495,
      "num_tokens": 68074176.0,
      "step": 56610
    },
    {
      "entropy": 1.8924739733338356,
      "epoch": 0.17551704603137577,
      "grad_norm": 4.301705360412598,
      "learning_rate": 6.0386083637895194e-06,
      "loss": 0.589,
      "mean_token_accuracy": 0.8193011194467544,
      "num_tokens": 68087007.0,
      "step": 56620
    },
    {
      "entropy": 1.9513219073414803,
      "epoch": 0.17554804515642547,
      "grad_norm": 8.441484451293945,
      "learning_rate": 6.038075167461339e-06,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8327566117048264,
      "num_tokens": 68098862.0,
      "step": 56630
    },
    {
      "entropy": 1.927374078333378,
      "epoch": 0.17557904428147517,
      "grad_norm": 9.029617309570312,
      "learning_rate": 6.037542112348537e-06,
      "loss": 0.5306,
      "mean_token_accuracy": 0.831793662905693,
      "num_tokens": 68111512.0,
      "step": 56640
    },
    {
      "entropy": 1.9223608300089836,
      "epoch": 0.17561004340652486,
      "grad_norm": 4.942859172821045,
      "learning_rate": 6.0370091983887946e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8408562824130058,
      "num_tokens": 68123365.0,
      "step": 56650
    },
    {
      "entropy": 1.9207605987787246,
      "epoch": 0.17564104253157456,
      "grad_norm": 8.518644332885742,
      "learning_rate": 6.036476425519826e-06,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8278803005814552,
      "num_tokens": 68135831.0,
      "step": 56660
    },
    {
      "entropy": 1.8496787279844285,
      "epoch": 0.17567204165662423,
      "grad_norm": 8.98222541809082,
      "learning_rate": 6.0359437936793865e-06,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8472736284136773,
      "num_tokens": 68147503.0,
      "step": 56670
    },
    {
      "entropy": 1.9197917222976684,
      "epoch": 0.17570304078167392,
      "grad_norm": 8.030874252319336,
      "learning_rate": 6.0354113028052645e-06,
      "loss": 0.5854,
      "mean_token_accuracy": 0.8222957968711853,
      "num_tokens": 68159296.0,
      "step": 56680
    },
    {
      "entropy": 1.7842533797025681,
      "epoch": 0.17573403990672362,
      "grad_norm": 4.326035976409912,
      "learning_rate": 6.034878952835293e-06,
      "loss": 0.452,
      "mean_token_accuracy": 0.8420939773321152,
      "num_tokens": 68173065.0,
      "step": 56690
    },
    {
      "entropy": 1.832868304848671,
      "epoch": 0.17576503903177332,
      "grad_norm": 7.654922962188721,
      "learning_rate": 6.0343467437073435e-06,
      "loss": 0.4637,
      "mean_token_accuracy": 0.8458761259913444,
      "num_tokens": 68185736.0,
      "step": 56700
    },
    {
      "entropy": 1.8796532317996024,
      "epoch": 0.17579603815682301,
      "grad_norm": 8.131157875061035,
      "learning_rate": 6.03381467535932e-06,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8416437700390815,
      "num_tokens": 68197051.0,
      "step": 56710
    },
    {
      "entropy": 1.8534429788589477,
      "epoch": 0.1758270372818727,
      "grad_norm": 10.127676963806152,
      "learning_rate": 6.03328274772917e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8437806442379951,
      "num_tokens": 68209732.0,
      "step": 56720
    },
    {
      "entropy": 1.9118575572967529,
      "epoch": 0.1758580364069224,
      "grad_norm": 8.467833518981934,
      "learning_rate": 6.0327509607548775e-06,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8397658735513687,
      "num_tokens": 68220339.0,
      "step": 56730
    },
    {
      "entropy": 1.844566436111927,
      "epoch": 0.1758890355319721,
      "grad_norm": 9.010001182556152,
      "learning_rate": 6.032219314374463e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8425126999616623,
      "num_tokens": 68232942.0,
      "step": 56740
    },
    {
      "entropy": 1.874419206380844,
      "epoch": 0.1759200346570218,
      "grad_norm": 3.708150625228882,
      "learning_rate": 6.03168780852599e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8311620220541954,
      "num_tokens": 68245087.0,
      "step": 56750
    },
    {
      "entropy": 1.936698080599308,
      "epoch": 0.1759510337820715,
      "grad_norm": 8.663212776184082,
      "learning_rate": 6.0311564431475544e-06,
      "loss": 0.5413,
      "mean_token_accuracy": 0.8401820287108421,
      "num_tokens": 68256104.0,
      "step": 56760
    },
    {
      "entropy": 1.7919904850423336,
      "epoch": 0.1759820329071212,
      "grad_norm": 8.531180381774902,
      "learning_rate": 6.030625218177295e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8482810765504837,
      "num_tokens": 68269355.0,
      "step": 56770
    },
    {
      "entropy": 1.8165376424789428,
      "epoch": 0.1760130320321709,
      "grad_norm": 8.21648120880127,
      "learning_rate": 6.030094133553386e-06,
      "loss": 0.3991,
      "mean_token_accuracy": 0.8565915554761887,
      "num_tokens": 68282068.0,
      "step": 56780
    },
    {
      "entropy": 1.9260249942541123,
      "epoch": 0.1760440311572206,
      "grad_norm": 10.875053405761719,
      "learning_rate": 6.029563189214042e-06,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8371611222624779,
      "num_tokens": 68293043.0,
      "step": 56790
    },
    {
      "entropy": 1.894984395802021,
      "epoch": 0.17607503028227028,
      "grad_norm": 4.284290790557861,
      "learning_rate": 6.0290323850975115e-06,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8329278856515885,
      "num_tokens": 68304689.0,
      "step": 56800
    },
    {
      "entropy": 1.826059702038765,
      "epoch": 0.17610602940731998,
      "grad_norm": 10.755820274353027,
      "learning_rate": 6.028501721142086e-06,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8308580055832863,
      "num_tokens": 68317294.0,
      "step": 56810
    },
    {
      "entropy": 1.8452559620141984,
      "epoch": 0.17613702853236968,
      "grad_norm": 7.7797040939331055,
      "learning_rate": 6.027971197286092e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8381189867854119,
      "num_tokens": 68329628.0,
      "step": 56820
    },
    {
      "entropy": 1.96290685236454,
      "epoch": 0.17616802765741937,
      "grad_norm": 9.668147087097168,
      "learning_rate": 6.027440813467895e-06,
      "loss": 0.5797,
      "mean_token_accuracy": 0.822065281867981,
      "num_tokens": 68341030.0,
      "step": 56830
    },
    {
      "entropy": 1.869129341840744,
      "epoch": 0.17619902678246907,
      "grad_norm": 8.411051750183105,
      "learning_rate": 6.026910569625899e-06,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8431349366903305,
      "num_tokens": 68352755.0,
      "step": 56840
    },
    {
      "entropy": 1.8781275868415832,
      "epoch": 0.17623002590751877,
      "grad_norm": 8.966902732849121,
      "learning_rate": 6.026380465698544e-06,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8468410953879356,
      "num_tokens": 68364276.0,
      "step": 56850
    },
    {
      "entropy": 1.7951830908656121,
      "epoch": 0.17626102503256846,
      "grad_norm": 7.949598789215088,
      "learning_rate": 6.025850501624308e-06,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8530111536383629,
      "num_tokens": 68377483.0,
      "step": 56860
    },
    {
      "entropy": 1.8381464675068855,
      "epoch": 0.17629202415761816,
      "grad_norm": 8.990347862243652,
      "learning_rate": 6.025320677341711e-06,
      "loss": 0.477,
      "mean_token_accuracy": 0.8483751729130745,
      "num_tokens": 68390308.0,
      "step": 56870
    },
    {
      "entropy": 1.8240345925092698,
      "epoch": 0.17632302328266786,
      "grad_norm": 7.0853071212768555,
      "learning_rate": 6.024790992789304e-06,
      "loss": 0.4229,
      "mean_token_accuracy": 0.8524978414177895,
      "num_tokens": 68402956.0,
      "step": 56880
    },
    {
      "entropy": 1.8852536737918855,
      "epoch": 0.17635402240771755,
      "grad_norm": 7.855372905731201,
      "learning_rate": 6.024261447905683e-06,
      "loss": 0.529,
      "mean_token_accuracy": 0.830386458337307,
      "num_tokens": 68415648.0,
      "step": 56890
    },
    {
      "entropy": 1.903229682147503,
      "epoch": 0.17638502153276725,
      "grad_norm": 9.828873634338379,
      "learning_rate": 6.0237320426294755e-06,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8335711434483528,
      "num_tokens": 68427530.0,
      "step": 56900
    },
    {
      "entropy": 1.9013172283768653,
      "epoch": 0.17641602065781695,
      "grad_norm": 8.85859203338623,
      "learning_rate": 6.023202776899353e-06,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8388579234480857,
      "num_tokens": 68439199.0,
      "step": 56910
    },
    {
      "entropy": 1.7970546633005142,
      "epoch": 0.17644701978286662,
      "grad_norm": 7.830367088317871,
      "learning_rate": 6.0226736506540186e-06,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8480477169156074,
      "num_tokens": 68452889.0,
      "step": 56920
    },
    {
      "entropy": 1.9641900137066841,
      "epoch": 0.1764780189079163,
      "grad_norm": 9.607945442199707,
      "learning_rate": 6.022144663832216e-06,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8301128730177879,
      "num_tokens": 68464241.0,
      "step": 56930
    },
    {
      "entropy": 1.8105278745293618,
      "epoch": 0.176509018032966,
      "grad_norm": 8.978128433227539,
      "learning_rate": 6.0216158163727265e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8478241801261902,
      "num_tokens": 68477760.0,
      "step": 56940
    },
    {
      "entropy": 1.8664100661873817,
      "epoch": 0.1765400171580157,
      "grad_norm": 9.166288375854492,
      "learning_rate": 6.021087108214369e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8363597363233566,
      "num_tokens": 68489199.0,
      "step": 56950
    },
    {
      "entropy": 1.86256393045187,
      "epoch": 0.1765710162830654,
      "grad_norm": 11.094559669494629,
      "learning_rate": 6.020558539296e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8493609294295311,
      "num_tokens": 68500857.0,
      "step": 56960
    },
    {
      "entropy": 1.9616552650928498,
      "epoch": 0.1766020154081151,
      "grad_norm": 9.776663780212402,
      "learning_rate": 6.020030109556513e-06,
      "loss": 0.5729,
      "mean_token_accuracy": 0.8265084490180016,
      "num_tokens": 68512216.0,
      "step": 56970
    },
    {
      "entropy": 1.9011280700564384,
      "epoch": 0.1766330145331648,
      "grad_norm": 10.027840614318848,
      "learning_rate": 6.019501818934841e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.83545922935009,
      "num_tokens": 68524348.0,
      "step": 56980
    },
    {
      "entropy": 1.898293286561966,
      "epoch": 0.1766640136582145,
      "grad_norm": 8.679490089416504,
      "learning_rate": 6.018973667369951e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.837830375134945,
      "num_tokens": 68536357.0,
      "step": 56990
    },
    {
      "entropy": 1.8468020305037498,
      "epoch": 0.1766950127832642,
      "grad_norm": 10.170060157775879,
      "learning_rate": 6.01844565480085e-06,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8466181620955467,
      "num_tokens": 68548639.0,
      "step": 57000
    },
    {
      "entropy": 1.8502362087368964,
      "epoch": 0.17672601190831388,
      "grad_norm": 9.918789863586426,
      "learning_rate": 6.017917781166582e-06,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8434646666049957,
      "num_tokens": 68560221.0,
      "step": 57010
    },
    {
      "entropy": 1.847883252799511,
      "epoch": 0.17675701103336358,
      "grad_norm": 3.5449788570404053,
      "learning_rate": 6.017390046406228e-06,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8291626140475273,
      "num_tokens": 68573365.0,
      "step": 57020
    },
    {
      "entropy": 1.8949782311916352,
      "epoch": 0.17678801015841328,
      "grad_norm": 8.271769523620605,
      "learning_rate": 6.016862450458908e-06,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8308848381042481,
      "num_tokens": 68585450.0,
      "step": 57030
    },
    {
      "entropy": 1.8946194440126418,
      "epoch": 0.17681900928346297,
      "grad_norm": 4.019629955291748,
      "learning_rate": 6.016334993263777e-06,
      "loss": 0.5717,
      "mean_token_accuracy": 0.832112868130207,
      "num_tokens": 68597707.0,
      "step": 57040
    },
    {
      "entropy": 1.8897631257772445,
      "epoch": 0.17685000840851267,
      "grad_norm": 7.272465705871582,
      "learning_rate": 6.015807674760029e-06,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8315581545233727,
      "num_tokens": 68610457.0,
      "step": 57050
    },
    {
      "entropy": 1.925081330537796,
      "epoch": 0.17688100753356237,
      "grad_norm": 3.892751932144165,
      "learning_rate": 6.015280494886894e-06,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8374591827392578,
      "num_tokens": 68621838.0,
      "step": 57060
    },
    {
      "entropy": 1.9003553301095963,
      "epoch": 0.17691200665861206,
      "grad_norm": 8.307111740112305,
      "learning_rate": 6.01475345358364e-06,
      "loss": 0.549,
      "mean_token_accuracy": 0.8233644008636475,
      "num_tokens": 68633799.0,
      "step": 57070
    },
    {
      "entropy": 1.8528232917189598,
      "epoch": 0.17694300578366176,
      "grad_norm": 5.3729071617126465,
      "learning_rate": 6.014226550789571e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8408889621496201,
      "num_tokens": 68645858.0,
      "step": 57080
    },
    {
      "entropy": 1.8784978061914444,
      "epoch": 0.17697400490871146,
      "grad_norm": 9.919316291809082,
      "learning_rate": 6.013699786444032e-06,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8381554082036018,
      "num_tokens": 68657900.0,
      "step": 57090
    },
    {
      "entropy": 1.870173905789852,
      "epoch": 0.17700500403376115,
      "grad_norm": 8.955236434936523,
      "learning_rate": 6.013173160486402e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8371668472886086,
      "num_tokens": 68670085.0,
      "step": 57100
    },
    {
      "entropy": 1.836390271782875,
      "epoch": 0.17703600315881085,
      "grad_norm": 8.182779312133789,
      "learning_rate": 6.012646672856096e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8486859187483787,
      "num_tokens": 68682961.0,
      "step": 57110
    },
    {
      "entropy": 1.8538921266794204,
      "epoch": 0.17706700228386055,
      "grad_norm": 9.032723426818848,
      "learning_rate": 6.012120323492569e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8346615865826607,
      "num_tokens": 68695560.0,
      "step": 57120
    },
    {
      "entropy": 1.7396719381213188,
      "epoch": 0.17709800140891024,
      "grad_norm": 9.137755393981934,
      "learning_rate": 6.0115941123353115e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.8600768327713013,
      "num_tokens": 68709891.0,
      "step": 57130
    },
    {
      "entropy": 1.9093570321798325,
      "epoch": 0.17712900053395994,
      "grad_norm": 8.14976978302002,
      "learning_rate": 6.011068039323853e-06,
      "loss": 0.5074,
      "mean_token_accuracy": 0.845938429236412,
      "num_tokens": 68721365.0,
      "step": 57140
    },
    {
      "entropy": 1.9436588928103447,
      "epoch": 0.17715999965900964,
      "grad_norm": 7.569816589355469,
      "learning_rate": 6.010542104397757e-06,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8280878692865372,
      "num_tokens": 68733349.0,
      "step": 57150
    },
    {
      "entropy": 1.8442727386951447,
      "epoch": 0.17719099878405933,
      "grad_norm": 8.786954879760742,
      "learning_rate": 6.0100163074966265e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8567957922816276,
      "num_tokens": 68746143.0,
      "step": 57160
    },
    {
      "entropy": 1.9321425527334213,
      "epoch": 0.177221997909109,
      "grad_norm": 5.256120681762695,
      "learning_rate": 6.009490648560099e-06,
      "loss": 0.5664,
      "mean_token_accuracy": 0.8285882025957108,
      "num_tokens": 68757817.0,
      "step": 57170
    },
    {
      "entropy": 1.8465531781315803,
      "epoch": 0.1772529970341587,
      "grad_norm": 9.743329048156738,
      "learning_rate": 6.008965127527853e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.842889928817749,
      "num_tokens": 68771328.0,
      "step": 57180
    },
    {
      "entropy": 1.9369317084550857,
      "epoch": 0.1772839961592084,
      "grad_norm": 3.970804452896118,
      "learning_rate": 6.008439744339599e-06,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8342833817005157,
      "num_tokens": 68782338.0,
      "step": 57190
    },
    {
      "entropy": 1.8963096350431443,
      "epoch": 0.1773149952842581,
      "grad_norm": 10.260050773620605,
      "learning_rate": 6.007914498935089e-06,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8387321591377258,
      "num_tokens": 68793787.0,
      "step": 57200
    },
    {
      "entropy": 1.9125867441296578,
      "epoch": 0.1773459944093078,
      "grad_norm": 8.963105201721191,
      "learning_rate": 6.007389391254107e-06,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8316601321101189,
      "num_tokens": 68805415.0,
      "step": 57210
    },
    {
      "entropy": 1.9211559012532233,
      "epoch": 0.17737699353435749,
      "grad_norm": 8.377967834472656,
      "learning_rate": 6.006864421236479e-06,
      "loss": 0.5096,
      "mean_token_accuracy": 0.84478460252285,
      "num_tokens": 68816320.0,
      "step": 57220
    },
    {
      "entropy": 1.9259533405303955,
      "epoch": 0.17740799265940718,
      "grad_norm": 8.574922561645508,
      "learning_rate": 6.0063395888220646e-06,
      "loss": 0.6228,
      "mean_token_accuracy": 0.8221300706267357,
      "num_tokens": 68828212.0,
      "step": 57230
    },
    {
      "entropy": 1.94825499355793,
      "epoch": 0.17743899178445688,
      "grad_norm": 7.872236251831055,
      "learning_rate": 6.00581489395076e-06,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8307785525918007,
      "num_tokens": 68839479.0,
      "step": 57240
    },
    {
      "entropy": 1.8726135820150376,
      "epoch": 0.17746999090950658,
      "grad_norm": 9.208822250366211,
      "learning_rate": 6.005290336562501e-06,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8378603532910347,
      "num_tokens": 68851565.0,
      "step": 57250
    },
    {
      "entropy": 1.899706956744194,
      "epoch": 0.17750099003455627,
      "grad_norm": 8.506078720092773,
      "learning_rate": 6.004765916597255e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.842448279261589,
      "num_tokens": 68863462.0,
      "step": 57260
    },
    {
      "entropy": 1.8596750691533088,
      "epoch": 0.17753198915960597,
      "grad_norm": 9.010077476501465,
      "learning_rate": 6.004241633995031e-06,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8384414002299309,
      "num_tokens": 68875441.0,
      "step": 57270
    },
    {
      "entropy": 1.9451793283224106,
      "epoch": 0.17756298828465567,
      "grad_norm": 14.787067413330078,
      "learning_rate": 6.0037174886958745e-06,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8292500406503678,
      "num_tokens": 68886706.0,
      "step": 57280
    },
    {
      "entropy": 1.9556769192218781,
      "epoch": 0.17759398740970536,
      "grad_norm": 9.24488353729248,
      "learning_rate": 6.003193480639865e-06,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8255641996860504,
      "num_tokens": 68899129.0,
      "step": 57290
    },
    {
      "entropy": 1.9216732487082482,
      "epoch": 0.17762498653475506,
      "grad_norm": 9.25411605834961,
      "learning_rate": 6.0026696097671166e-06,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8260634735226631,
      "num_tokens": 68911411.0,
      "step": 57300
    },
    {
      "entropy": 1.8534406289458274,
      "epoch": 0.17765598565980475,
      "grad_norm": 9.491771697998047,
      "learning_rate": 6.002145876017787e-06,
      "loss": 0.539,
      "mean_token_accuracy": 0.8245239853858948,
      "num_tokens": 68923378.0,
      "step": 57310
    },
    {
      "entropy": 1.8846858441829681,
      "epoch": 0.17768698478485445,
      "grad_norm": 8.265665054321289,
      "learning_rate": 6.001622279332065e-06,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8445610210299492,
      "num_tokens": 68934573.0,
      "step": 57320
    },
    {
      "entropy": 1.88294820189476,
      "epoch": 0.17771798390990415,
      "grad_norm": 6.753918170928955,
      "learning_rate": 6.001098819650178e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.846784770488739,
      "num_tokens": 68946288.0,
      "step": 57330
    },
    {
      "entropy": 1.8599652536213398,
      "epoch": 0.17774898303495384,
      "grad_norm": 3.2606992721557617,
      "learning_rate": 6.000575496912389e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8316014409065247,
      "num_tokens": 68958131.0,
      "step": 57340
    },
    {
      "entropy": 1.8438660085201264,
      "epoch": 0.17777998216000354,
      "grad_norm": 4.111084461212158,
      "learning_rate": 6.000052311058995e-06,
      "loss": 0.5221,
      "mean_token_accuracy": 0.8313129052519799,
      "num_tokens": 68970796.0,
      "step": 57350
    },
    {
      "entropy": 1.95847550034523,
      "epoch": 0.17781098128505324,
      "grad_norm": 8.4618501663208,
      "learning_rate": 5.999529262030336e-06,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8282546654343605,
      "num_tokens": 68982324.0,
      "step": 57360
    },
    {
      "entropy": 1.8982812404632567,
      "epoch": 0.17784198041010293,
      "grad_norm": 9.305404663085938,
      "learning_rate": 5.999006349766783e-06,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8398912832140922,
      "num_tokens": 68993770.0,
      "step": 57370
    },
    {
      "entropy": 1.9295596539974214,
      "epoch": 0.17787297953515263,
      "grad_norm": 8.27092456817627,
      "learning_rate": 5.998483574208745e-06,
      "loss": 0.521,
      "mean_token_accuracy": 0.8346618101000786,
      "num_tokens": 69005186.0,
      "step": 57380
    },
    {
      "entropy": 1.8877547219395638,
      "epoch": 0.17790397866020233,
      "grad_norm": 9.524073600769043,
      "learning_rate": 5.997960935296666e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8391301646828652,
      "num_tokens": 69016767.0,
      "step": 57390
    },
    {
      "entropy": 1.8745584458112716,
      "epoch": 0.17793497778525202,
      "grad_norm": 9.270411491394043,
      "learning_rate": 5.99743843297103e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8369218707084656,
      "num_tokens": 69028704.0,
      "step": 57400
    },
    {
      "entropy": 1.910376462340355,
      "epoch": 0.1779659769103017,
      "grad_norm": 9.455318450927734,
      "learning_rate": 5.9969160671723535e-06,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8310431107878685,
      "num_tokens": 69039809.0,
      "step": 57410
    },
    {
      "entropy": 1.8097993165254593,
      "epoch": 0.1779969760353514,
      "grad_norm": 6.114592552185059,
      "learning_rate": 5.996393837841191e-06,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8315378293395043,
      "num_tokens": 69053275.0,
      "step": 57420
    },
    {
      "entropy": 1.8377384558320045,
      "epoch": 0.1780279751604011,
      "grad_norm": 10.482206344604492,
      "learning_rate": 5.995871744918132e-06,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8455873817205429,
      "num_tokens": 69065797.0,
      "step": 57430
    },
    {
      "entropy": 1.8282571971416473,
      "epoch": 0.17805897428545078,
      "grad_norm": 8.590948104858398,
      "learning_rate": 5.995349788343804e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8498968616127968,
      "num_tokens": 69078035.0,
      "step": 57440
    },
    {
      "entropy": 1.8669711872935295,
      "epoch": 0.17808997341050048,
      "grad_norm": 8.990492820739746,
      "learning_rate": 5.994827968058869e-06,
      "loss": 0.532,
      "mean_token_accuracy": 0.8290546640753746,
      "num_tokens": 69090734.0,
      "step": 57450
    },
    {
      "entropy": 1.8619839981198312,
      "epoch": 0.17812097253555018,
      "grad_norm": 3.9954440593719482,
      "learning_rate": 5.9943062840040275e-06,
      "loss": 0.4637,
      "mean_token_accuracy": 0.8474828630685807,
      "num_tokens": 69103255.0,
      "step": 57460
    },
    {
      "entropy": 1.7928624168038367,
      "epoch": 0.17815197166059987,
      "grad_norm": 9.835958480834961,
      "learning_rate": 5.993784736120013e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8395032018423081,
      "num_tokens": 69115875.0,
      "step": 57470
    },
    {
      "entropy": 1.8707625821232796,
      "epoch": 0.17818297078564957,
      "grad_norm": 8.842391014099121,
      "learning_rate": 5.9932633243475954e-06,
      "loss": 0.4943,
      "mean_token_accuracy": 0.842163647711277,
      "num_tokens": 69127843.0,
      "step": 57480
    },
    {
      "entropy": 1.7798241943120956,
      "epoch": 0.17821396991069927,
      "grad_norm": 9.086782455444336,
      "learning_rate": 5.992742048627585e-06,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8449828371405601,
      "num_tokens": 69140178.0,
      "step": 57490
    },
    {
      "entropy": 1.885751624405384,
      "epoch": 0.17824496903574896,
      "grad_norm": 7.488559722900391,
      "learning_rate": 5.992220908900822e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8432123690843583,
      "num_tokens": 69151411.0,
      "step": 57500
    },
    {
      "entropy": 1.9364232003688813,
      "epoch": 0.17827596816079866,
      "grad_norm": 7.98928165435791,
      "learning_rate": 5.991699905108188e-06,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8352708846330643,
      "num_tokens": 69162326.0,
      "step": 57510
    },
    {
      "entropy": 1.7974600687623024,
      "epoch": 0.17830696728584836,
      "grad_norm": 8.123475074768066,
      "learning_rate": 5.991179037190596e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8448081642389298,
      "num_tokens": 69174698.0,
      "step": 57520
    },
    {
      "entropy": 1.8795274257659913,
      "epoch": 0.17833796641089805,
      "grad_norm": 5.942895412445068,
      "learning_rate": 5.9906583050889985e-06,
      "loss": 0.5132,
      "mean_token_accuracy": 0.847043776512146,
      "num_tokens": 69186725.0,
      "step": 57530
    },
    {
      "entropy": 1.895848709344864,
      "epoch": 0.17836896553594775,
      "grad_norm": 10.697137832641602,
      "learning_rate": 5.990137708744383e-06,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8229943498969078,
      "num_tokens": 69198477.0,
      "step": 57540
    },
    {
      "entropy": 1.8663012593984605,
      "epoch": 0.17839996466099745,
      "grad_norm": 10.071105003356934,
      "learning_rate": 5.989617248097771e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8398059472441674,
      "num_tokens": 69210496.0,
      "step": 57550
    },
    {
      "entropy": 1.8629123076796532,
      "epoch": 0.17843096378604714,
      "grad_norm": 8.073697090148926,
      "learning_rate": 5.989096923090223e-06,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8346897527575493,
      "num_tokens": 69222754.0,
      "step": 57560
    },
    {
      "entropy": 1.947044813632965,
      "epoch": 0.17846196291109684,
      "grad_norm": 10.283122062683105,
      "learning_rate": 5.988576733662831e-06,
      "loss": 0.5859,
      "mean_token_accuracy": 0.8340685039758682,
      "num_tokens": 69233747.0,
      "step": 57570
    },
    {
      "entropy": 1.8908654794096946,
      "epoch": 0.17849296203614654,
      "grad_norm": 8.264286041259766,
      "learning_rate": 5.988056679756728e-06,
      "loss": 0.573,
      "mean_token_accuracy": 0.8300464496016502,
      "num_tokens": 69244865.0,
      "step": 57580
    },
    {
      "entropy": 1.9280391097068788,
      "epoch": 0.17852396116119623,
      "grad_norm": 8.630804061889648,
      "learning_rate": 5.9875367613130775e-06,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8240269988775253,
      "num_tokens": 69255702.0,
      "step": 57590
    },
    {
      "entropy": 1.935716523230076,
      "epoch": 0.17855496028624593,
      "grad_norm": 11.321623802185059,
      "learning_rate": 5.987016978273085e-06,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8363067075610161,
      "num_tokens": 69266825.0,
      "step": 57600
    },
    {
      "entropy": 1.9014330476522445,
      "epoch": 0.17858595941129563,
      "grad_norm": 4.646116733551025,
      "learning_rate": 5.986497330577986e-06,
      "loss": 0.522,
      "mean_token_accuracy": 0.8390189468860626,
      "num_tokens": 69279081.0,
      "step": 57610
    },
    {
      "entropy": 1.9175360828638077,
      "epoch": 0.17861695853634532,
      "grad_norm": 9.152368545532227,
      "learning_rate": 5.985977818169053e-06,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8228761807084084,
      "num_tokens": 69290443.0,
      "step": 57620
    },
    {
      "entropy": 1.8552556172013284,
      "epoch": 0.17864795766139502,
      "grad_norm": 7.568161487579346,
      "learning_rate": 5.985458440987597e-06,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8391215577721596,
      "num_tokens": 69303861.0,
      "step": 57630
    },
    {
      "entropy": 1.7614154547452927,
      "epoch": 0.17867895678644471,
      "grad_norm": 3.9971959590911865,
      "learning_rate": 5.984939198974961e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8515268057584763,
      "num_tokens": 69317877.0,
      "step": 57640
    },
    {
      "entropy": 1.9185373350977897,
      "epoch": 0.1787099559114944,
      "grad_norm": 11.149886131286621,
      "learning_rate": 5.984420092072528e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8318190723657608,
      "num_tokens": 69329893.0,
      "step": 57650
    },
    {
      "entropy": 1.9363491863012314,
      "epoch": 0.17874095503654408,
      "grad_norm": 9.132575035095215,
      "learning_rate": 5.983901120221711e-06,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8298118308186531,
      "num_tokens": 69341278.0,
      "step": 57660
    },
    {
      "entropy": 1.833200243115425,
      "epoch": 0.17877195416159378,
      "grad_norm": 4.589186668395996,
      "learning_rate": 5.983382283363963e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8535523906350135,
      "num_tokens": 69354018.0,
      "step": 57670
    },
    {
      "entropy": 1.952676109969616,
      "epoch": 0.17880295328664347,
      "grad_norm": 8.8577241897583,
      "learning_rate": 5.9828635814407695e-06,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8286620289087295,
      "num_tokens": 69365463.0,
      "step": 57680
    },
    {
      "entropy": 1.901480646431446,
      "epoch": 0.17883395241169317,
      "grad_norm": 8.063117027282715,
      "learning_rate": 5.9823450143936555e-06,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8349184259772301,
      "num_tokens": 69377683.0,
      "step": 57690
    },
    {
      "entropy": 1.8743933662772179,
      "epoch": 0.17886495153674287,
      "grad_norm": 8.365898132324219,
      "learning_rate": 5.981826582164176e-06,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8368162304162979,
      "num_tokens": 69390015.0,
      "step": 57700
    },
    {
      "entropy": 1.776149820536375,
      "epoch": 0.17889595066179256,
      "grad_norm": 9.224735260009766,
      "learning_rate": 5.9813082846939264e-06,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8448210716247558,
      "num_tokens": 69402894.0,
      "step": 57710
    },
    {
      "entropy": 1.7384593583643437,
      "epoch": 0.17892694978684226,
      "grad_norm": 8.464068412780762,
      "learning_rate": 5.980790121924534e-06,
      "loss": 0.3974,
      "mean_token_accuracy": 0.8530717685818672,
      "num_tokens": 69416819.0,
      "step": 57720
    },
    {
      "entropy": 1.8666021645069122,
      "epoch": 0.17895794891189196,
      "grad_norm": 10.071202278137207,
      "learning_rate": 5.980272093797667e-06,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8388993337750434,
      "num_tokens": 69428622.0,
      "step": 57730
    },
    {
      "entropy": 1.9105965510010718,
      "epoch": 0.17898894803694165,
      "grad_norm": 7.725157737731934,
      "learning_rate": 5.979754200255019e-06,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8272575065493584,
      "num_tokens": 69439573.0,
      "step": 57740
    },
    {
      "entropy": 1.880422416329384,
      "epoch": 0.17901994716199135,
      "grad_norm": 4.6876630783081055,
      "learning_rate": 5.979236441238329e-06,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8352604731917381,
      "num_tokens": 69451040.0,
      "step": 57750
    },
    {
      "entropy": 1.877747993171215,
      "epoch": 0.17905094628704105,
      "grad_norm": 3.8717169761657715,
      "learning_rate": 5.978718816689365e-06,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8352397963404655,
      "num_tokens": 69462601.0,
      "step": 57760
    },
    {
      "entropy": 1.9101334929466247,
      "epoch": 0.17908194541209074,
      "grad_norm": 9.124340057373047,
      "learning_rate": 5.978201326549935e-06,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8258951917290688,
      "num_tokens": 69474464.0,
      "step": 57770
    },
    {
      "entropy": 1.7846893429756165,
      "epoch": 0.17911294453714044,
      "grad_norm": 3.8451902866363525,
      "learning_rate": 5.977683970761876e-06,
      "loss": 0.465,
      "mean_token_accuracy": 0.8453226387500763,
      "num_tokens": 69487808.0,
      "step": 57780
    },
    {
      "entropy": 1.8912085384130477,
      "epoch": 0.17914394366219014,
      "grad_norm": 9.206360816955566,
      "learning_rate": 5.9771667492670675e-06,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8258654251694679,
      "num_tokens": 69499784.0,
      "step": 57790
    },
    {
      "entropy": 1.8862205877900124,
      "epoch": 0.17917494278723983,
      "grad_norm": 4.152833461761475,
      "learning_rate": 5.97664966200742e-06,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8353832781314849,
      "num_tokens": 69511090.0,
      "step": 57800
    },
    {
      "entropy": 1.8243310183286667,
      "epoch": 0.17920594191228953,
      "grad_norm": 11.02452278137207,
      "learning_rate": 5.9761327089248786e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.840731629729271,
      "num_tokens": 69523788.0,
      "step": 57810
    },
    {
      "entropy": 1.9363866940140724,
      "epoch": 0.17923694103733923,
      "grad_norm": 4.2471923828125,
      "learning_rate": 5.975615889961425e-06,
      "loss": 0.5913,
      "mean_token_accuracy": 0.8229147881269455,
      "num_tokens": 69534659.0,
      "step": 57820
    },
    {
      "entropy": 1.8675603151321412,
      "epoch": 0.17926794016238892,
      "grad_norm": 8.263809204101562,
      "learning_rate": 5.9750992050590765e-06,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8418093547224998,
      "num_tokens": 69546361.0,
      "step": 57830
    },
    {
      "entropy": 1.8654843851923943,
      "epoch": 0.17929893928743862,
      "grad_norm": 9.333740234375,
      "learning_rate": 5.974582654159884e-06,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8490580499172211,
      "num_tokens": 69558369.0,
      "step": 57840
    },
    {
      "entropy": 1.7780871927738189,
      "epoch": 0.17932993841248832,
      "grad_norm": 3.7377822399139404,
      "learning_rate": 5.974066237205935e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8493701010942459,
      "num_tokens": 69571586.0,
      "step": 57850
    },
    {
      "entropy": 1.8709719255566597,
      "epoch": 0.179360937537538,
      "grad_norm": 9.209651947021484,
      "learning_rate": 5.9735499541393515e-06,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8328555643558502,
      "num_tokens": 69584134.0,
      "step": 57860
    },
    {
      "entropy": 1.9305126339197158,
      "epoch": 0.1793919366625877,
      "grad_norm": 8.458016395568848,
      "learning_rate": 5.9730338049022905e-06,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8329010114073754,
      "num_tokens": 69595889.0,
      "step": 57870
    },
    {
      "entropy": 1.9079819098114967,
      "epoch": 0.1794229357876374,
      "grad_norm": 8.438733100891113,
      "learning_rate": 5.972517789436941e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8301453992724419,
      "num_tokens": 69607872.0,
      "step": 57880
    },
    {
      "entropy": 1.8880163133144379,
      "epoch": 0.1794539349126871,
      "grad_norm": 7.341186046600342,
      "learning_rate": 5.972001907685534e-06,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8447798684239387,
      "num_tokens": 69619608.0,
      "step": 57890
    },
    {
      "entropy": 1.9878419309854507,
      "epoch": 0.1794849340377368,
      "grad_norm": 8.431941032409668,
      "learning_rate": 5.9714861595903275e-06,
      "loss": 0.5709,
      "mean_token_accuracy": 0.8319448336958886,
      "num_tokens": 69630342.0,
      "step": 57900
    },
    {
      "entropy": 1.9403752774000167,
      "epoch": 0.17951593316278647,
      "grad_norm": 8.231103897094727,
      "learning_rate": 5.9709705450936195e-06,
      "loss": 0.5535,
      "mean_token_accuracy": 0.8287773564457893,
      "num_tokens": 69641589.0,
      "step": 57910
    },
    {
      "entropy": 1.8349755868315696,
      "epoch": 0.17954693228783616,
      "grad_norm": 8.210265159606934,
      "learning_rate": 5.9704550641377414e-06,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8326278194785118,
      "num_tokens": 69654695.0,
      "step": 57920
    },
    {
      "entropy": 1.8748487085103989,
      "epoch": 0.17957793141288586,
      "grad_norm": 4.568990230560303,
      "learning_rate": 5.96993971666506e-06,
      "loss": 0.516,
      "mean_token_accuracy": 0.8239820554852486,
      "num_tokens": 69667546.0,
      "step": 57930
    },
    {
      "entropy": 1.957274827361107,
      "epoch": 0.17960893053793556,
      "grad_norm": 8.795912742614746,
      "learning_rate": 5.969424502617975e-06,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8339650616049766,
      "num_tokens": 69678423.0,
      "step": 57940
    },
    {
      "entropy": 1.9020697817206382,
      "epoch": 0.17963992966298525,
      "grad_norm": 8.028707504272461,
      "learning_rate": 5.968909421938924e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.83273094445467,
      "num_tokens": 69690504.0,
      "step": 57950
    },
    {
      "entropy": 1.9501001119613648,
      "epoch": 0.17967092878803495,
      "grad_norm": 8.39437198638916,
      "learning_rate": 5.968394474570377e-06,
      "loss": 0.5423,
      "mean_token_accuracy": 0.8293194517493248,
      "num_tokens": 69701909.0,
      "step": 57960
    },
    {
      "entropy": 1.8961822897195817,
      "epoch": 0.17970192791308465,
      "grad_norm": 3.989229202270508,
      "learning_rate": 5.9678796604548385e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8313602998852729,
      "num_tokens": 69713928.0,
      "step": 57970
    },
    {
      "entropy": 1.9584363132715226,
      "epoch": 0.17973292703813434,
      "grad_norm": 10.755518913269043,
      "learning_rate": 5.967364979534849e-06,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8270615398883819,
      "num_tokens": 69725093.0,
      "step": 57980
    },
    {
      "entropy": 1.883443070948124,
      "epoch": 0.17976392616318404,
      "grad_norm": 8.810373306274414,
      "learning_rate": 5.966850431752984e-06,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8435442551970482,
      "num_tokens": 69736710.0,
      "step": 57990
    },
    {
      "entropy": 1.838029107451439,
      "epoch": 0.17979492528823374,
      "grad_norm": 7.354464530944824,
      "learning_rate": 5.9663360170518524e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8321570068597793,
      "num_tokens": 69749525.0,
      "step": 58000
    },
    {
      "entropy": 1.860961727797985,
      "epoch": 0.17982592441328343,
      "grad_norm": 8.58539867401123,
      "learning_rate": 5.965821735374097e-06,
      "loss": 0.5197,
      "mean_token_accuracy": 0.842749148607254,
      "num_tokens": 69761090.0,
      "step": 58010
    },
    {
      "entropy": 1.888116455078125,
      "epoch": 0.17985692353833313,
      "grad_norm": 8.116665840148926,
      "learning_rate": 5.965307586662398e-06,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8538775756955147,
      "num_tokens": 69773218.0,
      "step": 58020
    },
    {
      "entropy": 1.88920701444149,
      "epoch": 0.17988792266338283,
      "grad_norm": 8.906111717224121,
      "learning_rate": 5.964793570859469e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8328505247831345,
      "num_tokens": 69785392.0,
      "step": 58030
    },
    {
      "entropy": 1.9361517250537872,
      "epoch": 0.17991892178843252,
      "grad_norm": 10.055381774902344,
      "learning_rate": 5.964279687908057e-06,
      "loss": 0.5385,
      "mean_token_accuracy": 0.8345837727189064,
      "num_tokens": 69796770.0,
      "step": 58040
    },
    {
      "entropy": 1.856403675675392,
      "epoch": 0.17994992091348222,
      "grad_norm": 2.6345648765563965,
      "learning_rate": 5.963765937750943e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8511771202087403,
      "num_tokens": 69808863.0,
      "step": 58050
    },
    {
      "entropy": 1.8975456178188324,
      "epoch": 0.17998092003853192,
      "grad_norm": 9.3054780960083,
      "learning_rate": 5.963252320330947e-06,
      "loss": 0.5478,
      "mean_token_accuracy": 0.827577319741249,
      "num_tokens": 69820718.0,
      "step": 58060
    },
    {
      "entropy": 1.8229490399360657,
      "epoch": 0.1800119191635816,
      "grad_norm": 8.946834564208984,
      "learning_rate": 5.962738835590917e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8495317012071609,
      "num_tokens": 69834215.0,
      "step": 58070
    },
    {
      "entropy": 1.9095330134034156,
      "epoch": 0.1800429182886313,
      "grad_norm": 8.070923805236816,
      "learning_rate": 5.962225483473742e-06,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8391377106308937,
      "num_tokens": 69846084.0,
      "step": 58080
    },
    {
      "entropy": 1.7783651992678642,
      "epoch": 0.180073917413681,
      "grad_norm": 7.508241176605225,
      "learning_rate": 5.961712263922337e-06,
      "loss": 0.4244,
      "mean_token_accuracy": 0.8533773466944694,
      "num_tokens": 69859333.0,
      "step": 58090
    },
    {
      "entropy": 1.926021693646908,
      "epoch": 0.1801049165387307,
      "grad_norm": 9.783551216125488,
      "learning_rate": 5.961199176879661e-06,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8343341812491417,
      "num_tokens": 69870605.0,
      "step": 58100
    },
    {
      "entropy": 1.9040902271866798,
      "epoch": 0.1801359156637804,
      "grad_norm": 8.443004608154297,
      "learning_rate": 5.960686222288703e-06,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8259869500994682,
      "num_tokens": 69883254.0,
      "step": 58110
    },
    {
      "entropy": 1.887841096520424,
      "epoch": 0.1801669147888301,
      "grad_norm": 9.13963794708252,
      "learning_rate": 5.960173400092483e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8409109726548195,
      "num_tokens": 69895422.0,
      "step": 58120
    },
    {
      "entropy": 1.8949230402708053,
      "epoch": 0.1801979139138798,
      "grad_norm": 8.608329772949219,
      "learning_rate": 5.9596607102340605e-06,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8356387764215469,
      "num_tokens": 69907555.0,
      "step": 58130
    },
    {
      "entropy": 1.9181236669421196,
      "epoch": 0.1802289130389295,
      "grad_norm": 7.1196513175964355,
      "learning_rate": 5.959148152656526e-06,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8519378170371056,
      "num_tokens": 69919079.0,
      "step": 58140
    },
    {
      "entropy": 1.9497553408145905,
      "epoch": 0.18025991216397916,
      "grad_norm": 9.516104698181152,
      "learning_rate": 5.958635727303008e-06,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8375312581658363,
      "num_tokens": 69929946.0,
      "step": 58150
    },
    {
      "entropy": 1.8249136477708816,
      "epoch": 0.18029091128902885,
      "grad_norm": 4.427674770355225,
      "learning_rate": 5.958123434116665e-06,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8447049587965012,
      "num_tokens": 69942319.0,
      "step": 58160
    },
    {
      "entropy": 1.8562700033187867,
      "epoch": 0.18032191041407855,
      "grad_norm": 9.44267749786377,
      "learning_rate": 5.957611273040691e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8396990388631821,
      "num_tokens": 69954897.0,
      "step": 58170
    },
    {
      "entropy": 1.9479858607053757,
      "epoch": 0.18035290953912825,
      "grad_norm": 4.776163101196289,
      "learning_rate": 5.9570992440183166e-06,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8349523320794106,
      "num_tokens": 69966494.0,
      "step": 58180
    },
    {
      "entropy": 1.9115686953067779,
      "epoch": 0.18038390866417794,
      "grad_norm": 8.399941444396973,
      "learning_rate": 5.956587346992802e-06,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8438914701342582,
      "num_tokens": 69978598.0,
      "step": 58190
    },
    {
      "entropy": 1.8765361204743385,
      "epoch": 0.18041490778922764,
      "grad_norm": 9.405889511108398,
      "learning_rate": 5.956075581907446e-06,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8236325919628144,
      "num_tokens": 69990680.0,
      "step": 58200
    },
    {
      "entropy": 1.9489685088396071,
      "epoch": 0.18044590691427734,
      "grad_norm": 8.940106391906738,
      "learning_rate": 5.955563948705578e-06,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8357775121927261,
      "num_tokens": 70001208.0,
      "step": 58210
    },
    {
      "entropy": 1.8073627695441246,
      "epoch": 0.18047690603932703,
      "grad_norm": 7.959947109222412,
      "learning_rate": 5.955052447330566e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.856120876967907,
      "num_tokens": 70015168.0,
      "step": 58220
    },
    {
      "entropy": 1.912624678015709,
      "epoch": 0.18050790516437673,
      "grad_norm": 10.096214294433594,
      "learning_rate": 5.954541077725806e-06,
      "loss": 0.5486,
      "mean_token_accuracy": 0.821705561876297,
      "num_tokens": 70027352.0,
      "step": 58230
    },
    {
      "entropy": 1.7947795525193215,
      "epoch": 0.18053890428942643,
      "grad_norm": 3.570014715194702,
      "learning_rate": 5.954029839834733e-06,
      "loss": 0.395,
      "mean_token_accuracy": 0.8589225232601165,
      "num_tokens": 70040539.0,
      "step": 58240
    },
    {
      "entropy": 1.9037404909729958,
      "epoch": 0.18056990341447612,
      "grad_norm": 7.7414655685424805,
      "learning_rate": 5.953518733600813e-06,
      "loss": 0.5514,
      "mean_token_accuracy": 0.828503304719925,
      "num_tokens": 70052279.0,
      "step": 58250
    },
    {
      "entropy": 1.898376226425171,
      "epoch": 0.18060090253952582,
      "grad_norm": 4.189719200134277,
      "learning_rate": 5.953007758967547e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8411821514368057,
      "num_tokens": 70064479.0,
      "step": 58260
    },
    {
      "entropy": 1.8012950256466866,
      "epoch": 0.18063190166457552,
      "grad_norm": 2.9318971633911133,
      "learning_rate": 5.952496915878471e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8473293244838714,
      "num_tokens": 70077477.0,
      "step": 58270
    },
    {
      "entropy": 1.883352592587471,
      "epoch": 0.1806629007896252,
      "grad_norm": 8.893994331359863,
      "learning_rate": 5.951986204277154e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8392975389957428,
      "num_tokens": 70089437.0,
      "step": 58280
    },
    {
      "entropy": 1.9155563935637474,
      "epoch": 0.1806938999146749,
      "grad_norm": 9.788602828979492,
      "learning_rate": 5.951475624107198e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8240778625011445,
      "num_tokens": 70101237.0,
      "step": 58290
    },
    {
      "entropy": 1.8512123331427575,
      "epoch": 0.1807248990397246,
      "grad_norm": 8.938436508178711,
      "learning_rate": 5.950965175312241e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.836862875521183,
      "num_tokens": 70114119.0,
      "step": 58300
    },
    {
      "entropy": 1.8680644989013673,
      "epoch": 0.1807558981647743,
      "grad_norm": 11.42545223236084,
      "learning_rate": 5.950454857835951e-06,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8473235473036766,
      "num_tokens": 70125999.0,
      "step": 58310
    },
    {
      "entropy": 1.8300606310367584,
      "epoch": 0.180786897289824,
      "grad_norm": 6.41502046585083,
      "learning_rate": 5.949944671622034e-06,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8300027817487716,
      "num_tokens": 70139069.0,
      "step": 58320
    },
    {
      "entropy": 1.9075913026928901,
      "epoch": 0.1808178964148737,
      "grad_norm": 3.9195168018341064,
      "learning_rate": 5.949434616614229e-06,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8323316410183906,
      "num_tokens": 70151975.0,
      "step": 58330
    },
    {
      "entropy": 1.979878196120262,
      "epoch": 0.1808488955399234,
      "grad_norm": 8.486638069152832,
      "learning_rate": 5.948924692756306e-06,
      "loss": 0.5894,
      "mean_token_accuracy": 0.8257222607731819,
      "num_tokens": 70162362.0,
      "step": 58340
    },
    {
      "entropy": 1.891125389933586,
      "epoch": 0.1808798946649731,
      "grad_norm": 4.342275142669678,
      "learning_rate": 5.948414899992072e-06,
      "loss": 0.542,
      "mean_token_accuracy": 0.8310712307691575,
      "num_tokens": 70174516.0,
      "step": 58350
    },
    {
      "entropy": 1.8640663996338844,
      "epoch": 0.1809108937900228,
      "grad_norm": 2.7677135467529297,
      "learning_rate": 5.947905238265366e-06,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8343622460961342,
      "num_tokens": 70186361.0,
      "step": 58360
    },
    {
      "entropy": 1.863309583067894,
      "epoch": 0.18094189291507248,
      "grad_norm": 9.252532958984375,
      "learning_rate": 5.947395707520059e-06,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8355363115668297,
      "num_tokens": 70198676.0,
      "step": 58370
    },
    {
      "entropy": 1.9271214351058006,
      "epoch": 0.18097289204012218,
      "grad_norm": 8.357312202453613,
      "learning_rate": 5.946886307700062e-06,
      "loss": 0.535,
      "mean_token_accuracy": 0.8216869860887528,
      "num_tokens": 70210561.0,
      "step": 58380
    },
    {
      "entropy": 1.8037690997123719,
      "epoch": 0.18100389116517188,
      "grad_norm": 8.436007499694824,
      "learning_rate": 5.94637703874931e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8438222482800484,
      "num_tokens": 70224444.0,
      "step": 58390
    },
    {
      "entropy": 1.9045485824346542,
      "epoch": 0.18103489029022155,
      "grad_norm": 10.353751182556152,
      "learning_rate": 5.9458679006117815e-06,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8378313854336739,
      "num_tokens": 70236240.0,
      "step": 58400
    },
    {
      "entropy": 1.9282925948500633,
      "epoch": 0.18106588941527124,
      "grad_norm": 9.726922035217285,
      "learning_rate": 5.94535889323148e-06,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8426464855670929,
      "num_tokens": 70247776.0,
      "step": 58410
    },
    {
      "entropy": 1.9179017692804337,
      "epoch": 0.18109688854032094,
      "grad_norm": 8.368406295776367,
      "learning_rate": 5.944850016552449e-06,
      "loss": 0.5916,
      "mean_token_accuracy": 0.820852018892765,
      "num_tokens": 70259710.0,
      "step": 58420
    },
    {
      "entropy": 1.9104688361287117,
      "epoch": 0.18112788766537063,
      "grad_norm": 9.314393997192383,
      "learning_rate": 5.944341270518763e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8437484547495842,
      "num_tokens": 70271807.0,
      "step": 58430
    },
    {
      "entropy": 1.9487433806061745,
      "epoch": 0.18115888679042033,
      "grad_norm": 9.065957069396973,
      "learning_rate": 5.943832655074528e-06,
      "loss": 0.5531,
      "mean_token_accuracy": 0.822340051829815,
      "num_tokens": 70283681.0,
      "step": 58440
    },
    {
      "entropy": 1.8573688492178917,
      "epoch": 0.18118988591547003,
      "grad_norm": 8.39069652557373,
      "learning_rate": 5.943324170163888e-06,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8431822940707207,
      "num_tokens": 70296587.0,
      "step": 58450
    },
    {
      "entropy": 1.8324483245611192,
      "epoch": 0.18122088504051972,
      "grad_norm": 4.025021076202393,
      "learning_rate": 5.942815815731015e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8505256190896034,
      "num_tokens": 70308977.0,
      "step": 58460
    },
    {
      "entropy": 1.9166489720344544,
      "epoch": 0.18125188416556942,
      "grad_norm": 7.720203876495361,
      "learning_rate": 5.942307591720121e-06,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8491469100117683,
      "num_tokens": 70320331.0,
      "step": 58470
    },
    {
      "entropy": 1.8525561004877091,
      "epoch": 0.18128288329061912,
      "grad_norm": 8.706543922424316,
      "learning_rate": 5.941799498075445e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8432158127427101,
      "num_tokens": 70331972.0,
      "step": 58480
    },
    {
      "entropy": 1.734443087875843,
      "epoch": 0.18131388241566881,
      "grad_norm": 4.216269493103027,
      "learning_rate": 5.9412915347412624e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8574797987937928,
      "num_tokens": 70345958.0,
      "step": 58490
    },
    {
      "entropy": 1.885965469479561,
      "epoch": 0.1813448815407185,
      "grad_norm": 8.010455131530762,
      "learning_rate": 5.940783701661882e-06,
      "loss": 0.4823,
      "mean_token_accuracy": 0.8380753323435783,
      "num_tokens": 70357917.0,
      "step": 58500
    },
    {
      "entropy": 1.9255839720368386,
      "epoch": 0.1813758806657682,
      "grad_norm": 8.077774047851562,
      "learning_rate": 5.940275998781646e-06,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8398918956518173,
      "num_tokens": 70369190.0,
      "step": 58510
    },
    {
      "entropy": 1.8902644366025925,
      "epoch": 0.1814068797908179,
      "grad_norm": 9.07833480834961,
      "learning_rate": 5.939768426044928e-06,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8410121381282807,
      "num_tokens": 70381045.0,
      "step": 58520
    },
    {
      "entropy": 1.8733600035309792,
      "epoch": 0.1814378789158676,
      "grad_norm": 7.863108158111572,
      "learning_rate": 5.939260983396139e-06,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8383708387613297,
      "num_tokens": 70394169.0,
      "step": 58530
    },
    {
      "entropy": 1.9640895150601865,
      "epoch": 0.1814688780409173,
      "grad_norm": 9.28256607055664,
      "learning_rate": 5.938753670779716e-06,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8161120370030404,
      "num_tokens": 70405772.0,
      "step": 58540
    },
    {
      "entropy": 1.8138402692973614,
      "epoch": 0.181499877165967,
      "grad_norm": 9.445598602294922,
      "learning_rate": 5.938246488140139e-06,
      "loss": 0.4378,
      "mean_token_accuracy": 0.8572772964835167,
      "num_tokens": 70418813.0,
      "step": 58550
    },
    {
      "entropy": 1.9521236777305604,
      "epoch": 0.1815308762910167,
      "grad_norm": 6.890706539154053,
      "learning_rate": 5.937739435421913e-06,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8317166283726692,
      "num_tokens": 70430085.0,
      "step": 58560
    },
    {
      "entropy": 1.8753592044115066,
      "epoch": 0.1815618754160664,
      "grad_norm": Infinity,
      "learning_rate": 5.9372325125695796e-06,
      "loss": 0.513,
      "mean_token_accuracy": 0.8267385110259056,
      "num_tokens": 70442432.0,
      "step": 58570
    },
    {
      "entropy": 1.951469287276268,
      "epoch": 0.18159287454111608,
      "grad_norm": 10.176626205444336,
      "learning_rate": 5.936725719527712e-06,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8351786851882934,
      "num_tokens": 70453518.0,
      "step": 58580
    },
    {
      "entropy": 2.006668972969055,
      "epoch": 0.18162387366616578,
      "grad_norm": 10.691234588623047,
      "learning_rate": 5.9362190562409196e-06,
      "loss": 0.6187,
      "mean_token_accuracy": 0.8179316386580467,
      "num_tokens": 70464169.0,
      "step": 58590
    },
    {
      "entropy": 1.9340445652604104,
      "epoch": 0.18165487279121548,
      "grad_norm": 8.558653831481934,
      "learning_rate": 5.93571252265384e-06,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8301935106515884,
      "num_tokens": 70476295.0,
      "step": 58600
    },
    {
      "entropy": 2.03122977912426,
      "epoch": 0.18168587191626517,
      "grad_norm": 8.362186431884766,
      "learning_rate": 5.935206118711151e-06,
      "loss": 0.6264,
      "mean_token_accuracy": 0.8174452885985375,
      "num_tokens": 70487260.0,
      "step": 58610
    },
    {
      "entropy": 1.9458183541893959,
      "epoch": 0.18171687104131487,
      "grad_norm": 4.3328633308410645,
      "learning_rate": 5.934699844357555e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.834861546754837,
      "num_tokens": 70498573.0,
      "step": 58620
    },
    {
      "entropy": 1.9818433836102485,
      "epoch": 0.18174787016636457,
      "grad_norm": 9.87380313873291,
      "learning_rate": 5.934193699537794e-06,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8219478592276573,
      "num_tokens": 70510126.0,
      "step": 58630
    },
    {
      "entropy": 1.9035624399781228,
      "epoch": 0.18177886929141426,
      "grad_norm": 7.710408687591553,
      "learning_rate": 5.933687684196638e-06,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8381109595298767,
      "num_tokens": 70522421.0,
      "step": 58640
    },
    {
      "entropy": 1.9757555976510048,
      "epoch": 0.18180986841646393,
      "grad_norm": 8.179011344909668,
      "learning_rate": 5.933181798278895e-06,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8351363927125931,
      "num_tokens": 70534084.0,
      "step": 58650
    },
    {
      "entropy": 1.9270098477602005,
      "epoch": 0.18184086754151363,
      "grad_norm": 4.406381130218506,
      "learning_rate": 5.9326760417294036e-06,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8375926703214646,
      "num_tokens": 70545476.0,
      "step": 58660
    },
    {
      "entropy": 1.9433436125516892,
      "epoch": 0.18187186666656333,
      "grad_norm": 8.257979393005371,
      "learning_rate": 5.9321704144930335e-06,
      "loss": 0.6162,
      "mean_token_accuracy": 0.8256099238991738,
      "num_tokens": 70557302.0,
      "step": 58670
    },
    {
      "entropy": 1.8725889652967453,
      "epoch": 0.18190286579161302,
      "grad_norm": 8.043081283569336,
      "learning_rate": 5.931664916514689e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8415656134486198,
      "num_tokens": 70569142.0,
      "step": 58680
    },
    {
      "entropy": 1.922675184905529,
      "epoch": 0.18193386491666272,
      "grad_norm": 9.857270240783691,
      "learning_rate": 5.931159547739309e-06,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8421023860573769,
      "num_tokens": 70580427.0,
      "step": 58690
    },
    {
      "entropy": 1.8865813314914703,
      "epoch": 0.18196486404171242,
      "grad_norm": 8.099855422973633,
      "learning_rate": 5.9306543081118605e-06,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8428427502512932,
      "num_tokens": 70591829.0,
      "step": 58700
    },
    {
      "entropy": 1.7871639668941497,
      "epoch": 0.1819958631667621,
      "grad_norm": 8.50505256652832,
      "learning_rate": 5.9301491975773485e-06,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8578494325280189,
      "num_tokens": 70605530.0,
      "step": 58710
    },
    {
      "entropy": 1.9226329445838928,
      "epoch": 0.1820268622918118,
      "grad_norm": 9.982844352722168,
      "learning_rate": 5.929644216080808e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8366420567035675,
      "num_tokens": 70616806.0,
      "step": 58720
    },
    {
      "entropy": 1.8149694345891476,
      "epoch": 0.1820578614168615,
      "grad_norm": 8.859916687011719,
      "learning_rate": 5.9291393635673065e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8594744309782982,
      "num_tokens": 70629601.0,
      "step": 58730
    },
    {
      "entropy": 1.8165232956409454,
      "epoch": 0.1820888605419112,
      "grad_norm": 8.397076606750488,
      "learning_rate": 5.928634639981946e-06,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8303966209292412,
      "num_tokens": 70642458.0,
      "step": 58740
    },
    {
      "entropy": 1.8342901065945625,
      "epoch": 0.1821198596669609,
      "grad_norm": 4.7208123207092285,
      "learning_rate": 5.9281300452698584e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.850199481844902,
      "num_tokens": 70655561.0,
      "step": 58750
    },
    {
      "entropy": 1.959836632013321,
      "epoch": 0.1821508587920106,
      "grad_norm": 7.517179489135742,
      "learning_rate": 5.927625579376213e-06,
      "loss": 0.5588,
      "mean_token_accuracy": 0.8339650839567184,
      "num_tokens": 70666015.0,
      "step": 58760
    },
    {
      "entropy": 1.8640755087137222,
      "epoch": 0.1821818579170603,
      "grad_norm": 8.22607135772705,
      "learning_rate": 5.927121242246206e-06,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8419701069593429,
      "num_tokens": 70678336.0,
      "step": 58770
    },
    {
      "entropy": 1.8295077085494995,
      "epoch": 0.18221285704211,
      "grad_norm": 7.90572452545166,
      "learning_rate": 5.9266170338250715e-06,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8469627618789672,
      "num_tokens": 70691062.0,
      "step": 58780
    },
    {
      "entropy": 1.8726282447576523,
      "epoch": 0.18224385616715968,
      "grad_norm": 9.512833595275879,
      "learning_rate": 5.926112954058072e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.840190976858139,
      "num_tokens": 70702498.0,
      "step": 58790
    },
    {
      "entropy": 1.8806493178009986,
      "epoch": 0.18227485529220938,
      "grad_norm": 9.236995697021484,
      "learning_rate": 5.925609002890504e-06,
      "loss": 0.5467,
      "mean_token_accuracy": 0.8324322685599327,
      "num_tokens": 70713997.0,
      "step": 58800
    },
    {
      "entropy": 1.9413601398468017,
      "epoch": 0.18230585441725908,
      "grad_norm": 8.431058883666992,
      "learning_rate": 5.9251051802677e-06,
      "loss": 0.5951,
      "mean_token_accuracy": 0.8179917827248573,
      "num_tokens": 70725865.0,
      "step": 58810
    },
    {
      "entropy": 1.8975177273154258,
      "epoch": 0.18233685354230877,
      "grad_norm": 9.826079368591309,
      "learning_rate": 5.9246014861350176e-06,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8294125840067863,
      "num_tokens": 70737735.0,
      "step": 58820
    },
    {
      "entropy": 1.8288067810237407,
      "epoch": 0.18236785266735847,
      "grad_norm": 8.987369537353516,
      "learning_rate": 5.924097920437855e-06,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8347997352480888,
      "num_tokens": 70750964.0,
      "step": 58830
    },
    {
      "entropy": 1.9133558467030525,
      "epoch": 0.18239885179240817,
      "grad_norm": 8.020203590393066,
      "learning_rate": 5.923594483121636e-06,
      "loss": 0.553,
      "mean_token_accuracy": 0.8251760572195053,
      "num_tokens": 70762325.0,
      "step": 58840
    },
    {
      "entropy": 1.9130362540483474,
      "epoch": 0.18242985091745786,
      "grad_norm": 4.7962446212768555,
      "learning_rate": 5.923091174131822e-06,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8287200018763542,
      "num_tokens": 70773880.0,
      "step": 58850
    },
    {
      "entropy": 1.8333944439888001,
      "epoch": 0.18246085004250756,
      "grad_norm": 8.215493202209473,
      "learning_rate": 5.922587993413905e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8558073043823242,
      "num_tokens": 70786236.0,
      "step": 58860
    },
    {
      "entropy": 1.8580266639590264,
      "epoch": 0.18249184916755726,
      "grad_norm": 8.1721830368042,
      "learning_rate": 5.922084940913409e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8527326479554176,
      "num_tokens": 70798107.0,
      "step": 58870
    },
    {
      "entropy": 1.8850371599197389,
      "epoch": 0.18252284829260695,
      "grad_norm": 9.118043899536133,
      "learning_rate": 5.921582016575889e-06,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8236063331365585,
      "num_tokens": 70809614.0,
      "step": 58880
    },
    {
      "entropy": 1.819055077433586,
      "epoch": 0.18255384741765662,
      "grad_norm": 6.715834617614746,
      "learning_rate": 5.921079220346936e-06,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8425466433167458,
      "num_tokens": 70821709.0,
      "step": 58890
    },
    {
      "entropy": 1.9392906963825225,
      "epoch": 0.18258484654270632,
      "grad_norm": 9.14821720123291,
      "learning_rate": 5.920576552172171e-06,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8290189146995545,
      "num_tokens": 70833465.0,
      "step": 58900
    },
    {
      "entropy": 1.8775691345334053,
      "epoch": 0.18261584566775602,
      "grad_norm": 9.709592819213867,
      "learning_rate": 5.920074011997246e-06,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8243531733751297,
      "num_tokens": 70845723.0,
      "step": 58910
    },
    {
      "entropy": 1.8897256642580031,
      "epoch": 0.1826468447928057,
      "grad_norm": 8.386919975280762,
      "learning_rate": 5.919571599767849e-06,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8361333101987839,
      "num_tokens": 70858171.0,
      "step": 58920
    },
    {
      "entropy": 1.810855358839035,
      "epoch": 0.1826778439178554,
      "grad_norm": 7.87431001663208,
      "learning_rate": 5.919069315429698e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8387002602219582,
      "num_tokens": 70870960.0,
      "step": 58930
    },
    {
      "entropy": 1.812534037232399,
      "epoch": 0.1827088430429051,
      "grad_norm": 9.197452545166016,
      "learning_rate": 5.91856715892854e-06,
      "loss": 0.411,
      "mean_token_accuracy": 0.8409740015864372,
      "num_tokens": 70884301.0,
      "step": 58940
    },
    {
      "entropy": 1.8245099663734436,
      "epoch": 0.1827398421679548,
      "grad_norm": 3.566840410232544,
      "learning_rate": 5.918065130210162e-06,
      "loss": 0.435,
      "mean_token_accuracy": 0.8494891449809074,
      "num_tokens": 70897472.0,
      "step": 58950
    },
    {
      "entropy": 1.8804219841957093,
      "epoch": 0.1827708412930045,
      "grad_norm": 9.882930755615234,
      "learning_rate": 5.917563229220378e-06,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8421027392148972,
      "num_tokens": 70909937.0,
      "step": 58960
    },
    {
      "entropy": 1.8350219413638116,
      "epoch": 0.1828018404180542,
      "grad_norm": 5.378752708435059,
      "learning_rate": 5.917061455905032e-06,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8302841424942017,
      "num_tokens": 70922611.0,
      "step": 58970
    },
    {
      "entropy": 1.9114596903324128,
      "epoch": 0.1828328395431039,
      "grad_norm": 9.584024429321289,
      "learning_rate": 5.9165598102100065e-06,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8317157745361328,
      "num_tokens": 70934342.0,
      "step": 58980
    },
    {
      "entropy": 1.9586676806211472,
      "epoch": 0.1828638386681536,
      "grad_norm": 9.2459135055542,
      "learning_rate": 5.916058292081212e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8311464220285416,
      "num_tokens": 70945014.0,
      "step": 58990
    },
    {
      "entropy": 1.8728288680315017,
      "epoch": 0.18289483779320329,
      "grad_norm": 3.3124911785125732,
      "learning_rate": 5.91555690146459e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8396729901432991,
      "num_tokens": 70957244.0,
      "step": 59000
    },
    {
      "entropy": 1.8486259281635284,
      "epoch": 0.18292583691825298,
      "grad_norm": 7.671779155731201,
      "learning_rate": 5.9150556383061166e-06,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8414855524897575,
      "num_tokens": 70969181.0,
      "step": 59010
    },
    {
      "entropy": 1.8501243054866792,
      "epoch": 0.18295683604330268,
      "grad_norm": 7.360293865203857,
      "learning_rate": 5.914554502551802e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8381767392158508,
      "num_tokens": 70981480.0,
      "step": 59020
    },
    {
      "entropy": 1.9539805203676224,
      "epoch": 0.18298783516835238,
      "grad_norm": 8.94968032836914,
      "learning_rate": 5.91405349414768e-06,
      "loss": 0.6436,
      "mean_token_accuracy": 0.8138018161058426,
      "num_tokens": 70992634.0,
      "step": 59030
    },
    {
      "entropy": 1.7947911590337753,
      "epoch": 0.18301883429340207,
      "grad_norm": 8.198612213134766,
      "learning_rate": 5.913552613039827e-06,
      "loss": 0.4641,
      "mean_token_accuracy": 0.8452375203371048,
      "num_tokens": 71006127.0,
      "step": 59040
    },
    {
      "entropy": 1.837024737894535,
      "epoch": 0.18304983341845177,
      "grad_norm": 3.7017059326171875,
      "learning_rate": 5.913051859174345e-06,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8444261506199837,
      "num_tokens": 71018041.0,
      "step": 59050
    },
    {
      "entropy": 1.7747617959976196,
      "epoch": 0.18308083254350146,
      "grad_norm": 7.044408321380615,
      "learning_rate": 5.9125512324973685e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8478129550814628,
      "num_tokens": 71031917.0,
      "step": 59060
    },
    {
      "entropy": 1.8644176304340363,
      "epoch": 0.18311183166855116,
      "grad_norm": 9.228703498840332,
      "learning_rate": 5.9120507329550645e-06,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8401506051421166,
      "num_tokens": 71044161.0,
      "step": 59070
    },
    {
      "entropy": 1.8438007518649102,
      "epoch": 0.18314283079360086,
      "grad_norm": 4.598083019256592,
      "learning_rate": 5.9115503604936345e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8480850771069527,
      "num_tokens": 71057367.0,
      "step": 59080
    },
    {
      "entropy": 1.9214376494288445,
      "epoch": 0.18317382991865055,
      "grad_norm": 11.023812294006348,
      "learning_rate": 5.911050115059307e-06,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8322188884019852,
      "num_tokens": 71068645.0,
      "step": 59090
    },
    {
      "entropy": 1.9257762372493743,
      "epoch": 0.18320482904370025,
      "grad_norm": 7.3912129402160645,
      "learning_rate": 5.910549996598346e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8420733541250229,
      "num_tokens": 71079883.0,
      "step": 59100
    },
    {
      "entropy": 1.9468917578458786,
      "epoch": 0.18323582816874995,
      "grad_norm": 8.324677467346191,
      "learning_rate": 5.910050005057045e-06,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8290826484560967,
      "num_tokens": 71090907.0,
      "step": 59110
    },
    {
      "entropy": 1.930706176161766,
      "epoch": 0.18326682729379964,
      "grad_norm": 9.13598918914795,
      "learning_rate": 5.909550140381733e-06,
      "loss": 0.5705,
      "mean_token_accuracy": 0.8350569799542427,
      "num_tokens": 71101884.0,
      "step": 59120
    },
    {
      "entropy": 1.8789242595434188,
      "epoch": 0.18329782641884934,
      "grad_norm": 10.5730619430542,
      "learning_rate": 5.9090504025187655e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8466743901371956,
      "num_tokens": 71113590.0,
      "step": 59130
    },
    {
      "entropy": 1.901440866291523,
      "epoch": 0.183328825543899,
      "grad_norm": 12.086167335510254,
      "learning_rate": 5.908550791414533e-06,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8337482139468193,
      "num_tokens": 71125988.0,
      "step": 59140
    },
    {
      "entropy": 1.8326066508889198,
      "epoch": 0.1833598246689487,
      "grad_norm": 2.8919966220855713,
      "learning_rate": 5.908051307015459e-06,
      "loss": 0.4791,
      "mean_token_accuracy": 0.844254644215107,
      "num_tokens": 71139896.0,
      "step": 59150
    },
    {
      "entropy": 1.9314706429839135,
      "epoch": 0.1833908237939984,
      "grad_norm": 8.493422508239746,
      "learning_rate": 5.907551949267995e-06,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8385980620980262,
      "num_tokens": 71151553.0,
      "step": 59160
    },
    {
      "entropy": 1.8806381478905678,
      "epoch": 0.1834218229190481,
      "grad_norm": 8.074383735656738,
      "learning_rate": 5.907052718118627e-06,
      "loss": 0.4823,
      "mean_token_accuracy": 0.8440838649868965,
      "num_tokens": 71163731.0,
      "step": 59170
    },
    {
      "entropy": 1.845897839963436,
      "epoch": 0.1834528220440978,
      "grad_norm": 10.374253273010254,
      "learning_rate": 5.9065536135138725e-06,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8441713094711304,
      "num_tokens": 71175735.0,
      "step": 59180
    },
    {
      "entropy": 1.9331976994872093,
      "epoch": 0.1834838211691475,
      "grad_norm": 8.653754234313965,
      "learning_rate": 5.906054635400278e-06,
      "loss": 0.5448,
      "mean_token_accuracy": 0.834554848074913,
      "num_tokens": 71187616.0,
      "step": 59190
    },
    {
      "entropy": 1.8752469643950462,
      "epoch": 0.1835148202941972,
      "grad_norm": 9.133798599243164,
      "learning_rate": 5.905555783724424e-06,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8436867401003838,
      "num_tokens": 71199959.0,
      "step": 59200
    },
    {
      "entropy": 1.8521633878350259,
      "epoch": 0.1835458194192469,
      "grad_norm": 8.649680137634277,
      "learning_rate": 5.905057058432922e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.842129784822464,
      "num_tokens": 71212099.0,
      "step": 59210
    },
    {
      "entropy": 1.8513922914862633,
      "epoch": 0.18357681854429658,
      "grad_norm": 9.206934928894043,
      "learning_rate": 5.904558459472417e-06,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8332608088850975,
      "num_tokens": 71224564.0,
      "step": 59220
    },
    {
      "entropy": 1.8147864386439323,
      "epoch": 0.18360781766934628,
      "grad_norm": 7.711428165435791,
      "learning_rate": 5.904059986789582e-06,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8427559122443199,
      "num_tokens": 71237203.0,
      "step": 59230
    },
    {
      "entropy": 1.8857770830392837,
      "epoch": 0.18363881679439598,
      "grad_norm": 8.450910568237305,
      "learning_rate": 5.903561640331122e-06,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8380788192152977,
      "num_tokens": 71249147.0,
      "step": 59240
    },
    {
      "entropy": 1.8396921649575233,
      "epoch": 0.18366981591944567,
      "grad_norm": 2.8776183128356934,
      "learning_rate": 5.9030634200437765e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8463966697454453,
      "num_tokens": 71260897.0,
      "step": 59250
    },
    {
      "entropy": 1.8520937889814377,
      "epoch": 0.18370081504449537,
      "grad_norm": 6.755105018615723,
      "learning_rate": 5.902565325874313e-06,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8342436105012894,
      "num_tokens": 71273021.0,
      "step": 59260
    },
    {
      "entropy": 1.8717355713248254,
      "epoch": 0.18373181416954507,
      "grad_norm": 8.250826835632324,
      "learning_rate": 5.902067357769535e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8344789668917656,
      "num_tokens": 71285282.0,
      "step": 59270
    },
    {
      "entropy": 1.911249254643917,
      "epoch": 0.18376281329459476,
      "grad_norm": 11.614958763122559,
      "learning_rate": 5.901569515676272e-06,
      "loss": 0.5307,
      "mean_token_accuracy": 0.839310847222805,
      "num_tokens": 71297503.0,
      "step": 59280
    },
    {
      "entropy": 1.9195268914103507,
      "epoch": 0.18379381241964446,
      "grad_norm": 8.41645336151123,
      "learning_rate": 5.901071799541385e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8358474299311638,
      "num_tokens": 71309385.0,
      "step": 59290
    },
    {
      "entropy": 1.90712169110775,
      "epoch": 0.18382481154469416,
      "grad_norm": 8.294679641723633,
      "learning_rate": 5.900574209311775e-06,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8497764050960541,
      "num_tokens": 71320129.0,
      "step": 59300
    },
    {
      "entropy": 1.935383716225624,
      "epoch": 0.18385581066974385,
      "grad_norm": 9.08409309387207,
      "learning_rate": 5.900076744934362e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8406318157911301,
      "num_tokens": 71331528.0,
      "step": 59310
    },
    {
      "entropy": 1.9089769005775452,
      "epoch": 0.18388680979479355,
      "grad_norm": 12.19558048248291,
      "learning_rate": 5.899579406356107e-06,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8409374266862869,
      "num_tokens": 71342190.0,
      "step": 59320
    },
    {
      "entropy": 1.9273348927497864,
      "epoch": 0.18391780891984325,
      "grad_norm": 10.214935302734375,
      "learning_rate": 5.8990821935239975e-06,
      "loss": 0.5118,
      "mean_token_accuracy": 0.8436912566423416,
      "num_tokens": 71353668.0,
      "step": 59330
    },
    {
      "entropy": 1.9410814076662064,
      "epoch": 0.18394880804489294,
      "grad_norm": 9.362590789794922,
      "learning_rate": 5.898585106385053e-06,
      "loss": 0.5704,
      "mean_token_accuracy": 0.8309288114309311,
      "num_tokens": 71364044.0,
      "step": 59340
    },
    {
      "entropy": 1.9330678835511208,
      "epoch": 0.18397980716994264,
      "grad_norm": 7.387380599975586,
      "learning_rate": 5.898088144886326e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8400380194187165,
      "num_tokens": 71375184.0,
      "step": 59350
    },
    {
      "entropy": 1.9176151275634765,
      "epoch": 0.18401080629499234,
      "grad_norm": 8.243695259094238,
      "learning_rate": 5.897591308974896e-06,
      "loss": 0.5476,
      "mean_token_accuracy": 0.8329030573368073,
      "num_tokens": 71386183.0,
      "step": 59360
    },
    {
      "entropy": 1.7748358353972435,
      "epoch": 0.18404180542004203,
      "grad_norm": 4.477342128753662,
      "learning_rate": 5.897094598597879e-06,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8530562177300454,
      "num_tokens": 71399368.0,
      "step": 59370
    },
    {
      "entropy": 1.8144743472337723,
      "epoch": 0.18407280454509173,
      "grad_norm": 9.520360946655273,
      "learning_rate": 5.896598013702419e-06,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8422923222184181,
      "num_tokens": 71412727.0,
      "step": 59380
    },
    {
      "entropy": 1.8014607191085816,
      "epoch": 0.1841038036701414,
      "grad_norm": 6.801212787628174,
      "learning_rate": 5.8961015542356925e-06,
      "loss": 0.3999,
      "mean_token_accuracy": 0.8570068180561066,
      "num_tokens": 71425857.0,
      "step": 59390
    },
    {
      "entropy": 1.927448120713234,
      "epoch": 0.1841348027951911,
      "grad_norm": 8.77550220489502,
      "learning_rate": 5.895605220144907e-06,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8222044169902801,
      "num_tokens": 71436727.0,
      "step": 59400
    },
    {
      "entropy": 1.9893068850040436,
      "epoch": 0.1841658019202408,
      "grad_norm": 9.02739143371582,
      "learning_rate": 5.8951090113772976e-06,
      "loss": 0.5979,
      "mean_token_accuracy": 0.8225637242197991,
      "num_tokens": 71447373.0,
      "step": 59410
    },
    {
      "entropy": 1.8202601961791516,
      "epoch": 0.1841968010452905,
      "grad_norm": 8.421029090881348,
      "learning_rate": 5.894612927880137e-06,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8546611994504929,
      "num_tokens": 71460105.0,
      "step": 59420
    },
    {
      "entropy": 1.8888828486204148,
      "epoch": 0.18422780017034018,
      "grad_norm": 9.308785438537598,
      "learning_rate": 5.8941169696007235e-06,
      "loss": 0.53,
      "mean_token_accuracy": 0.8378401637077332,
      "num_tokens": 71472520.0,
      "step": 59430
    },
    {
      "entropy": 1.8498116791248322,
      "epoch": 0.18425879929538988,
      "grad_norm": 8.532123565673828,
      "learning_rate": 5.893621136486389e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.844611656665802,
      "num_tokens": 71484511.0,
      "step": 59440
    },
    {
      "entropy": 1.8172200858592986,
      "epoch": 0.18428979842043958,
      "grad_norm": 4.122251033782959,
      "learning_rate": 5.893125428484495e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8459887281060219,
      "num_tokens": 71496637.0,
      "step": 59450
    },
    {
      "entropy": 1.8752017796039582,
      "epoch": 0.18432079754548927,
      "grad_norm": 9.73798942565918,
      "learning_rate": 5.892629845542437e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8463060513138772,
      "num_tokens": 71508739.0,
      "step": 59460
    },
    {
      "entropy": 1.8498885072767735,
      "epoch": 0.18435179667053897,
      "grad_norm": 8.824564933776855,
      "learning_rate": 5.892134387607636e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.835616897046566,
      "num_tokens": 71521174.0,
      "step": 59470
    },
    {
      "entropy": 1.863853606581688,
      "epoch": 0.18438279579558867,
      "grad_norm": 9.248029708862305,
      "learning_rate": 5.891639054627551e-06,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8358003750443459,
      "num_tokens": 71533111.0,
      "step": 59480
    },
    {
      "entropy": 1.9517934620380402,
      "epoch": 0.18441379492063836,
      "grad_norm": 9.871654510498047,
      "learning_rate": 5.891143846549664e-06,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8294711738824845,
      "num_tokens": 71544244.0,
      "step": 59490
    },
    {
      "entropy": 1.8825423762202262,
      "epoch": 0.18444479404568806,
      "grad_norm": 7.173405170440674,
      "learning_rate": 5.890648763321495e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8478507623076439,
      "num_tokens": 71555732.0,
      "step": 59500
    },
    {
      "entropy": 1.9470236003398895,
      "epoch": 0.18447579317073776,
      "grad_norm": 10.64505672454834,
      "learning_rate": 5.8901538048905904e-06,
      "loss": 0.5637,
      "mean_token_accuracy": 0.8279279932379723,
      "num_tokens": 71566717.0,
      "step": 59510
    },
    {
      "entropy": 1.8414243504405021,
      "epoch": 0.18450679229578745,
      "grad_norm": 8.195703506469727,
      "learning_rate": 5.8896589712045306e-06,
      "loss": 0.4448,
      "mean_token_accuracy": 0.8501287072896957,
      "num_tokens": 71578437.0,
      "step": 59520
    },
    {
      "entropy": 1.8852709889411927,
      "epoch": 0.18453779142083715,
      "grad_norm": 10.737709045410156,
      "learning_rate": 5.8891642622109235e-06,
      "loss": 0.4645,
      "mean_token_accuracy": 0.831559669971466,
      "num_tokens": 71590564.0,
      "step": 59530
    },
    {
      "entropy": 1.8008535578846931,
      "epoch": 0.18456879054588685,
      "grad_norm": 12.071378707885742,
      "learning_rate": 5.888669677857409e-06,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8527644023299217,
      "num_tokens": 71604258.0,
      "step": 59540
    },
    {
      "entropy": 1.8511207103729248,
      "epoch": 0.18459978967093654,
      "grad_norm": 4.7530012130737305,
      "learning_rate": 5.88817521809166e-06,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8385688096284867,
      "num_tokens": 71616859.0,
      "step": 59550
    },
    {
      "entropy": 1.9494419798254967,
      "epoch": 0.18463078879598624,
      "grad_norm": 10.362981796264648,
      "learning_rate": 5.887680882861378e-06,
      "loss": 0.537,
      "mean_token_accuracy": 0.833903931081295,
      "num_tokens": 71628020.0,
      "step": 59560
    },
    {
      "entropy": 1.9751142218708992,
      "epoch": 0.18466178792103594,
      "grad_norm": 10.168465614318848,
      "learning_rate": 5.887186672114294e-06,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8342203125357628,
      "num_tokens": 71639405.0,
      "step": 59570
    },
    {
      "entropy": 1.9470524042844772,
      "epoch": 0.18469278704608563,
      "grad_norm": 7.566704273223877,
      "learning_rate": 5.886692585798173e-06,
      "loss": 0.5758,
      "mean_token_accuracy": 0.8385176613926888,
      "num_tokens": 71650529.0,
      "step": 59580
    },
    {
      "entropy": 1.8561659947037696,
      "epoch": 0.18472378617113533,
      "grad_norm": 8.916204452514648,
      "learning_rate": 5.886198623860807e-06,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8419315591454506,
      "num_tokens": 71663274.0,
      "step": 59590
    },
    {
      "entropy": 1.9323371350765228,
      "epoch": 0.18475478529618503,
      "grad_norm": 10.571834564208984,
      "learning_rate": 5.8857047862500226e-06,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8282590746879578,
      "num_tokens": 71674739.0,
      "step": 59600
    },
    {
      "entropy": 1.9620111271739007,
      "epoch": 0.18478578442123472,
      "grad_norm": 8.114259719848633,
      "learning_rate": 5.885211072913676e-06,
      "loss": 0.5633,
      "mean_token_accuracy": 0.8262233078479767,
      "num_tokens": 71685925.0,
      "step": 59610
    },
    {
      "entropy": 1.8970046997070313,
      "epoch": 0.18481678354628442,
      "grad_norm": 3.8900442123413086,
      "learning_rate": 5.884717483799649e-06,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8467764779925346,
      "num_tokens": 71697604.0,
      "step": 59620
    },
    {
      "entropy": 1.8912265598773956,
      "epoch": 0.18484778267133412,
      "grad_norm": 9.921353340148926,
      "learning_rate": 5.884224018855862e-06,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8240482151508332,
      "num_tokens": 71708748.0,
      "step": 59630
    },
    {
      "entropy": 1.873260524868965,
      "epoch": 0.18487878179638378,
      "grad_norm": 7.441627025604248,
      "learning_rate": 5.883730678030261e-06,
      "loss": 0.5304,
      "mean_token_accuracy": 0.831610731780529,
      "num_tokens": 71720631.0,
      "step": 59640
    },
    {
      "entropy": 1.8897306516766548,
      "epoch": 0.18490978092143348,
      "grad_norm": 8.014190673828125,
      "learning_rate": 5.883237461270822e-06,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8387016162276268,
      "num_tokens": 71732006.0,
      "step": 59650
    },
    {
      "entropy": 1.8416223376989365,
      "epoch": 0.18494078004648318,
      "grad_norm": 6.837216854095459,
      "learning_rate": 5.882744368525556e-06,
      "loss": 0.5324,
      "mean_token_accuracy": 0.832354225218296,
      "num_tokens": 71744288.0,
      "step": 59660
    },
    {
      "entropy": 1.7982579827308656,
      "epoch": 0.18497177917153287,
      "grad_norm": 9.943193435668945,
      "learning_rate": 5.882251399742499e-06,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8525391161441803,
      "num_tokens": 71757879.0,
      "step": 59670
    },
    {
      "entropy": 1.7894408702850342,
      "epoch": 0.18500277829658257,
      "grad_norm": 8.48753833770752,
      "learning_rate": 5.881758554869721e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8433696299791336,
      "num_tokens": 71772009.0,
      "step": 59680
    },
    {
      "entropy": 1.8790398433804512,
      "epoch": 0.18503377742163227,
      "grad_norm": 8.779352188110352,
      "learning_rate": 5.881265833855321e-06,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8341058731079102,
      "num_tokens": 71784020.0,
      "step": 59690
    },
    {
      "entropy": 1.8383540600538253,
      "epoch": 0.18506477654668196,
      "grad_norm": 4.482400894165039,
      "learning_rate": 5.880773236647431e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8363133147358894,
      "num_tokens": 71795570.0,
      "step": 59700
    },
    {
      "entropy": 1.807381245493889,
      "epoch": 0.18509577567173166,
      "grad_norm": 7.861814498901367,
      "learning_rate": 5.8802807631942095e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8491622850298881,
      "num_tokens": 71808715.0,
      "step": 59710
    },
    {
      "entropy": 1.9171263113617898,
      "epoch": 0.18512677479678136,
      "grad_norm": 9.038065910339355,
      "learning_rate": 5.879788413443846e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8278371721506119,
      "num_tokens": 71820652.0,
      "step": 59720
    },
    {
      "entropy": 1.9229667693376542,
      "epoch": 0.18515777392183105,
      "grad_norm": 7.926364898681641,
      "learning_rate": 5.879296187344564e-06,
      "loss": 0.5683,
      "mean_token_accuracy": 0.8299353420734406,
      "num_tokens": 71832006.0,
      "step": 59730
    },
    {
      "entropy": 1.8270016878843307,
      "epoch": 0.18518877304688075,
      "grad_norm": 4.967191219329834,
      "learning_rate": 5.878804084844616e-06,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8319668889045715,
      "num_tokens": 71845433.0,
      "step": 59740
    },
    {
      "entropy": 1.9323490008711814,
      "epoch": 0.18521977217193045,
      "grad_norm": 7.835849285125732,
      "learning_rate": 5.878312105892281e-06,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8331502959132194,
      "num_tokens": 71856737.0,
      "step": 59750
    },
    {
      "entropy": 1.7635822594165802,
      "epoch": 0.18525077129698014,
      "grad_norm": 10.747203826904297,
      "learning_rate": 5.8778202504358716e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8548522099852562,
      "num_tokens": 71870471.0,
      "step": 59760
    },
    {
      "entropy": 1.7823803335428239,
      "epoch": 0.18528177042202984,
      "grad_norm": 3.434664249420166,
      "learning_rate": 5.8773285184237316e-06,
      "loss": 0.426,
      "mean_token_accuracy": 0.846599979698658,
      "num_tokens": 71884046.0,
      "step": 59770
    },
    {
      "entropy": 1.8083126276731492,
      "epoch": 0.18531276954707954,
      "grad_norm": 4.187022686004639,
      "learning_rate": 5.876836909804231e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8312985822558403,
      "num_tokens": 71897333.0,
      "step": 59780
    },
    {
      "entropy": 1.9598889589309691,
      "epoch": 0.18534376867212923,
      "grad_norm": 8.239096641540527,
      "learning_rate": 5.876345424525776e-06,
      "loss": 0.6611,
      "mean_token_accuracy": 0.8133495360612869,
      "num_tokens": 71908395.0,
      "step": 59790
    },
    {
      "entropy": 1.9233538269996644,
      "epoch": 0.18537476779717893,
      "grad_norm": 3.7987241744995117,
      "learning_rate": 5.8758540625367965e-06,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8402821734547615,
      "num_tokens": 71920345.0,
      "step": 59800
    },
    {
      "entropy": 1.9380876675248147,
      "epoch": 0.18540576692222863,
      "grad_norm": 7.9834208488464355,
      "learning_rate": 5.875362823785758e-06,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8361145481467247,
      "num_tokens": 71932372.0,
      "step": 59810
    },
    {
      "entropy": 1.9172768041491508,
      "epoch": 0.18543676604727832,
      "grad_norm": 10.48983097076416,
      "learning_rate": 5.8748717082211516e-06,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8384231805801392,
      "num_tokens": 71944733.0,
      "step": 59820
    },
    {
      "entropy": 1.8952128738164902,
      "epoch": 0.18546776517232802,
      "grad_norm": 6.812406539916992,
      "learning_rate": 5.874380715791502e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.846713088452816,
      "num_tokens": 71956856.0,
      "step": 59830
    },
    {
      "entropy": 1.780382940173149,
      "epoch": 0.18549876429737772,
      "grad_norm": 4.57772159576416,
      "learning_rate": 5.8738898464453644e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8531007096171379,
      "num_tokens": 71970503.0,
      "step": 59840
    },
    {
      "entropy": 1.910321943461895,
      "epoch": 0.1855297634224274,
      "grad_norm": 10.066534042358398,
      "learning_rate": 5.87339910013132e-06,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8341235533356667,
      "num_tokens": 71982378.0,
      "step": 59850
    },
    {
      "entropy": 1.927718922495842,
      "epoch": 0.1855607625474771,
      "grad_norm": 9.32041072845459,
      "learning_rate": 5.872908476797983e-06,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8251764044165611,
      "num_tokens": 71993478.0,
      "step": 59860
    },
    {
      "entropy": 1.9356975913047791,
      "epoch": 0.1855917616725268,
      "grad_norm": 4.789025783538818,
      "learning_rate": 5.872417976393997e-06,
      "loss": 0.5355,
      "mean_token_accuracy": 0.8277438759803772,
      "num_tokens": 72005113.0,
      "step": 59870
    },
    {
      "entropy": 1.8369565188884736,
      "epoch": 0.18562276079757647,
      "grad_norm": 7.20404577255249,
      "learning_rate": 5.871927598868036e-06,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8378855675458908,
      "num_tokens": 72017827.0,
      "step": 59880
    },
    {
      "entropy": 1.8494772240519524,
      "epoch": 0.18565375992262617,
      "grad_norm": 8.194475173950195,
      "learning_rate": 5.871437344168805e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8221006825566292,
      "num_tokens": 72030285.0,
      "step": 59890
    },
    {
      "entropy": 1.9028642885386944,
      "epoch": 0.18568475904767587,
      "grad_norm": 8.854440689086914,
      "learning_rate": 5.870947212245036e-06,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8278263434767723,
      "num_tokens": 72043015.0,
      "step": 59900
    },
    {
      "entropy": 1.8917630180716514,
      "epoch": 0.18571575817272556,
      "grad_norm": 8.142960548400879,
      "learning_rate": 5.870457203045496e-06,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8398408219218254,
      "num_tokens": 72054537.0,
      "step": 59910
    },
    {
      "entropy": 1.8536349281668663,
      "epoch": 0.18574675729777526,
      "grad_norm": 8.590570449829102,
      "learning_rate": 5.869967316518973e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8522145226597786,
      "num_tokens": 72067016.0,
      "step": 59920
    },
    {
      "entropy": 1.9314019471406936,
      "epoch": 0.18577775642282496,
      "grad_norm": 4.44788932800293,
      "learning_rate": 5.869477552614296e-06,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8373445898294449,
      "num_tokens": 72078617.0,
      "step": 59930
    },
    {
      "entropy": 1.8379703059792518,
      "epoch": 0.18580875554787465,
      "grad_norm": 8.737577438354492,
      "learning_rate": 5.868987911280315e-06,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8353394463658332,
      "num_tokens": 72090576.0,
      "step": 59940
    },
    {
      "entropy": 1.9440933406352996,
      "epoch": 0.18583975467292435,
      "grad_norm": 8.472302436828613,
      "learning_rate": 5.868498392465915e-06,
      "loss": 0.5734,
      "mean_token_accuracy": 0.8333403542637825,
      "num_tokens": 72101344.0,
      "step": 59950
    },
    {
      "entropy": 1.8688756361603738,
      "epoch": 0.18587075379797405,
      "grad_norm": 9.316754341125488,
      "learning_rate": 5.868008996120008e-06,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8337069198489189,
      "num_tokens": 72113227.0,
      "step": 59960
    },
    {
      "entropy": 1.834533803164959,
      "epoch": 0.18590175292302374,
      "grad_norm": 8.468538284301758,
      "learning_rate": 5.867519722191538e-06,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8497491136193276,
      "num_tokens": 72125271.0,
      "step": 59970
    },
    {
      "entropy": 1.7592488691210746,
      "epoch": 0.18593275204807344,
      "grad_norm": 8.312585830688477,
      "learning_rate": 5.867030570629477e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8604243487119675,
      "num_tokens": 72139001.0,
      "step": 59980
    },
    {
      "entropy": 1.9161117985844611,
      "epoch": 0.18596375117312314,
      "grad_norm": 4.209480285644531,
      "learning_rate": 5.866541541382829e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8296124190092087,
      "num_tokens": 72151289.0,
      "step": 59990
    },
    {
      "entropy": 1.9577097177505494,
      "epoch": 0.18599475029817283,
      "grad_norm": 9.150534629821777,
      "learning_rate": 5.866052634400624e-06,
      "loss": 0.58,
      "mean_token_accuracy": 0.8270249828696251,
      "num_tokens": 72161958.0,
      "step": 60000
    },
    {
      "entropy": 1.896253764629364,
      "epoch": 0.18602574942322253,
      "grad_norm": 5.07069206237793,
      "learning_rate": 5.865563849631925e-06,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8247905030846596,
      "num_tokens": 72175097.0,
      "step": 60010
    },
    {
      "entropy": 1.7840982712805271,
      "epoch": 0.18605674854827223,
      "grad_norm": 8.719794273376465,
      "learning_rate": 5.865075187025824e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8515463337302208,
      "num_tokens": 72188721.0,
      "step": 60020
    },
    {
      "entropy": 1.994665226340294,
      "epoch": 0.18608774767332192,
      "grad_norm": 8.010587692260742,
      "learning_rate": 5.864586646531443e-06,
      "loss": 0.5875,
      "mean_token_accuracy": 0.8243452772498131,
      "num_tokens": 72199455.0,
      "step": 60030
    },
    {
      "entropy": 1.869076819717884,
      "epoch": 0.18611874679837162,
      "grad_norm": 8.577561378479004,
      "learning_rate": 5.864098228097931e-06,
      "loss": 0.4793,
      "mean_token_accuracy": 0.843721067905426,
      "num_tokens": 72211035.0,
      "step": 60040
    },
    {
      "entropy": 1.9081630438566208,
      "epoch": 0.18614974592342132,
      "grad_norm": 8.139861106872559,
      "learning_rate": 5.863609931674471e-06,
      "loss": 0.4883,
      "mean_token_accuracy": 0.847606098651886,
      "num_tokens": 72222707.0,
      "step": 60050
    },
    {
      "entropy": 1.9403681397438048,
      "epoch": 0.186180745048471,
      "grad_norm": 9.84714412689209,
      "learning_rate": 5.8631217572102716e-06,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8385643243789673,
      "num_tokens": 72232917.0,
      "step": 60060
    },
    {
      "entropy": 1.9238607689738274,
      "epoch": 0.1862117441735207,
      "grad_norm": 8.912215232849121,
      "learning_rate": 5.862633704654573e-06,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8306085541844368,
      "num_tokens": 72244516.0,
      "step": 60070
    },
    {
      "entropy": 1.9544622465968131,
      "epoch": 0.1862427432985704,
      "grad_norm": 8.447044372558594,
      "learning_rate": 5.862145773956647e-06,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8342535004019738,
      "num_tokens": 72255606.0,
      "step": 60080
    },
    {
      "entropy": 1.8915927335619926,
      "epoch": 0.1862737424236201,
      "grad_norm": 9.219244003295898,
      "learning_rate": 5.861657965065788e-06,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8286745131015778,
      "num_tokens": 72267179.0,
      "step": 60090
    },
    {
      "entropy": 1.9081023514270783,
      "epoch": 0.1863047415486698,
      "grad_norm": 3.8178694248199463,
      "learning_rate": 5.861170277931328e-06,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8304624617099762,
      "num_tokens": 72279385.0,
      "step": 60100
    },
    {
      "entropy": 1.8515421822667122,
      "epoch": 0.1863357406737195,
      "grad_norm": 6.892448902130127,
      "learning_rate": 5.8606827125026256e-06,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8440054371953011,
      "num_tokens": 72292367.0,
      "step": 60110
    },
    {
      "entropy": 1.941430465877056,
      "epoch": 0.1863667397987692,
      "grad_norm": 9.434157371520996,
      "learning_rate": 5.860195268729066e-06,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8339191779494286,
      "num_tokens": 72303613.0,
      "step": 60120
    },
    {
      "entropy": 1.9135318726301194,
      "epoch": 0.18639773892381886,
      "grad_norm": 11.788899421691895,
      "learning_rate": 5.859707946560067e-06,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8330492198467254,
      "num_tokens": 72314613.0,
      "step": 60130
    },
    {
      "entropy": 1.8537031307816505,
      "epoch": 0.18642873804886856,
      "grad_norm": 9.07436466217041,
      "learning_rate": 5.859220745945075e-06,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8319351330399514,
      "num_tokens": 72326917.0,
      "step": 60140
    },
    {
      "entropy": 2.000509098172188,
      "epoch": 0.18645973717391826,
      "grad_norm": 11.218595504760742,
      "learning_rate": 5.858733666833567e-06,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8274472177028656,
      "num_tokens": 72338272.0,
      "step": 60150
    },
    {
      "entropy": 1.9531633704900742,
      "epoch": 0.18649073629896795,
      "grad_norm": 8.319474220275879,
      "learning_rate": 5.858246709175047e-06,
      "loss": 0.5806,
      "mean_token_accuracy": 0.8167789533734322,
      "num_tokens": 72349717.0,
      "step": 60160
    },
    {
      "entropy": 1.8039245098829269,
      "epoch": 0.18652173542401765,
      "grad_norm": 5.849046230316162,
      "learning_rate": 5.8577598729190496e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8554789200425148,
      "num_tokens": 72362915.0,
      "step": 60170
    },
    {
      "entropy": 1.87906823605299,
      "epoch": 0.18655273454906734,
      "grad_norm": 8.501765251159668,
      "learning_rate": 5.85727315801514e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8379226326942444,
      "num_tokens": 72374922.0,
      "step": 60180
    },
    {
      "entropy": 1.8856094121932983,
      "epoch": 0.18658373367411704,
      "grad_norm": 8.660506248474121,
      "learning_rate": 5.8567865644129095e-06,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8400798097252846,
      "num_tokens": 72387230.0,
      "step": 60190
    },
    {
      "entropy": 1.843792949616909,
      "epoch": 0.18661473279916674,
      "grad_norm": 7.968148231506348,
      "learning_rate": 5.856300092061984e-06,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8412106573581696,
      "num_tokens": 72399496.0,
      "step": 60200
    },
    {
      "entropy": 1.863570548593998,
      "epoch": 0.18664573192421643,
      "grad_norm": 8.22105598449707,
      "learning_rate": 5.855813740912011e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8438234731554985,
      "num_tokens": 72412130.0,
      "step": 60210
    },
    {
      "entropy": 1.903498760610819,
      "epoch": 0.18667673104926613,
      "grad_norm": 8.098082542419434,
      "learning_rate": 5.855327510912675e-06,
      "loss": 0.487,
      "mean_token_accuracy": 0.8345488205552101,
      "num_tokens": 72425418.0,
      "step": 60220
    },
    {
      "entropy": 1.8605897799134254,
      "epoch": 0.18670773017431583,
      "grad_norm": 8.592905044555664,
      "learning_rate": 5.854841402013685e-06,
      "loss": 0.5436,
      "mean_token_accuracy": 0.8296092748641968,
      "num_tokens": 72437861.0,
      "step": 60230
    },
    {
      "entropy": 1.8651499822735786,
      "epoch": 0.18673872929936552,
      "grad_norm": 3.829653739929199,
      "learning_rate": 5.8543554141647814e-06,
      "loss": 0.4924,
      "mean_token_accuracy": 0.837422613799572,
      "num_tokens": 72450429.0,
      "step": 60240
    },
    {
      "entropy": 1.8985691770911217,
      "epoch": 0.18676972842441522,
      "grad_norm": 8.132059097290039,
      "learning_rate": 5.853869547315731e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8453919142484665,
      "num_tokens": 72462168.0,
      "step": 60250
    },
    {
      "entropy": 1.8822806507349015,
      "epoch": 0.18680072754946492,
      "grad_norm": 8.20763111114502,
      "learning_rate": 5.853383801416336e-06,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8344938531517982,
      "num_tokens": 72473992.0,
      "step": 60260
    },
    {
      "entropy": 1.9301311939954757,
      "epoch": 0.18683172667451461,
      "grad_norm": 8.083572387695312,
      "learning_rate": 5.8528981764164205e-06,
      "loss": 0.5355,
      "mean_token_accuracy": 0.841622045636177,
      "num_tokens": 72484567.0,
      "step": 60270
    },
    {
      "entropy": 1.8997415751218796,
      "epoch": 0.1868627257995643,
      "grad_norm": 7.791327953338623,
      "learning_rate": 5.85241267226584e-06,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8344151824712753,
      "num_tokens": 72495039.0,
      "step": 60280
    },
    {
      "entropy": 1.8864543735980988,
      "epoch": 0.186893724924614,
      "grad_norm": 7.940762042999268,
      "learning_rate": 5.851927288914482e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8393915817141533,
      "num_tokens": 72507012.0,
      "step": 60290
    },
    {
      "entropy": 1.8426005780696868,
      "epoch": 0.1869247240496637,
      "grad_norm": 8.484749794006348,
      "learning_rate": 5.85144202631226e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8417939618229866,
      "num_tokens": 72519213.0,
      "step": 60300
    },
    {
      "entropy": 1.860968105494976,
      "epoch": 0.1869557231747134,
      "grad_norm": 8.830284118652344,
      "learning_rate": 5.850956884409118e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.848306556046009,
      "num_tokens": 72531214.0,
      "step": 60310
    },
    {
      "entropy": 1.921893371641636,
      "epoch": 0.1869867222997631,
      "grad_norm": 4.533517837524414,
      "learning_rate": 5.8504718631550285e-06,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8360114693641663,
      "num_tokens": 72543240.0,
      "step": 60320
    },
    {
      "entropy": 1.93612492531538,
      "epoch": 0.1870177214248128,
      "grad_norm": 8.422826766967773,
      "learning_rate": 5.849986962499992e-06,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8314146876335144,
      "num_tokens": 72554502.0,
      "step": 60330
    },
    {
      "entropy": 1.887687975168228,
      "epoch": 0.1870487205498625,
      "grad_norm": 8.538383483886719,
      "learning_rate": 5.8495021823940416e-06,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8397644698619843,
      "num_tokens": 72565656.0,
      "step": 60340
    },
    {
      "entropy": 1.744446012377739,
      "epoch": 0.1870797196749122,
      "grad_norm": 8.830626487731934,
      "learning_rate": 5.849017522787233e-06,
      "loss": 0.512,
      "mean_token_accuracy": 0.8492047414183617,
      "num_tokens": 72579158.0,
      "step": 60350
    },
    {
      "entropy": 1.8573518849909305,
      "epoch": 0.18711071879996188,
      "grad_norm": 8.745538711547852,
      "learning_rate": 5.848532983629661e-06,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8321371108293534,
      "num_tokens": 72591425.0,
      "step": 60360
    },
    {
      "entropy": 1.7826851338148118,
      "epoch": 0.18714171792501158,
      "grad_norm": 8.527620315551758,
      "learning_rate": 5.848048564871437e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8485151454806328,
      "num_tokens": 72604518.0,
      "step": 60370
    },
    {
      "entropy": 1.8439786598086356,
      "epoch": 0.18717271705006125,
      "grad_norm": 8.76579475402832,
      "learning_rate": 5.84756426646271e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8471570834517479,
      "num_tokens": 72616133.0,
      "step": 60380
    },
    {
      "entropy": 1.9413674965500831,
      "epoch": 0.18720371617511095,
      "grad_norm": 9.924832344055176,
      "learning_rate": 5.847080088353656e-06,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8391248330473899,
      "num_tokens": 72627275.0,
      "step": 60390
    },
    {
      "entropy": 1.803322871029377,
      "epoch": 0.18723471530016064,
      "grad_norm": 7.1156816482543945,
      "learning_rate": 5.846596030494477e-06,
      "loss": 0.5098,
      "mean_token_accuracy": 0.831206327676773,
      "num_tokens": 72640255.0,
      "step": 60400
    },
    {
      "entropy": 1.9137111112475396,
      "epoch": 0.18726571442521034,
      "grad_norm": 9.406431198120117,
      "learning_rate": 5.8461120928354075e-06,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8288335934281349,
      "num_tokens": 72651695.0,
      "step": 60410
    },
    {
      "entropy": 1.988877174258232,
      "epoch": 0.18729671355026004,
      "grad_norm": 8.406907081604004,
      "learning_rate": 5.845628275326711e-06,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8310790181159973,
      "num_tokens": 72662701.0,
      "step": 60420
    },
    {
      "entropy": 1.844719012081623,
      "epoch": 0.18732771267530973,
      "grad_norm": 4.9071431159973145,
      "learning_rate": 5.845144577918675e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8463888078927994,
      "num_tokens": 72675062.0,
      "step": 60430
    },
    {
      "entropy": 1.9066783770918847,
      "epoch": 0.18735871180035943,
      "grad_norm": 9.20982551574707,
      "learning_rate": 5.844661000561621e-06,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8450051352381707,
      "num_tokens": 72687250.0,
      "step": 60440
    },
    {
      "entropy": 1.8505268752574922,
      "epoch": 0.18738971092540913,
      "grad_norm": 12.914068222045898,
      "learning_rate": 5.844177543205897e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8434465497732162,
      "num_tokens": 72699919.0,
      "step": 60450
    },
    {
      "entropy": 1.7514374867081641,
      "epoch": 0.18742071005045882,
      "grad_norm": 8.30848503112793,
      "learning_rate": 5.843694205801879e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8658733040094375,
      "num_tokens": 72713733.0,
      "step": 60460
    },
    {
      "entropy": 1.89697934538126,
      "epoch": 0.18745170917550852,
      "grad_norm": 9.627775192260742,
      "learning_rate": 5.843210988299973e-06,
      "loss": 0.5452,
      "mean_token_accuracy": 0.8306502267718315,
      "num_tokens": 72726134.0,
      "step": 60470
    },
    {
      "entropy": 1.9438308894634246,
      "epoch": 0.18748270830055822,
      "grad_norm": 10.055462837219238,
      "learning_rate": 5.842727890650615e-06,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8255834862589836,
      "num_tokens": 72736983.0,
      "step": 60480
    },
    {
      "entropy": 1.8450178682804108,
      "epoch": 0.1875137074256079,
      "grad_norm": 9.322096824645996,
      "learning_rate": 5.8422449128042654e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.845012141764164,
      "num_tokens": 72749747.0,
      "step": 60490
    },
    {
      "entropy": 1.8929164052009582,
      "epoch": 0.1875447065506576,
      "grad_norm": 3.979761838912964,
      "learning_rate": 5.84176205471142e-06,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8384419724345207,
      "num_tokens": 72761568.0,
      "step": 60500
    },
    {
      "entropy": 1.8335536420345306,
      "epoch": 0.1875757056757073,
      "grad_norm": 7.59586238861084,
      "learning_rate": 5.841279316322594e-06,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8467094838619232,
      "num_tokens": 72774261.0,
      "step": 60510
    },
    {
      "entropy": 1.882728861272335,
      "epoch": 0.187606704800757,
      "grad_norm": 8.7779541015625,
      "learning_rate": 5.840796697588341e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8436086297035217,
      "num_tokens": 72786714.0,
      "step": 60520
    },
    {
      "entropy": 1.8310616135597229,
      "epoch": 0.1876377039258067,
      "grad_norm": 5.086270809173584,
      "learning_rate": 5.840314198459236e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8549797222018242,
      "num_tokens": 72799986.0,
      "step": 60530
    },
    {
      "entropy": 1.9134382233023643,
      "epoch": 0.1876687030508564,
      "grad_norm": 8.873359680175781,
      "learning_rate": 5.839831818885886e-06,
      "loss": 0.5306,
      "mean_token_accuracy": 0.8400856927037239,
      "num_tokens": 72811382.0,
      "step": 60540
    },
    {
      "entropy": 1.809613211452961,
      "epoch": 0.1876997021759061,
      "grad_norm": 4.331747055053711,
      "learning_rate": 5.839349558818926e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8486540347337723,
      "num_tokens": 72824270.0,
      "step": 60550
    },
    {
      "entropy": 1.908320277929306,
      "epoch": 0.1877307013009558,
      "grad_norm": 8.491447448730469,
      "learning_rate": 5.838867418209019e-06,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8320151507854462,
      "num_tokens": 72836395.0,
      "step": 60560
    },
    {
      "entropy": 1.8660949409008025,
      "epoch": 0.18776170042600548,
      "grad_norm": 4.656119346618652,
      "learning_rate": 5.838385397006855e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8481364041566849,
      "num_tokens": 72849113.0,
      "step": 60570
    },
    {
      "entropy": 1.8910485938191415,
      "epoch": 0.18779269955105518,
      "grad_norm": 6.919294834136963,
      "learning_rate": 5.837903495163157e-06,
      "loss": 0.489,
      "mean_token_accuracy": 0.8406860768795014,
      "num_tokens": 72860770.0,
      "step": 60580
    },
    {
      "entropy": 1.9625378847122192,
      "epoch": 0.18782369867610488,
      "grad_norm": 10.646431922912598,
      "learning_rate": 5.837421712628675e-06,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8330793261528016,
      "num_tokens": 72871585.0,
      "step": 60590
    },
    {
      "entropy": 1.8914886653423308,
      "epoch": 0.18785469780115457,
      "grad_norm": 7.903103828430176,
      "learning_rate": 5.836940049354182e-06,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8413411438465118,
      "num_tokens": 72883535.0,
      "step": 60600
    },
    {
      "entropy": 1.8991345196962357,
      "epoch": 0.18788569692620427,
      "grad_norm": 4.898101329803467,
      "learning_rate": 5.8364585052904845e-06,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8374688118696213,
      "num_tokens": 72895721.0,
      "step": 60610
    },
    {
      "entropy": 1.8644976392388344,
      "epoch": 0.18791669605125394,
      "grad_norm": 4.19233512878418,
      "learning_rate": 5.8359770803884175e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.8408412501215935,
      "num_tokens": 72907421.0,
      "step": 60620
    },
    {
      "entropy": 1.891925212740898,
      "epoch": 0.18794769517630364,
      "grad_norm": 7.999198913574219,
      "learning_rate": 5.835495774598844e-06,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8336058259010315,
      "num_tokens": 72919575.0,
      "step": 60630
    },
    {
      "entropy": 1.8591307654976845,
      "epoch": 0.18797869430135333,
      "grad_norm": 8.912029266357422,
      "learning_rate": 5.8350145878726545e-06,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8295843943953514,
      "num_tokens": 72932425.0,
      "step": 60640
    },
    {
      "entropy": 1.957919180393219,
      "epoch": 0.18800969342640303,
      "grad_norm": 8.599974632263184,
      "learning_rate": 5.834533520160769e-06,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8301822498440743,
      "num_tokens": 72943322.0,
      "step": 60650
    },
    {
      "entropy": 1.9159081429243088,
      "epoch": 0.18804069255145273,
      "grad_norm": 7.521796226501465,
      "learning_rate": 5.834052571414132e-06,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8309228405356407,
      "num_tokens": 72954379.0,
      "step": 60660
    },
    {
      "entropy": 1.8881125912070273,
      "epoch": 0.18807169167650242,
      "grad_norm": 8.066596984863281,
      "learning_rate": 5.833571741583721e-06,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8445821896195411,
      "num_tokens": 72965758.0,
      "step": 60670
    },
    {
      "entropy": 1.8443811774253844,
      "epoch": 0.18810269080155212,
      "grad_norm": 9.628620147705078,
      "learning_rate": 5.8330910306205405e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8472337678074837,
      "num_tokens": 72977927.0,
      "step": 60680
    },
    {
      "entropy": 1.869968481361866,
      "epoch": 0.18813368992660182,
      "grad_norm": 8.748429298400879,
      "learning_rate": 5.832610438475622e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8413064509630204,
      "num_tokens": 72989566.0,
      "step": 60690
    },
    {
      "entropy": 1.881579375267029,
      "epoch": 0.1881646890516515,
      "grad_norm": 8.547125816345215,
      "learning_rate": 5.832129965100026e-06,
      "loss": 0.5417,
      "mean_token_accuracy": 0.8311194002628326,
      "num_tokens": 73000552.0,
      "step": 60700
    },
    {
      "entropy": 1.9194360464811324,
      "epoch": 0.1881956881767012,
      "grad_norm": 10.575980186462402,
      "learning_rate": 5.831649610444842e-06,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8376916989684104,
      "num_tokens": 73011919.0,
      "step": 60710
    },
    {
      "entropy": 1.936641664803028,
      "epoch": 0.1882266873017509,
      "grad_norm": 7.608135223388672,
      "learning_rate": 5.831169374461185e-06,
      "loss": 0.5898,
      "mean_token_accuracy": 0.821084663271904,
      "num_tokens": 73023762.0,
      "step": 60720
    },
    {
      "entropy": 1.9660951375961304,
      "epoch": 0.1882576864268006,
      "grad_norm": 8.96996784210205,
      "learning_rate": 5.8306892571002025e-06,
      "loss": 0.5683,
      "mean_token_accuracy": 0.835235033929348,
      "num_tokens": 73034510.0,
      "step": 60730
    },
    {
      "entropy": 1.8755255818367005,
      "epoch": 0.1882886855518503,
      "grad_norm": 8.350106239318848,
      "learning_rate": 5.830209258313067e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8453781142830848,
      "num_tokens": 73046102.0,
      "step": 60740
    },
    {
      "entropy": 1.901422207057476,
      "epoch": 0.1883196846769,
      "grad_norm": 8.605900764465332,
      "learning_rate": 5.829729378050978e-06,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8291002050042152,
      "num_tokens": 73057508.0,
      "step": 60750
    },
    {
      "entropy": 1.9229997634887694,
      "epoch": 0.1883506838019497,
      "grad_norm": 11.104288101196289,
      "learning_rate": 5.829249616265167e-06,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8307598143815994,
      "num_tokens": 73068665.0,
      "step": 60760
    },
    {
      "entropy": 1.8038654774427414,
      "epoch": 0.1883816829269994,
      "grad_norm": 9.37712287902832,
      "learning_rate": 5.828769972906891e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8444581896066665,
      "num_tokens": 73081347.0,
      "step": 60770
    },
    {
      "entropy": 1.8795530632138253,
      "epoch": 0.18841268205204909,
      "grad_norm": 4.818871021270752,
      "learning_rate": 5.828290447927437e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8303244650363922,
      "num_tokens": 73093545.0,
      "step": 60780
    },
    {
      "entropy": 1.8115891009569167,
      "epoch": 0.18844368117709878,
      "grad_norm": 8.035147666931152,
      "learning_rate": 5.827811041278115e-06,
      "loss": 0.426,
      "mean_token_accuracy": 0.8465625256299972,
      "num_tokens": 73106680.0,
      "step": 60790
    },
    {
      "entropy": 1.9304178357124329,
      "epoch": 0.18847468030214848,
      "grad_norm": 9.54056167602539,
      "learning_rate": 5.827331752910272e-06,
      "loss": 0.5425,
      "mean_token_accuracy": 0.839380769431591,
      "num_tokens": 73117528.0,
      "step": 60800
    },
    {
      "entropy": 1.9495259881019593,
      "epoch": 0.18850567942719818,
      "grad_norm": 9.372381210327148,
      "learning_rate": 5.826852582775273e-06,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8166373401880265,
      "num_tokens": 73128356.0,
      "step": 60810
    },
    {
      "entropy": 1.8412368908524512,
      "epoch": 0.18853667855224787,
      "grad_norm": 7.743039608001709,
      "learning_rate": 5.826373530824517e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8510013222694397,
      "num_tokens": 73140590.0,
      "step": 60820
    },
    {
      "entropy": 1.7964115411043167,
      "epoch": 0.18856767767729757,
      "grad_norm": 8.170472145080566,
      "learning_rate": 5.825894597009432e-06,
      "loss": 0.468,
      "mean_token_accuracy": 0.8488139227032662,
      "num_tokens": 73153546.0,
      "step": 60830
    },
    {
      "entropy": 1.755212776362896,
      "epoch": 0.18859867680234726,
      "grad_norm": 7.8742451667785645,
      "learning_rate": 5.825415781281467e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8522952437400818,
      "num_tokens": 73166595.0,
      "step": 60840
    },
    {
      "entropy": 1.8175699785351753,
      "epoch": 0.18862967592739696,
      "grad_norm": 8.804338455200195,
      "learning_rate": 5.824937083592109e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.8330580577254295,
      "num_tokens": 73178927.0,
      "step": 60850
    },
    {
      "entropy": 1.8207954451441766,
      "epoch": 0.18866067505244666,
      "grad_norm": 8.167973518371582,
      "learning_rate": 5.824458503892864e-06,
      "loss": 0.4969,
      "mean_token_accuracy": 0.8418833702802658,
      "num_tokens": 73190925.0,
      "step": 60860
    },
    {
      "entropy": 1.8893669456243516,
      "epoch": 0.18869167417749633,
      "grad_norm": 8.711833953857422,
      "learning_rate": 5.82398004213527e-06,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8303925588726997,
      "num_tokens": 73202125.0,
      "step": 60870
    },
    {
      "entropy": 1.8512268111109733,
      "epoch": 0.18872267330254602,
      "grad_norm": 7.86924409866333,
      "learning_rate": 5.823501698270892e-06,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8379854589700699,
      "num_tokens": 73214492.0,
      "step": 60880
    },
    {
      "entropy": 1.8381731614470482,
      "epoch": 0.18875367242759572,
      "grad_norm": 6.331354141235352,
      "learning_rate": 5.8230234722513236e-06,
      "loss": 0.518,
      "mean_token_accuracy": 0.8383203312754631,
      "num_tokens": 73226919.0,
      "step": 60890
    },
    {
      "entropy": 1.8610471919178964,
      "epoch": 0.18878467155264542,
      "grad_norm": 8.457162857055664,
      "learning_rate": 5.822545364028186e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8471423760056496,
      "num_tokens": 73238963.0,
      "step": 60900
    },
    {
      "entropy": 1.8293138653039933,
      "epoch": 0.1888156706776951,
      "grad_norm": 4.284148216247559,
      "learning_rate": 5.822067373553127e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8517989009618759,
      "num_tokens": 73251832.0,
      "step": 60910
    },
    {
      "entropy": 1.856334713101387,
      "epoch": 0.1888466698027448,
      "grad_norm": 8.815077781677246,
      "learning_rate": 5.8215895007778235e-06,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8325720369815827,
      "num_tokens": 73264796.0,
      "step": 60920
    },
    {
      "entropy": 1.6921759322285652,
      "epoch": 0.1888776689277945,
      "grad_norm": 7.586294174194336,
      "learning_rate": 5.821111745653979e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8488885432481765,
      "num_tokens": 73279601.0,
      "step": 60930
    },
    {
      "entropy": 1.9192057803273201,
      "epoch": 0.1889086680528442,
      "grad_norm": 8.917458534240723,
      "learning_rate": 5.820634108133325e-06,
      "loss": 0.5522,
      "mean_token_accuracy": 0.825607393682003,
      "num_tokens": 73291372.0,
      "step": 60940
    },
    {
      "entropy": 1.8995835900306701,
      "epoch": 0.1889396671778939,
      "grad_norm": 9.669869422912598,
      "learning_rate": 5.820156588167624e-06,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8387888163328171,
      "num_tokens": 73302760.0,
      "step": 60950
    },
    {
      "entropy": 1.9761246114969253,
      "epoch": 0.1889706663029436,
      "grad_norm": 9.00790786743164,
      "learning_rate": 5.819679185708661e-06,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8249676078557968,
      "num_tokens": 73313701.0,
      "step": 60960
    },
    {
      "entropy": 1.9618436962366104,
      "epoch": 0.1890016654279933,
      "grad_norm": 9.204923629760742,
      "learning_rate": 5.8192019007082515e-06,
      "loss": 0.5978,
      "mean_token_accuracy": 0.8198722168803215,
      "num_tokens": 73324576.0,
      "step": 60970
    },
    {
      "entropy": 1.8996351152658462,
      "epoch": 0.189032664553043,
      "grad_norm": 8.046465873718262,
      "learning_rate": 5.818724733118237e-06,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8394160747528077,
      "num_tokens": 73335871.0,
      "step": 60980
    },
    {
      "entropy": 1.8664459705352783,
      "epoch": 0.18906366367809269,
      "grad_norm": 9.74695873260498,
      "learning_rate": 5.8182476828904896e-06,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8346919432282448,
      "num_tokens": 73347256.0,
      "step": 60990
    },
    {
      "entropy": 1.9113014698028565,
      "epoch": 0.18909466280314238,
      "grad_norm": 9.092202186584473,
      "learning_rate": 5.817770749976909e-06,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8376618474721909,
      "num_tokens": 73358623.0,
      "step": 61000
    },
    {
      "entropy": 1.8848463252186776,
      "epoch": 0.18912566192819208,
      "grad_norm": 4.147286891937256,
      "learning_rate": 5.817293934329417e-06,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8313426792621612,
      "num_tokens": 73370417.0,
      "step": 61010
    },
    {
      "entropy": 1.9214195176959037,
      "epoch": 0.18915666105324178,
      "grad_norm": 7.7611403465271,
      "learning_rate": 5.816817235899967e-06,
      "loss": 0.5371,
      "mean_token_accuracy": 0.83161641061306,
      "num_tokens": 73382133.0,
      "step": 61020
    },
    {
      "entropy": 1.9057343572378158,
      "epoch": 0.18918766017829147,
      "grad_norm": 9.498258590698242,
      "learning_rate": 5.816340654640542e-06,
      "loss": 0.5504,
      "mean_token_accuracy": 0.8384771049022675,
      "num_tokens": 73393440.0,
      "step": 61030
    },
    {
      "entropy": 1.834838719666004,
      "epoch": 0.18921865930334117,
      "grad_norm": 4.797052383422852,
      "learning_rate": 5.815864190503149e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8543123677372932,
      "num_tokens": 73406284.0,
      "step": 61040
    },
    {
      "entropy": 1.8589958116412162,
      "epoch": 0.18924965842839087,
      "grad_norm": 8.2996826171875,
      "learning_rate": 5.815387843439824e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8410072878003121,
      "num_tokens": 73419184.0,
      "step": 61050
    },
    {
      "entropy": 1.9609655499458314,
      "epoch": 0.18928065755344056,
      "grad_norm": 8.885133743286133,
      "learning_rate": 5.814911613402629e-06,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8247525811195373,
      "num_tokens": 73430755.0,
      "step": 61060
    },
    {
      "entropy": 1.8497852653265,
      "epoch": 0.18931165667849026,
      "grad_norm": 8.99361801147461,
      "learning_rate": 5.814435500343657e-06,
      "loss": 0.467,
      "mean_token_accuracy": 0.8407298862934113,
      "num_tokens": 73443198.0,
      "step": 61070
    },
    {
      "entropy": 1.920815998315811,
      "epoch": 0.18934265580353996,
      "grad_norm": 8.852625846862793,
      "learning_rate": 5.813959504215025e-06,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8310668468475342,
      "num_tokens": 73453909.0,
      "step": 61080
    },
    {
      "entropy": 1.8594837307929992,
      "epoch": 0.18937365492858965,
      "grad_norm": 5.037121772766113,
      "learning_rate": 5.813483624968877e-06,
      "loss": 0.5775,
      "mean_token_accuracy": 0.82839694917202,
      "num_tokens": 73466187.0,
      "step": 61090
    },
    {
      "entropy": 1.816783571243286,
      "epoch": 0.18940465405363935,
      "grad_norm": 2.801055669784546,
      "learning_rate": 5.813007862557388e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.8508325532078743,
      "num_tokens": 73479306.0,
      "step": 61100
    },
    {
      "entropy": 1.9225141122937202,
      "epoch": 0.18943565317868905,
      "grad_norm": 8.705913543701172,
      "learning_rate": 5.812532216932759e-06,
      "loss": 0.6119,
      "mean_token_accuracy": 0.8328218296170234,
      "num_tokens": 73491113.0,
      "step": 61110
    },
    {
      "entropy": 1.8322004929184914,
      "epoch": 0.18946665230373871,
      "grad_norm": 3.7541184425354004,
      "learning_rate": 5.8120566880472155e-06,
      "loss": 0.4586,
      "mean_token_accuracy": 0.848419138789177,
      "num_tokens": 73503751.0,
      "step": 61120
    },
    {
      "entropy": 1.910209448635578,
      "epoch": 0.1894976514287884,
      "grad_norm": 9.48783016204834,
      "learning_rate": 5.811581275853014e-06,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8345631062984467,
      "num_tokens": 73515104.0,
      "step": 61130
    },
    {
      "entropy": 1.8471285864710807,
      "epoch": 0.1895286505538381,
      "grad_norm": 3.989353656768799,
      "learning_rate": 5.811105980302438e-06,
      "loss": 0.497,
      "mean_token_accuracy": 0.8420290291309357,
      "num_tokens": 73527180.0,
      "step": 61140
    },
    {
      "entropy": 1.8412006407976151,
      "epoch": 0.1895596496788878,
      "grad_norm": 9.421629905700684,
      "learning_rate": 5.810630801347794e-06,
      "loss": 0.5035,
      "mean_token_accuracy": 0.843493039906025,
      "num_tokens": 73540379.0,
      "step": 61150
    },
    {
      "entropy": 1.89210002720356,
      "epoch": 0.1895906488039375,
      "grad_norm": 4.73581075668335,
      "learning_rate": 5.810155738941422e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8357082083821297,
      "num_tokens": 73552901.0,
      "step": 61160
    },
    {
      "entropy": 1.899315556883812,
      "epoch": 0.1896216479289872,
      "grad_norm": 7.832334518432617,
      "learning_rate": 5.809680793035686e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8473002672195434,
      "num_tokens": 73564938.0,
      "step": 61170
    },
    {
      "entropy": 1.9637476325035095,
      "epoch": 0.1896526470540369,
      "grad_norm": 10.968403816223145,
      "learning_rate": 5.8092059635829754e-06,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8195749655365944,
      "num_tokens": 73576160.0,
      "step": 61180
    },
    {
      "entropy": 1.9289003670215608,
      "epoch": 0.1896836461790866,
      "grad_norm": 9.405878067016602,
      "learning_rate": 5.8087312505357115e-06,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8394549310207366,
      "num_tokens": 73588394.0,
      "step": 61190
    },
    {
      "entropy": 1.9212181255221368,
      "epoch": 0.1897146453041363,
      "grad_norm": 9.955124855041504,
      "learning_rate": 5.80825665384634e-06,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8281317040324211,
      "num_tokens": 73600253.0,
      "step": 61200
    },
    {
      "entropy": 1.8745829716324807,
      "epoch": 0.18974564442918598,
      "grad_norm": 8.92733097076416,
      "learning_rate": 5.807782173467334e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8472078263759613,
      "num_tokens": 73613086.0,
      "step": 61210
    },
    {
      "entropy": 1.931848169863224,
      "epoch": 0.18977664355423568,
      "grad_norm": 7.844409465789795,
      "learning_rate": 5.807307809351192e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8336080491542817,
      "num_tokens": 73624754.0,
      "step": 61220
    },
    {
      "entropy": 1.915302050113678,
      "epoch": 0.18980764267928538,
      "grad_norm": 7.905981540679932,
      "learning_rate": 5.806833561450444e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8398157224059105,
      "num_tokens": 73636264.0,
      "step": 61230
    },
    {
      "entropy": 1.980039805173874,
      "epoch": 0.18983864180433507,
      "grad_norm": 7.911728858947754,
      "learning_rate": 5.806359429717643e-06,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8273883283138275,
      "num_tokens": 73647130.0,
      "step": 61240
    },
    {
      "entropy": 1.9102006241679192,
      "epoch": 0.18986964092938477,
      "grad_norm": 8.068841934204102,
      "learning_rate": 5.805885414105372e-06,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8346290215849876,
      "num_tokens": 73658915.0,
      "step": 61250
    },
    {
      "entropy": 1.8921155214309693,
      "epoch": 0.18990064005443447,
      "grad_norm": 9.479058265686035,
      "learning_rate": 5.805411514566239e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8326509952545166,
      "num_tokens": 73670760.0,
      "step": 61260
    },
    {
      "entropy": 1.8476255759596825,
      "epoch": 0.18993163917948416,
      "grad_norm": 3.8252689838409424,
      "learning_rate": 5.804937731052881e-06,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8441173121333122,
      "num_tokens": 73683779.0,
      "step": 61270
    },
    {
      "entropy": 1.8536283493041992,
      "epoch": 0.18996263830453386,
      "grad_norm": 9.392457008361816,
      "learning_rate": 5.80446406351796e-06,
      "loss": 0.446,
      "mean_token_accuracy": 0.8531845271587372,
      "num_tokens": 73695924.0,
      "step": 61280
    },
    {
      "entropy": 2.0113365948200226,
      "epoch": 0.18999363742958356,
      "grad_norm": 9.121764183044434,
      "learning_rate": 5.803990511914166e-06,
      "loss": 0.6117,
      "mean_token_accuracy": 0.825063693523407,
      "num_tokens": 73706995.0,
      "step": 61290
    },
    {
      "entropy": 1.8952318519353866,
      "epoch": 0.19002463655463325,
      "grad_norm": 8.177303314208984,
      "learning_rate": 5.803517076194217e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8383290365338325,
      "num_tokens": 73718874.0,
      "step": 61300
    },
    {
      "entropy": 1.9754302859306336,
      "epoch": 0.19005563567968295,
      "grad_norm": 9.628982543945312,
      "learning_rate": 5.803043756310858e-06,
      "loss": 0.6201,
      "mean_token_accuracy": 0.8155344128608704,
      "num_tokens": 73729826.0,
      "step": 61310
    },
    {
      "entropy": 1.8714706540107726,
      "epoch": 0.19008663480473265,
      "grad_norm": 8.074376106262207,
      "learning_rate": 5.802570552216857e-06,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8505250707268714,
      "num_tokens": 73741893.0,
      "step": 61320
    },
    {
      "entropy": 1.9628629803657531,
      "epoch": 0.19011763392978234,
      "grad_norm": 4.110509872436523,
      "learning_rate": 5.802097463865013e-06,
      "loss": 0.5459,
      "mean_token_accuracy": 0.8337266921997071,
      "num_tokens": 73753334.0,
      "step": 61330
    },
    {
      "entropy": 1.9020420521497727,
      "epoch": 0.19014863305483204,
      "grad_norm": 8.163873672485352,
      "learning_rate": 5.801624491208153e-06,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8378794327378273,
      "num_tokens": 73765817.0,
      "step": 61340
    },
    {
      "entropy": 1.8595923513174057,
      "epoch": 0.19017963217988174,
      "grad_norm": 9.472771644592285,
      "learning_rate": 5.8011516341991266e-06,
      "loss": 0.4882,
      "mean_token_accuracy": 0.8379010885953904,
      "num_tokens": 73778450.0,
      "step": 61350
    },
    {
      "entropy": 1.9552810430526733,
      "epoch": 0.1902106313049314,
      "grad_norm": 9.82465648651123,
      "learning_rate": 5.800678892790814e-06,
      "loss": 0.531,
      "mean_token_accuracy": 0.8336144149303436,
      "num_tokens": 73789111.0,
      "step": 61360
    },
    {
      "entropy": 1.8182827576994895,
      "epoch": 0.1902416304299811,
      "grad_norm": 7.453169822692871,
      "learning_rate": 5.80020626693612e-06,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8479637071490288,
      "num_tokens": 73801615.0,
      "step": 61370
    },
    {
      "entropy": 1.791535222530365,
      "epoch": 0.1902726295550308,
      "grad_norm": 3.9200282096862793,
      "learning_rate": 5.799733756587978e-06,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8588753417134285,
      "num_tokens": 73815428.0,
      "step": 61380
    },
    {
      "entropy": 1.8905212104320526,
      "epoch": 0.1903036286800805,
      "grad_norm": 6.597655773162842,
      "learning_rate": 5.799261361699344e-06,
      "loss": 0.4899,
      "mean_token_accuracy": 0.8383428543806076,
      "num_tokens": 73826951.0,
      "step": 61390
    },
    {
      "entropy": 1.8670665681362153,
      "epoch": 0.1903346278051302,
      "grad_norm": 8.293522834777832,
      "learning_rate": 5.798789082223209e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8391018345952034,
      "num_tokens": 73839103.0,
      "step": 61400
    },
    {
      "entropy": 1.7990123346447944,
      "epoch": 0.1903656269301799,
      "grad_norm": 2.1400680541992188,
      "learning_rate": 5.798316918112582e-06,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8424833819270134,
      "num_tokens": 73852403.0,
      "step": 61410
    },
    {
      "entropy": 1.777857731282711,
      "epoch": 0.19039662605522958,
      "grad_norm": 4.169025897979736,
      "learning_rate": 5.797844869320504e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8537265419960022,
      "num_tokens": 73866168.0,
      "step": 61420
    },
    {
      "entropy": 1.9535469591617585,
      "epoch": 0.19042762518027928,
      "grad_norm": 9.52925968170166,
      "learning_rate": 5.797372935800044e-06,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8247589409351349,
      "num_tokens": 73876934.0,
      "step": 61430
    },
    {
      "entropy": 1.8025505855679511,
      "epoch": 0.19045862430532898,
      "grad_norm": 3.67460036277771,
      "learning_rate": 5.796901117504291e-06,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8523954421281814,
      "num_tokens": 73889905.0,
      "step": 61440
    },
    {
      "entropy": 1.879899947345257,
      "epoch": 0.19048962343037867,
      "grad_norm": 8.983809471130371,
      "learning_rate": 5.796429414386368e-06,
      "loss": 0.5415,
      "mean_token_accuracy": 0.823103578388691,
      "num_tokens": 73901347.0,
      "step": 61450
    },
    {
      "entropy": 1.9289403408765793,
      "epoch": 0.19052062255542837,
      "grad_norm": 8.888519287109375,
      "learning_rate": 5.7959578263994186e-06,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8268095463514328,
      "num_tokens": 73912721.0,
      "step": 61460
    },
    {
      "entropy": 1.7729464322328568,
      "epoch": 0.19055162168047807,
      "grad_norm": 9.862566947937012,
      "learning_rate": 5.79548635349662e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8498085349798202,
      "num_tokens": 73926277.0,
      "step": 61470
    },
    {
      "entropy": 1.8908159032464027,
      "epoch": 0.19058262080552776,
      "grad_norm": 9.228188514709473,
      "learning_rate": 5.795014995631168e-06,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8412437319755555,
      "num_tokens": 73937675.0,
      "step": 61480
    },
    {
      "entropy": 1.8851223319768906,
      "epoch": 0.19061361993057746,
      "grad_norm": 8.568156242370605,
      "learning_rate": 5.794543752756292e-06,
      "loss": 0.5504,
      "mean_token_accuracy": 0.82836285084486,
      "num_tokens": 73949212.0,
      "step": 61490
    },
    {
      "entropy": 1.9255726218223572,
      "epoch": 0.19064461905562716,
      "grad_norm": 7.397071838378906,
      "learning_rate": 5.794072624825245e-06,
      "loss": 0.5881,
      "mean_token_accuracy": 0.8346537619829177,
      "num_tokens": 73960504.0,
      "step": 61500
    },
    {
      "entropy": 1.9002389639616013,
      "epoch": 0.19067561818067685,
      "grad_norm": 9.786722183227539,
      "learning_rate": 5.793601611791305e-06,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8421397104859352,
      "num_tokens": 73972536.0,
      "step": 61510
    },
    {
      "entropy": 1.7949311509728432,
      "epoch": 0.19070661730572655,
      "grad_norm": 9.29088306427002,
      "learning_rate": 5.7931307136077804e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8478746846318245,
      "num_tokens": 73985554.0,
      "step": 61520
    },
    {
      "entropy": 1.8513918846845627,
      "epoch": 0.19073761643077625,
      "grad_norm": 7.442810535430908,
      "learning_rate": 5.792659930228004e-06,
      "loss": 0.5175,
      "mean_token_accuracy": 0.829763513803482,
      "num_tokens": 73997021.0,
      "step": 61530
    },
    {
      "entropy": 1.8692805111408233,
      "epoch": 0.19076861555582594,
      "grad_norm": 6.96608304977417,
      "learning_rate": 5.792189261605333e-06,
      "loss": 0.4901,
      "mean_token_accuracy": 0.8437046900391578,
      "num_tokens": 74009028.0,
      "step": 61540
    },
    {
      "entropy": 1.9384054720401764,
      "epoch": 0.19079961468087564,
      "grad_norm": 8.266419410705566,
      "learning_rate": 5.791718707693156e-06,
      "loss": 0.5769,
      "mean_token_accuracy": 0.830648484826088,
      "num_tokens": 74020561.0,
      "step": 61550
    },
    {
      "entropy": 1.869347333908081,
      "epoch": 0.19083061380592534,
      "grad_norm": 4.296153545379639,
      "learning_rate": 5.7912482684448845e-06,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8308837920427322,
      "num_tokens": 74032486.0,
      "step": 61560
    },
    {
      "entropy": 1.8627609625458716,
      "epoch": 0.19086161293097503,
      "grad_norm": 8.509102821350098,
      "learning_rate": 5.790777943813958e-06,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8389568880200386,
      "num_tokens": 74044672.0,
      "step": 61570
    },
    {
      "entropy": 1.8437671825289725,
      "epoch": 0.19089261205602473,
      "grad_norm": 8.853963851928711,
      "learning_rate": 5.79030773375384e-06,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8353106841444969,
      "num_tokens": 74057007.0,
      "step": 61580
    },
    {
      "entropy": 1.9143078982830048,
      "epoch": 0.19092361118107443,
      "grad_norm": 7.705349445343018,
      "learning_rate": 5.789837638218024e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8365962967276573,
      "num_tokens": 74068293.0,
      "step": 61590
    },
    {
      "entropy": 1.8211023643612863,
      "epoch": 0.19095461030612412,
      "grad_norm": 8.679984092712402,
      "learning_rate": 5.78936765716003e-06,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8368862703442573,
      "num_tokens": 74080738.0,
      "step": 61600
    },
    {
      "entropy": 1.9360493808984756,
      "epoch": 0.1909856094311738,
      "grad_norm": 8.391457557678223,
      "learning_rate": 5.788897790533401e-06,
      "loss": 0.583,
      "mean_token_accuracy": 0.8265061900019646,
      "num_tokens": 74091611.0,
      "step": 61610
    },
    {
      "entropy": 1.8157276138663292,
      "epoch": 0.1910166085562235,
      "grad_norm": 8.302699089050293,
      "learning_rate": 5.788428038291707e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8514835327863693,
      "num_tokens": 74104091.0,
      "step": 61620
    },
    {
      "entropy": 1.9097129538655282,
      "epoch": 0.19104760768127318,
      "grad_norm": 10.248120307922363,
      "learning_rate": 5.787958400388546e-06,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8318827226758003,
      "num_tokens": 74115663.0,
      "step": 61630
    },
    {
      "entropy": 1.805907167494297,
      "epoch": 0.19107860680632288,
      "grad_norm": 9.584233283996582,
      "learning_rate": 5.787488876777544e-06,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8393665820360183,
      "num_tokens": 74127987.0,
      "step": 61640
    },
    {
      "entropy": 1.772979559749365,
      "epoch": 0.19110960593137258,
      "grad_norm": 7.392131805419922,
      "learning_rate": 5.78701946741235e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8417062953114509,
      "num_tokens": 74141723.0,
      "step": 61650
    },
    {
      "entropy": 1.9116898834705354,
      "epoch": 0.19114060505642227,
      "grad_norm": 9.686699867248535,
      "learning_rate": 5.786550172246639e-06,
      "loss": 0.5491,
      "mean_token_accuracy": 0.8334579989314079,
      "num_tokens": 74152615.0,
      "step": 61660
    },
    {
      "entropy": 1.8346798852086068,
      "epoch": 0.19117160418147197,
      "grad_norm": 7.035947799682617,
      "learning_rate": 5.786080991234115e-06,
      "loss": 0.49,
      "mean_token_accuracy": 0.8480042293667793,
      "num_tokens": 74164838.0,
      "step": 61670
    },
    {
      "entropy": 1.822737891972065,
      "epoch": 0.19120260330652167,
      "grad_norm": 9.475711822509766,
      "learning_rate": 5.785611924328507e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8468611136078834,
      "num_tokens": 74177191.0,
      "step": 61680
    },
    {
      "entropy": 1.8375527679920196,
      "epoch": 0.19123360243157136,
      "grad_norm": 6.906650543212891,
      "learning_rate": 5.785142971483572e-06,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8478382468223572,
      "num_tokens": 74189816.0,
      "step": 61690
    },
    {
      "entropy": 1.7865667924284936,
      "epoch": 0.19126460155662106,
      "grad_norm": 4.482153415679932,
      "learning_rate": 5.784674132653088e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8439031273126603,
      "num_tokens": 74202820.0,
      "step": 61700
    },
    {
      "entropy": 1.8899090319871903,
      "epoch": 0.19129560068167076,
      "grad_norm": 8.343757629394531,
      "learning_rate": 5.784205407790866e-06,
      "loss": 0.481,
      "mean_token_accuracy": 0.8452587991952896,
      "num_tokens": 74214375.0,
      "step": 61710
    },
    {
      "entropy": 1.83342125415802,
      "epoch": 0.19132659980672045,
      "grad_norm": 8.29771614074707,
      "learning_rate": 5.783736796850737e-06,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8335444629192352,
      "num_tokens": 74227751.0,
      "step": 61720
    },
    {
      "entropy": 1.961284738779068,
      "epoch": 0.19135759893177015,
      "grad_norm": 9.149755477905273,
      "learning_rate": 5.783268299786564e-06,
      "loss": 0.544,
      "mean_token_accuracy": 0.833545659482479,
      "num_tokens": 74238590.0,
      "step": 61730
    },
    {
      "entropy": 1.8501825496554374,
      "epoch": 0.19138859805681985,
      "grad_norm": 10.697173118591309,
      "learning_rate": 5.782799916552232e-06,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8350521355867386,
      "num_tokens": 74250667.0,
      "step": 61740
    },
    {
      "entropy": 1.9134044259786607,
      "epoch": 0.19141959718186954,
      "grad_norm": 6.616395950317383,
      "learning_rate": 5.782331647101653e-06,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8395699933171272,
      "num_tokens": 74261958.0,
      "step": 61750
    },
    {
      "entropy": 1.9204273253679276,
      "epoch": 0.19145059630691924,
      "grad_norm": 10.518773078918457,
      "learning_rate": 5.781863491388767e-06,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8439541757106781,
      "num_tokens": 74272746.0,
      "step": 61760
    },
    {
      "entropy": 1.8314485549926758,
      "epoch": 0.19148159543196894,
      "grad_norm": 8.12490463256836,
      "learning_rate": 5.781395449367536e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8424828752875329,
      "num_tokens": 74285787.0,
      "step": 61770
    },
    {
      "entropy": 1.894325715303421,
      "epoch": 0.19151259455701863,
      "grad_norm": 10.066198348999023,
      "learning_rate": 5.780927520991953e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8336468815803528,
      "num_tokens": 74298311.0,
      "step": 61780
    },
    {
      "entropy": 1.9529366761445999,
      "epoch": 0.19154359368206833,
      "grad_norm": 9.384862899780273,
      "learning_rate": 5.780459706216036e-06,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8310287207365036,
      "num_tokens": 74309228.0,
      "step": 61790
    },
    {
      "entropy": 1.962406338751316,
      "epoch": 0.19157459280711803,
      "grad_norm": 7.662502288818359,
      "learning_rate": 5.779992004993824e-06,
      "loss": 0.5665,
      "mean_token_accuracy": 0.8225008681416511,
      "num_tokens": 74320341.0,
      "step": 61800
    },
    {
      "entropy": 1.9238436996936799,
      "epoch": 0.19160559193216772,
      "grad_norm": 9.47153091430664,
      "learning_rate": 5.779524417279388e-06,
      "loss": 0.526,
      "mean_token_accuracy": 0.836972689628601,
      "num_tokens": 74332180.0,
      "step": 61810
    },
    {
      "entropy": 1.933986322581768,
      "epoch": 0.19163659105721742,
      "grad_norm": 7.185315132141113,
      "learning_rate": 5.779056943026824e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8494284927845002,
      "num_tokens": 74343712.0,
      "step": 61820
    },
    {
      "entropy": 1.920379176735878,
      "epoch": 0.19166759018226712,
      "grad_norm": 4.159258842468262,
      "learning_rate": 5.778589582190252e-06,
      "loss": 0.489,
      "mean_token_accuracy": 0.836681205034256,
      "num_tokens": 74356004.0,
      "step": 61830
    },
    {
      "entropy": 1.8799566522240638,
      "epoch": 0.1916985893073168,
      "grad_norm": 8.231870651245117,
      "learning_rate": 5.778122334723817e-06,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8473984330892563,
      "num_tokens": 74368305.0,
      "step": 61840
    },
    {
      "entropy": 1.893815641105175,
      "epoch": 0.1917295884323665,
      "grad_norm": 4.571141719818115,
      "learning_rate": 5.777655200581693e-06,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8461870953440667,
      "num_tokens": 74380533.0,
      "step": 61850
    },
    {
      "entropy": 1.8599497005343437,
      "epoch": 0.19176058755741618,
      "grad_norm": 4.433168411254883,
      "learning_rate": 5.77718817971808e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8428743481636047,
      "num_tokens": 74393564.0,
      "step": 61860
    },
    {
      "entropy": 1.9357849955558777,
      "epoch": 0.19179158668246588,
      "grad_norm": 8.343416213989258,
      "learning_rate": 5.776721272087201e-06,
      "loss": 0.5484,
      "mean_token_accuracy": 0.8426366299390793,
      "num_tokens": 74405577.0,
      "step": 61870
    },
    {
      "entropy": 2.0002988666296004,
      "epoch": 0.19182258580751557,
      "grad_norm": 8.28388500213623,
      "learning_rate": 5.776254477643307e-06,
      "loss": 0.5856,
      "mean_token_accuracy": 0.8197537645697593,
      "num_tokens": 74416662.0,
      "step": 61880
    },
    {
      "entropy": 1.8667000889778138,
      "epoch": 0.19185358493256527,
      "grad_norm": 10.070082664489746,
      "learning_rate": 5.775787796340675e-06,
      "loss": 0.5541,
      "mean_token_accuracy": 0.8392659932374954,
      "num_tokens": 74428842.0,
      "step": 61890
    },
    {
      "entropy": 1.8394625827670097,
      "epoch": 0.19188458405761497,
      "grad_norm": 2.4154021739959717,
      "learning_rate": 5.775321228133606e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8550085946917534,
      "num_tokens": 74441442.0,
      "step": 61900
    },
    {
      "entropy": 1.8814760237932204,
      "epoch": 0.19191558318266466,
      "grad_norm": 8.660712242126465,
      "learning_rate": 5.77485477297643e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8466302067041397,
      "num_tokens": 74453829.0,
      "step": 61910
    },
    {
      "entropy": 1.8994119971990586,
      "epoch": 0.19194658230771436,
      "grad_norm": 8.826970100402832,
      "learning_rate": 5.774388430823499e-06,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8399457424879074,
      "num_tokens": 74464456.0,
      "step": 61920
    },
    {
      "entropy": 1.9710954904556275,
      "epoch": 0.19197758143276406,
      "grad_norm": 7.8489155769348145,
      "learning_rate": 5.773922201629193e-06,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8248024433851242,
      "num_tokens": 74475138.0,
      "step": 61930
    },
    {
      "entropy": 1.8401651561260224,
      "epoch": 0.19200858055781375,
      "grad_norm": 9.654853820800781,
      "learning_rate": 5.7734560853479185e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8436680495738983,
      "num_tokens": 74488049.0,
      "step": 61940
    },
    {
      "entropy": 1.839670716226101,
      "epoch": 0.19203957968286345,
      "grad_norm": 8.641029357910156,
      "learning_rate": 5.772990081934104e-06,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8443516567349434,
      "num_tokens": 74500733.0,
      "step": 61950
    },
    {
      "entropy": 1.8809733077883721,
      "epoch": 0.19207057880791314,
      "grad_norm": 7.769062519073486,
      "learning_rate": 5.772524191342211e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8486372217535972,
      "num_tokens": 74512926.0,
      "step": 61960
    },
    {
      "entropy": 1.8726215928792953,
      "epoch": 0.19210157793296284,
      "grad_norm": 9.620684623718262,
      "learning_rate": 5.77205841352672e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8346716165542603,
      "num_tokens": 74524851.0,
      "step": 61970
    },
    {
      "entropy": 1.8348112776875496,
      "epoch": 0.19213257705801254,
      "grad_norm": 9.443613052368164,
      "learning_rate": 5.771592748442137e-06,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8425130292773246,
      "num_tokens": 74537835.0,
      "step": 61980
    },
    {
      "entropy": 1.871398164331913,
      "epoch": 0.19216357618306223,
      "grad_norm": 4.7039875984191895,
      "learning_rate": 5.771127196042999e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8458441883325577,
      "num_tokens": 74549780.0,
      "step": 61990
    },
    {
      "entropy": 1.8827469125390053,
      "epoch": 0.19219457530811193,
      "grad_norm": 10.308496475219727,
      "learning_rate": 5.770661756283866e-06,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8351684272289276,
      "num_tokens": 74561738.0,
      "step": 62000
    },
    {
      "entropy": 1.9043960571289062,
      "epoch": 0.19222557443316163,
      "grad_norm": 7.712822914123535,
      "learning_rate": 5.7701964291193214e-06,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8395509481430053,
      "num_tokens": 74573112.0,
      "step": 62010
    },
    {
      "entropy": 1.9056378319859504,
      "epoch": 0.19225657355821132,
      "grad_norm": 9.439181327819824,
      "learning_rate": 5.769731214503978e-06,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8385317623615265,
      "num_tokens": 74583987.0,
      "step": 62020
    },
    {
      "entropy": 1.8788242667913437,
      "epoch": 0.19228757268326102,
      "grad_norm": 3.4620392322540283,
      "learning_rate": 5.76926611239247e-06,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8428028956055641,
      "num_tokens": 74595794.0,
      "step": 62030
    },
    {
      "entropy": 1.8820303320884704,
      "epoch": 0.19231857180831072,
      "grad_norm": 10.090309143066406,
      "learning_rate": 5.7688011227394625e-06,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8451215922832489,
      "num_tokens": 74607298.0,
      "step": 62040
    },
    {
      "entropy": 1.886990125477314,
      "epoch": 0.19234957093336041,
      "grad_norm": 8.511155128479004,
      "learning_rate": 5.768336245499641e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8422086775302887,
      "num_tokens": 74619848.0,
      "step": 62050
    },
    {
      "entropy": 1.9199146822094917,
      "epoch": 0.1923805700584101,
      "grad_norm": 8.11715316772461,
      "learning_rate": 5.767871480627723e-06,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8385195031762123,
      "num_tokens": 74631619.0,
      "step": 62060
    },
    {
      "entropy": 1.8858480796217918,
      "epoch": 0.1924115691834598,
      "grad_norm": 7.366301536560059,
      "learning_rate": 5.767406828078441e-06,
      "loss": 0.489,
      "mean_token_accuracy": 0.8323479920625687,
      "num_tokens": 74643673.0,
      "step": 62070
    },
    {
      "entropy": 1.8879368424415588,
      "epoch": 0.1924425683085095,
      "grad_norm": 5.003079414367676,
      "learning_rate": 5.766942287806564e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8332347095012664,
      "num_tokens": 74656285.0,
      "step": 62080
    },
    {
      "entropy": 1.8710455060005189,
      "epoch": 0.1924735674335592,
      "grad_norm": 3.315375804901123,
      "learning_rate": 5.766477859766882e-06,
      "loss": 0.5307,
      "mean_token_accuracy": 0.829591977596283,
      "num_tokens": 74668414.0,
      "step": 62090
    },
    {
      "entropy": 1.8381486520171166,
      "epoch": 0.19250456655860887,
      "grad_norm": 7.80744743347168,
      "learning_rate": 5.766013543914207e-06,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8402635097503662,
      "num_tokens": 74680230.0,
      "step": 62100
    },
    {
      "entropy": 1.9462435692548752,
      "epoch": 0.19253556568365857,
      "grad_norm": 8.435437202453613,
      "learning_rate": 5.7655493402033836e-06,
      "loss": 0.5916,
      "mean_token_accuracy": 0.8307736337184906,
      "num_tokens": 74692003.0,
      "step": 62110
    },
    {
      "entropy": 1.903138768672943,
      "epoch": 0.19256656480870826,
      "grad_norm": 8.948243141174316,
      "learning_rate": 5.7650852485892765e-06,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8443427816033363,
      "num_tokens": 74703411.0,
      "step": 62120
    },
    {
      "entropy": 1.8868961855769157,
      "epoch": 0.19259756393375796,
      "grad_norm": 11.374164581298828,
      "learning_rate": 5.764621269026775e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8403722077608109,
      "num_tokens": 74715521.0,
      "step": 62130
    },
    {
      "entropy": 1.892562797665596,
      "epoch": 0.19262856305880766,
      "grad_norm": 7.841557025909424,
      "learning_rate": 5.764157401470803e-06,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8378325685858726,
      "num_tokens": 74727068.0,
      "step": 62140
    },
    {
      "entropy": 1.7926482424139976,
      "epoch": 0.19265956218385735,
      "grad_norm": 3.997764825820923,
      "learning_rate": 5.763693645876296e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8492718085646629,
      "num_tokens": 74740269.0,
      "step": 62150
    },
    {
      "entropy": 1.866290318965912,
      "epoch": 0.19269056130890705,
      "grad_norm": 4.6976494789123535,
      "learning_rate": 5.763230002198225e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8362017720937729,
      "num_tokens": 74751597.0,
      "step": 62160
    },
    {
      "entropy": 1.8112548857927322,
      "epoch": 0.19272156043395675,
      "grad_norm": 8.123526573181152,
      "learning_rate": 5.762766470391583e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8409424394369125,
      "num_tokens": 74764989.0,
      "step": 62170
    },
    {
      "entropy": 1.753453615307808,
      "epoch": 0.19275255955900644,
      "grad_norm": 4.4287848472595215,
      "learning_rate": 5.762303050411388e-06,
      "loss": 0.397,
      "mean_token_accuracy": 0.8451619282364845,
      "num_tokens": 74778762.0,
      "step": 62180
    },
    {
      "entropy": 1.929175502061844,
      "epoch": 0.19278355868405614,
      "grad_norm": 9.11049747467041,
      "learning_rate": 5.761839742212686e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.841756422817707,
      "num_tokens": 74789840.0,
      "step": 62190
    },
    {
      "entropy": 1.9000461488962173,
      "epoch": 0.19281455780910584,
      "grad_norm": 4.639410972595215,
      "learning_rate": 5.761376545750544e-06,
      "loss": 0.493,
      "mean_token_accuracy": 0.8386220842599869,
      "num_tokens": 74801445.0,
      "step": 62200
    },
    {
      "entropy": 1.8612179577350616,
      "epoch": 0.19284555693415553,
      "grad_norm": 9.239517211914062,
      "learning_rate": 5.760913460980057e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8468897223472596,
      "num_tokens": 74813988.0,
      "step": 62210
    },
    {
      "entropy": 1.7820834949612618,
      "epoch": 0.19287655605920523,
      "grad_norm": 3.959775447845459,
      "learning_rate": 5.760450487856346e-06,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8533436298370362,
      "num_tokens": 74827651.0,
      "step": 62220
    },
    {
      "entropy": 1.8692509412765503,
      "epoch": 0.19290755518425493,
      "grad_norm": 9.524608612060547,
      "learning_rate": 5.759987626334555e-06,
      "loss": 0.498,
      "mean_token_accuracy": 0.8503947392106056,
      "num_tokens": 74840095.0,
      "step": 62230
    },
    {
      "entropy": 1.9380140736699105,
      "epoch": 0.19293855430930462,
      "grad_norm": 8.502176284790039,
      "learning_rate": 5.759524876369853e-06,
      "loss": 0.5736,
      "mean_token_accuracy": 0.8298733696341515,
      "num_tokens": 74851878.0,
      "step": 62240
    },
    {
      "entropy": 1.928398758172989,
      "epoch": 0.19296955343435432,
      "grad_norm": 9.923539161682129,
      "learning_rate": 5.759062237917436e-06,
      "loss": 0.523,
      "mean_token_accuracy": 0.8408984065055847,
      "num_tokens": 74862822.0,
      "step": 62250
    },
    {
      "entropy": 1.8122706308960914,
      "epoch": 0.19300055255940401,
      "grad_norm": 8.388425827026367,
      "learning_rate": 5.758599710932528e-06,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8447504699230194,
      "num_tokens": 74875649.0,
      "step": 62260
    },
    {
      "entropy": 1.8911856144666672,
      "epoch": 0.1930315516844537,
      "grad_norm": 8.741740226745605,
      "learning_rate": 5.75813729537037e-06,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8428622677922248,
      "num_tokens": 74887627.0,
      "step": 62270
    },
    {
      "entropy": 1.86591185182333,
      "epoch": 0.1930625508095034,
      "grad_norm": 8.541808128356934,
      "learning_rate": 5.757674991186235e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8314554423093796,
      "num_tokens": 74900605.0,
      "step": 62280
    },
    {
      "entropy": 1.926130199432373,
      "epoch": 0.1930935499345531,
      "grad_norm": 7.839807510375977,
      "learning_rate": 5.75721279833542e-06,
      "loss": 0.5659,
      "mean_token_accuracy": 0.8310964539647102,
      "num_tokens": 74912462.0,
      "step": 62290
    },
    {
      "entropy": 1.8626934379339217,
      "epoch": 0.1931245490596028,
      "grad_norm": 11.213129997253418,
      "learning_rate": 5.756750716773244e-06,
      "loss": 0.4776,
      "mean_token_accuracy": 0.8312896370887757,
      "num_tokens": 74924678.0,
      "step": 62300
    },
    {
      "entropy": 1.9561289519071579,
      "epoch": 0.1931555481846525,
      "grad_norm": 9.022578239440918,
      "learning_rate": 5.7562887464550565e-06,
      "loss": 0.5684,
      "mean_token_accuracy": 0.83260398209095,
      "num_tokens": 74935522.0,
      "step": 62310
    },
    {
      "entropy": 1.935487399995327,
      "epoch": 0.1931865473097022,
      "grad_norm": 8.986858367919922,
      "learning_rate": 5.755826887336227e-06,
      "loss": 0.5735,
      "mean_token_accuracy": 0.8362299337983131,
      "num_tokens": 74947411.0,
      "step": 62320
    },
    {
      "entropy": 1.8337993949651719,
      "epoch": 0.1932175464347519,
      "grad_norm": 8.892833709716797,
      "learning_rate": 5.75536513937215e-06,
      "loss": 0.5033,
      "mean_token_accuracy": 0.840664692223072,
      "num_tokens": 74961058.0,
      "step": 62330
    },
    {
      "entropy": 1.9785827055573464,
      "epoch": 0.1932485455598016,
      "grad_norm": 8.412293434143066,
      "learning_rate": 5.7549035025182494e-06,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8272167652845382,
      "num_tokens": 74973029.0,
      "step": 62340
    },
    {
      "entropy": 1.8726650208234787,
      "epoch": 0.19327954468485126,
      "grad_norm": 3.809394598007202,
      "learning_rate": 5.754441976729972e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.846262401342392,
      "num_tokens": 74985904.0,
      "step": 62350
    },
    {
      "entropy": 1.8960716605186463,
      "epoch": 0.19331054380990095,
      "grad_norm": 4.894566535949707,
      "learning_rate": 5.753980561962787e-06,
      "loss": 0.5005,
      "mean_token_accuracy": 0.846526712179184,
      "num_tokens": 74997658.0,
      "step": 62360
    },
    {
      "entropy": 1.9618462428450585,
      "epoch": 0.19334154293495065,
      "grad_norm": 7.828151226043701,
      "learning_rate": 5.753519258172194e-06,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8398016721010209,
      "num_tokens": 75009078.0,
      "step": 62370
    },
    {
      "entropy": 1.8819516450166702,
      "epoch": 0.19337254206000035,
      "grad_norm": 8.260143280029297,
      "learning_rate": 5.753058065313714e-06,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8390106439590455,
      "num_tokens": 75021094.0,
      "step": 62380
    },
    {
      "entropy": 1.8420429840683936,
      "epoch": 0.19340354118505004,
      "grad_norm": 7.1242475509643555,
      "learning_rate": 5.7525969833428895e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.834058178961277,
      "num_tokens": 75033633.0,
      "step": 62390
    },
    {
      "entropy": 1.9346123069524765,
      "epoch": 0.19343454031009974,
      "grad_norm": 7.526602268218994,
      "learning_rate": 5.752136012215297e-06,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8317095413804054,
      "num_tokens": 75045262.0,
      "step": 62400
    },
    {
      "entropy": 1.9385161980986596,
      "epoch": 0.19346553943514944,
      "grad_norm": 7.26128625869751,
      "learning_rate": 5.751675151886529e-06,
      "loss": 0.492,
      "mean_token_accuracy": 0.8422947779297829,
      "num_tokens": 75057520.0,
      "step": 62410
    },
    {
      "entropy": 1.8083595961332322,
      "epoch": 0.19349653856019913,
      "grad_norm": 8.095551490783691,
      "learning_rate": 5.751214402312208e-06,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8498411178588867,
      "num_tokens": 75070750.0,
      "step": 62420
    },
    {
      "entropy": 1.8618351459503173,
      "epoch": 0.19352753768524883,
      "grad_norm": 9.302799224853516,
      "learning_rate": 5.750753763447981e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8428505852818489,
      "num_tokens": 75083142.0,
      "step": 62430
    },
    {
      "entropy": 1.874634511768818,
      "epoch": 0.19355853681029853,
      "grad_norm": 8.7201566696167,
      "learning_rate": 5.750293235249518e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8446508884429932,
      "num_tokens": 75094939.0,
      "step": 62440
    },
    {
      "entropy": 1.9367821574211121,
      "epoch": 0.19358953593534822,
      "grad_norm": 8.614350318908691,
      "learning_rate": 5.749832817672515e-06,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8392156153917313,
      "num_tokens": 75106355.0,
      "step": 62450
    },
    {
      "entropy": 1.873591238260269,
      "epoch": 0.19362053506039792,
      "grad_norm": 9.169925689697266,
      "learning_rate": 5.749372510672692e-06,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8400036081671715,
      "num_tokens": 75117711.0,
      "step": 62460
    },
    {
      "entropy": 1.8877021759748458,
      "epoch": 0.19365153418544762,
      "grad_norm": 4.023455619812012,
      "learning_rate": 5.748912314205795e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8442784741520881,
      "num_tokens": 75130177.0,
      "step": 62470
    },
    {
      "entropy": 1.9278222426772118,
      "epoch": 0.1936825333104973,
      "grad_norm": 8.48756217956543,
      "learning_rate": 5.748452228227594e-06,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8467386439442635,
      "num_tokens": 75141376.0,
      "step": 62480
    },
    {
      "entropy": 1.8919538147747517,
      "epoch": 0.193713532435547,
      "grad_norm": 8.26032829284668,
      "learning_rate": 5.7479922526938844e-06,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8378899380564689,
      "num_tokens": 75153872.0,
      "step": 62490
    },
    {
      "entropy": 1.975018060207367,
      "epoch": 0.1937445315605967,
      "grad_norm": 12.015787124633789,
      "learning_rate": 5.747532387560486e-06,
      "loss": 0.5763,
      "mean_token_accuracy": 0.8265995383262634,
      "num_tokens": 75164449.0,
      "step": 62500
    },
    {
      "entropy": 1.9173302993178367,
      "epoch": 0.1937755306856464,
      "grad_norm": 9.031746864318848,
      "learning_rate": 5.747072632783242e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.8385538429021835,
      "num_tokens": 75176299.0,
      "step": 62510
    },
    {
      "entropy": 1.9593890875577926,
      "epoch": 0.1938065298106961,
      "grad_norm": 9.437821388244629,
      "learning_rate": 5.746612988318023e-06,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8430341646075249,
      "num_tokens": 75187861.0,
      "step": 62520
    },
    {
      "entropy": 1.8832410350441933,
      "epoch": 0.1938375289357458,
      "grad_norm": 13.652206420898438,
      "learning_rate": 5.7461534541207234e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8412627220153809,
      "num_tokens": 75200163.0,
      "step": 62530
    },
    {
      "entropy": 1.9353638172149659,
      "epoch": 0.1938685280607955,
      "grad_norm": 7.707444667816162,
      "learning_rate": 5.745694030147259e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.839554651081562,
      "num_tokens": 75211123.0,
      "step": 62540
    },
    {
      "entropy": 1.8284053675830365,
      "epoch": 0.1938995271858452,
      "grad_norm": 8.445575714111328,
      "learning_rate": 5.745234716353575e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.84172955006361,
      "num_tokens": 75223852.0,
      "step": 62550
    },
    {
      "entropy": 1.9116384595632554,
      "epoch": 0.19393052631089489,
      "grad_norm": 8.664936065673828,
      "learning_rate": 5.744775512695639e-06,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8343887582421303,
      "num_tokens": 75235614.0,
      "step": 62560
    },
    {
      "entropy": 1.8100930273532867,
      "epoch": 0.19396152543594458,
      "grad_norm": 7.181854724884033,
      "learning_rate": 5.744316419129445e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8528478279709816,
      "num_tokens": 75248586.0,
      "step": 62570
    },
    {
      "entropy": 1.9202060773968697,
      "epoch": 0.19399252456099428,
      "grad_norm": 3.0128285884857178,
      "learning_rate": 5.743857435611008e-06,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8501577556133271,
      "num_tokens": 75259638.0,
      "step": 62580
    },
    {
      "entropy": 1.782825130224228,
      "epoch": 0.19402352368604397,
      "grad_norm": 9.068933486938477,
      "learning_rate": 5.743398562096369e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8459088578820229,
      "num_tokens": 75273019.0,
      "step": 62590
    },
    {
      "entropy": 1.912159901857376,
      "epoch": 0.19405452281109364,
      "grad_norm": 8.978429794311523,
      "learning_rate": 5.742939798541598e-06,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8394264042377472,
      "num_tokens": 75284329.0,
      "step": 62600
    },
    {
      "entropy": 1.9709818661212921,
      "epoch": 0.19408552193614334,
      "grad_norm": 8.043622970581055,
      "learning_rate": 5.742481144902782e-06,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8323390766978264,
      "num_tokens": 75295155.0,
      "step": 62610
    },
    {
      "entropy": 1.8821393936872481,
      "epoch": 0.19411652106119304,
      "grad_norm": 7.908102989196777,
      "learning_rate": 5.742022601136038e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8384183317422866,
      "num_tokens": 75306535.0,
      "step": 62620
    },
    {
      "entropy": 1.8917057454586028,
      "epoch": 0.19414752018624273,
      "grad_norm": 4.215620994567871,
      "learning_rate": 5.741564167197507e-06,
      "loss": 0.514,
      "mean_token_accuracy": 0.8402154177427292,
      "num_tokens": 75318253.0,
      "step": 62630
    },
    {
      "entropy": 1.9241933241486548,
      "epoch": 0.19417851931129243,
      "grad_norm": 9.891353607177734,
      "learning_rate": 5.741105843043353e-06,
      "loss": 0.5965,
      "mean_token_accuracy": 0.818701134622097,
      "num_tokens": 75329353.0,
      "step": 62640
    },
    {
      "entropy": 1.9373375624418259,
      "epoch": 0.19420951843634213,
      "grad_norm": 9.279706954956055,
      "learning_rate": 5.740647628629763e-06,
      "loss": 0.525,
      "mean_token_accuracy": 0.8327772691845894,
      "num_tokens": 75340932.0,
      "step": 62650
    },
    {
      "entropy": 1.8756248265504838,
      "epoch": 0.19424051756139182,
      "grad_norm": 5.068475723266602,
      "learning_rate": 5.740189523912952e-06,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8361865937709808,
      "num_tokens": 75353705.0,
      "step": 62660
    },
    {
      "entropy": 1.9200247406959534,
      "epoch": 0.19427151668644152,
      "grad_norm": 8.971196174621582,
      "learning_rate": 5.7397315288491585e-06,
      "loss": 0.553,
      "mean_token_accuracy": 0.8186924889683723,
      "num_tokens": 75365245.0,
      "step": 62670
    },
    {
      "entropy": 1.87558753490448,
      "epoch": 0.19430251581149122,
      "grad_norm": 8.281537055969238,
      "learning_rate": 5.7392736433946424e-06,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8481769561767578,
      "num_tokens": 75377140.0,
      "step": 62680
    },
    {
      "entropy": 1.904216541349888,
      "epoch": 0.1943335149365409,
      "grad_norm": 8.079618453979492,
      "learning_rate": 5.738815867505695e-06,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8357766851782799,
      "num_tokens": 75389067.0,
      "step": 62690
    },
    {
      "entropy": 1.8466911643743515,
      "epoch": 0.1943645140615906,
      "grad_norm": 8.569628715515137,
      "learning_rate": 5.7383582011386215e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8424016252160073,
      "num_tokens": 75401290.0,
      "step": 62700
    },
    {
      "entropy": 1.8429038256406785,
      "epoch": 0.1943955131866403,
      "grad_norm": 4.195886135101318,
      "learning_rate": 5.737900644249762e-06,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8370422974228859,
      "num_tokens": 75414571.0,
      "step": 62710
    },
    {
      "entropy": 1.870844414830208,
      "epoch": 0.19442651231169,
      "grad_norm": 10.276758193969727,
      "learning_rate": 5.737443196795474e-06,
      "loss": 0.5647,
      "mean_token_accuracy": 0.836325615644455,
      "num_tokens": 75426778.0,
      "step": 62720
    },
    {
      "entropy": 1.8725099056959151,
      "epoch": 0.1944575114367397,
      "grad_norm": 7.546666622161865,
      "learning_rate": 5.736985858732143e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8451737076044082,
      "num_tokens": 75438333.0,
      "step": 62730
    },
    {
      "entropy": 1.9273310750722885,
      "epoch": 0.1944885105617894,
      "grad_norm": 9.644882202148438,
      "learning_rate": 5.736528630016177e-06,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8291450396180153,
      "num_tokens": 75450773.0,
      "step": 62740
    },
    {
      "entropy": 1.9267635852098466,
      "epoch": 0.1945195096868391,
      "grad_norm": 3.730067729949951,
      "learning_rate": 5.7360715106040076e-06,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8472494944930077,
      "num_tokens": 75462100.0,
      "step": 62750
    },
    {
      "entropy": 1.8266283124685287,
      "epoch": 0.1945505088118888,
      "grad_norm": 7.616328239440918,
      "learning_rate": 5.735614500452095e-06,
      "loss": 0.437,
      "mean_token_accuracy": 0.8459511280059815,
      "num_tokens": 75475416.0,
      "step": 62760
    },
    {
      "entropy": 1.9178658738732337,
      "epoch": 0.19458150793693849,
      "grad_norm": 9.340888977050781,
      "learning_rate": 5.7351575995169186e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.835166348516941,
      "num_tokens": 75487885.0,
      "step": 62770
    },
    {
      "entropy": 1.8748321622610091,
      "epoch": 0.19461250706198818,
      "grad_norm": 3.991840124130249,
      "learning_rate": 5.734700807754984e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.8418037429451942,
      "num_tokens": 75499947.0,
      "step": 62780
    },
    {
      "entropy": 1.8018044173717498,
      "epoch": 0.19464350618703788,
      "grad_norm": 8.51059341430664,
      "learning_rate": 5.734244125122822e-06,
      "loss": 0.4169,
      "mean_token_accuracy": 0.8524055883288384,
      "num_tokens": 75513475.0,
      "step": 62790
    },
    {
      "entropy": 1.9655559062957764,
      "epoch": 0.19467450531208758,
      "grad_norm": 10.379049301147461,
      "learning_rate": 5.733787551576987e-06,
      "loss": 0.5602,
      "mean_token_accuracy": 0.8376842170953751,
      "num_tokens": 75524934.0,
      "step": 62800
    },
    {
      "entropy": 1.880786569416523,
      "epoch": 0.19470550443713727,
      "grad_norm": 8.091976165771484,
      "learning_rate": 5.7333310870740565e-06,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8456220507621766,
      "num_tokens": 75537225.0,
      "step": 62810
    },
    {
      "entropy": 1.9510624960064888,
      "epoch": 0.19473650356218697,
      "grad_norm": 9.744572639465332,
      "learning_rate": 5.732874731570633e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8450943067669868,
      "num_tokens": 75548663.0,
      "step": 62820
    },
    {
      "entropy": 1.9875933229923248,
      "epoch": 0.19476750268723667,
      "grad_norm": 8.468138694763184,
      "learning_rate": 5.732418485023345e-06,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8282896682620049,
      "num_tokens": 75559697.0,
      "step": 62830
    },
    {
      "entropy": 1.8670821204781531,
      "epoch": 0.19479850181228636,
      "grad_norm": 9.108405113220215,
      "learning_rate": 5.731962347388841e-06,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8396436840295791,
      "num_tokens": 75571652.0,
      "step": 62840
    },
    {
      "entropy": 1.8874829977750778,
      "epoch": 0.19482950093733603,
      "grad_norm": 7.285294055938721,
      "learning_rate": 5.731506318623798e-06,
      "loss": 0.5263,
      "mean_token_accuracy": 0.836583924293518,
      "num_tokens": 75583725.0,
      "step": 62850
    },
    {
      "entropy": 1.8510995037853717,
      "epoch": 0.19486050006238573,
      "grad_norm": 8.491140365600586,
      "learning_rate": 5.731050398684913e-06,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8445952758193016,
      "num_tokens": 75596812.0,
      "step": 62860
    },
    {
      "entropy": 1.9246593803167342,
      "epoch": 0.19489149918743542,
      "grad_norm": 8.031885147094727,
      "learning_rate": 5.7305945875289126e-06,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8263408780097962,
      "num_tokens": 75609185.0,
      "step": 62870
    },
    {
      "entropy": 1.9077463194727897,
      "epoch": 0.19492249831248512,
      "grad_norm": 8.196245193481445,
      "learning_rate": 5.730138885112542e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.837520606815815,
      "num_tokens": 75621376.0,
      "step": 62880
    },
    {
      "entropy": 1.9066799104213714,
      "epoch": 0.19495349743753482,
      "grad_norm": 8.017034530639648,
      "learning_rate": 5.729683291392573e-06,
      "loss": 0.522,
      "mean_token_accuracy": 0.8396532908082008,
      "num_tokens": 75634042.0,
      "step": 62890
    },
    {
      "entropy": 1.8118005082011224,
      "epoch": 0.19498449656258451,
      "grad_norm": 7.996087551116943,
      "learning_rate": 5.7292278063258015e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8566027864813804,
      "num_tokens": 75647509.0,
      "step": 62900
    },
    {
      "entropy": 1.9176707819104195,
      "epoch": 0.1950154956876342,
      "grad_norm": 3.844882011413574,
      "learning_rate": 5.728772429869045e-06,
      "loss": 0.5002,
      "mean_token_accuracy": 0.842312179505825,
      "num_tokens": 75659954.0,
      "step": 62910
    },
    {
      "entropy": 1.8584643453359604,
      "epoch": 0.1950464948126839,
      "grad_norm": 4.170616149902344,
      "learning_rate": 5.72831716197915e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8405888646841049,
      "num_tokens": 75672753.0,
      "step": 62920
    },
    {
      "entropy": 1.8920593187212944,
      "epoch": 0.1950774939377336,
      "grad_norm": 8.466096878051758,
      "learning_rate": 5.727862002612982e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8370663553476334,
      "num_tokens": 75685124.0,
      "step": 62930
    },
    {
      "entropy": 1.9039631336927414,
      "epoch": 0.1951084930627833,
      "grad_norm": 10.970901489257812,
      "learning_rate": 5.727406951727435e-06,
      "loss": 0.5118,
      "mean_token_accuracy": 0.837016536295414,
      "num_tokens": 75696871.0,
      "step": 62940
    },
    {
      "entropy": 1.8986446857452393,
      "epoch": 0.195139492187833,
      "grad_norm": 8.603690147399902,
      "learning_rate": 5.726952009279424e-06,
      "loss": 0.4586,
      "mean_token_accuracy": 0.845548364520073,
      "num_tokens": 75708850.0,
      "step": 62950
    },
    {
      "entropy": 1.879063467681408,
      "epoch": 0.1951704913128827,
      "grad_norm": 8.833934783935547,
      "learning_rate": 5.726497175225886e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8337208881974221,
      "num_tokens": 75721064.0,
      "step": 62960
    },
    {
      "entropy": 1.9283981889486312,
      "epoch": 0.1952014904379324,
      "grad_norm": 4.0578155517578125,
      "learning_rate": 5.726042449523786e-06,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8367056354880333,
      "num_tokens": 75732835.0,
      "step": 62970
    },
    {
      "entropy": 1.9276798009872436,
      "epoch": 0.1952324895629821,
      "grad_norm": 7.493443489074707,
      "learning_rate": 5.725587832130112e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.841886368393898,
      "num_tokens": 75743432.0,
      "step": 62980
    },
    {
      "entropy": 1.7954487293958663,
      "epoch": 0.19526348868803178,
      "grad_norm": 8.081300735473633,
      "learning_rate": 5.725133323001873e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8556428179144859,
      "num_tokens": 75757107.0,
      "step": 62990
    },
    {
      "entropy": 1.8096677586436272,
      "epoch": 0.19529448781308148,
      "grad_norm": 9.376117706298828,
      "learning_rate": 5.724678922096108e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8452500954270363,
      "num_tokens": 75769351.0,
      "step": 63000
    },
    {
      "entropy": 1.8731877133250237,
      "epoch": 0.19532548693813118,
      "grad_norm": 9.466216087341309,
      "learning_rate": 5.724224629369872e-06,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8412355482578278,
      "num_tokens": 75781771.0,
      "step": 63010
    },
    {
      "entropy": 1.8507128790020944,
      "epoch": 0.19535648606318087,
      "grad_norm": 10.853748321533203,
      "learning_rate": 5.72377044478025e-06,
      "loss": 0.498,
      "mean_token_accuracy": 0.8409447997808457,
      "num_tokens": 75794706.0,
      "step": 63020
    },
    {
      "entropy": 1.9449727565050126,
      "epoch": 0.19538748518823057,
      "grad_norm": 6.4505085945129395,
      "learning_rate": 5.723316368284348e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8548792108893395,
      "num_tokens": 75805711.0,
      "step": 63030
    },
    {
      "entropy": 1.8775279074907303,
      "epoch": 0.19541848431328027,
      "grad_norm": 7.87474250793457,
      "learning_rate": 5.722862399839298e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8526403203606605,
      "num_tokens": 75817522.0,
      "step": 63040
    },
    {
      "entropy": 1.856291452050209,
      "epoch": 0.19544948343832996,
      "grad_norm": 4.191627025604248,
      "learning_rate": 5.722408539402254e-06,
      "loss": 0.4899,
      "mean_token_accuracy": 0.8563273102045059,
      "num_tokens": 75829497.0,
      "step": 63050
    },
    {
      "entropy": 1.9053906798362732,
      "epoch": 0.19548048256337966,
      "grad_norm": 8.759142875671387,
      "learning_rate": 5.72195478693039e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8432946696877479,
      "num_tokens": 75840759.0,
      "step": 63060
    },
    {
      "entropy": 1.907120332121849,
      "epoch": 0.19551148168842936,
      "grad_norm": 9.412018775939941,
      "learning_rate": 5.721501142380913e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8373003304004669,
      "num_tokens": 75852503.0,
      "step": 63070
    },
    {
      "entropy": 2.0056669265031815,
      "epoch": 0.19554248081347905,
      "grad_norm": 7.475878715515137,
      "learning_rate": 5.721047605711047e-06,
      "loss": 0.57,
      "mean_token_accuracy": 0.8297986865043641,
      "num_tokens": 75863521.0,
      "step": 63080
    },
    {
      "entropy": 1.922423042356968,
      "epoch": 0.19557347993852872,
      "grad_norm": 4.381282329559326,
      "learning_rate": 5.720594176878039e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8460247412323951,
      "num_tokens": 75874691.0,
      "step": 63090
    },
    {
      "entropy": 1.9274744465947151,
      "epoch": 0.19560447906357842,
      "grad_norm": 10.508957862854004,
      "learning_rate": 5.720140855839166e-06,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8317728966474534,
      "num_tokens": 75887760.0,
      "step": 63100
    },
    {
      "entropy": 1.9512497708201408,
      "epoch": 0.19563547818862811,
      "grad_norm": 9.120226860046387,
      "learning_rate": 5.719687642551722e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8364842370152473,
      "num_tokens": 75899765.0,
      "step": 63110
    },
    {
      "entropy": 1.925400537252426,
      "epoch": 0.1956664773136778,
      "grad_norm": 5.570336818695068,
      "learning_rate": 5.719234536973028e-06,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8405450001358986,
      "num_tokens": 75911921.0,
      "step": 63120
    },
    {
      "entropy": 1.836472137272358,
      "epoch": 0.1956974764387275,
      "grad_norm": 2.734543561935425,
      "learning_rate": 5.718781539060429e-06,
      "loss": 0.4246,
      "mean_token_accuracy": 0.851534178853035,
      "num_tokens": 75925915.0,
      "step": 63130
    },
    {
      "entropy": 1.8467722043395043,
      "epoch": 0.1957284755637772,
      "grad_norm": 8.128966331481934,
      "learning_rate": 5.718328648771291e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.851094801723957,
      "num_tokens": 75938929.0,
      "step": 63140
    },
    {
      "entropy": 1.9085650324821473,
      "epoch": 0.1957594746888269,
      "grad_norm": 8.325636863708496,
      "learning_rate": 5.717875866063008e-06,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8383592694997788,
      "num_tokens": 75949645.0,
      "step": 63150
    },
    {
      "entropy": 1.8661453172564506,
      "epoch": 0.1957904738138766,
      "grad_norm": 4.130495548248291,
      "learning_rate": 5.717423190892991e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8449689701199532,
      "num_tokens": 75962507.0,
      "step": 63160
    },
    {
      "entropy": 1.9401205703616142,
      "epoch": 0.1958214729389263,
      "grad_norm": 10.611771583557129,
      "learning_rate": 5.716970623218681e-06,
      "loss": 0.5661,
      "mean_token_accuracy": 0.8341074407100677,
      "num_tokens": 75974430.0,
      "step": 63170
    },
    {
      "entropy": 1.834306775033474,
      "epoch": 0.195852472063976,
      "grad_norm": 8.28214168548584,
      "learning_rate": 5.716518162997542e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8515095382928848,
      "num_tokens": 75987170.0,
      "step": 63180
    },
    {
      "entropy": 1.8667117178440094,
      "epoch": 0.1958834711890257,
      "grad_norm": 4.622128963470459,
      "learning_rate": 5.716065810187056e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8383040621876716,
      "num_tokens": 75999364.0,
      "step": 63190
    },
    {
      "entropy": 1.9117874279618263,
      "epoch": 0.19591447031407538,
      "grad_norm": 7.746886730194092,
      "learning_rate": 5.715613564744734e-06,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8401778340339661,
      "num_tokens": 76010533.0,
      "step": 63200
    },
    {
      "entropy": 1.860894750058651,
      "epoch": 0.19594546943912508,
      "grad_norm": 3.8454928398132324,
      "learning_rate": 5.715161426628111e-06,
      "loss": 0.4673,
      "mean_token_accuracy": 0.847791762650013,
      "num_tokens": 76023243.0,
      "step": 63210
    },
    {
      "entropy": 1.8731217697262763,
      "epoch": 0.19597646856417478,
      "grad_norm": 4.6532745361328125,
      "learning_rate": 5.71470939579474e-06,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8273180708289146,
      "num_tokens": 76035128.0,
      "step": 63220
    },
    {
      "entropy": 1.892588709294796,
      "epoch": 0.19600746768922447,
      "grad_norm": 10.033417701721191,
      "learning_rate": 5.714257472202201e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8375523954629898,
      "num_tokens": 76047088.0,
      "step": 63230
    },
    {
      "entropy": 1.8054717764258386,
      "epoch": 0.19603846681427417,
      "grad_norm": 4.947404384613037,
      "learning_rate": 5.7138056558080975e-06,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8533313110470772,
      "num_tokens": 76060229.0,
      "step": 63240
    },
    {
      "entropy": 1.9267044529318809,
      "epoch": 0.19606946593932387,
      "grad_norm": 8.534299850463867,
      "learning_rate": 5.713353946570057e-06,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8383896484971046,
      "num_tokens": 76072124.0,
      "step": 63250
    },
    {
      "entropy": 1.9226173907518387,
      "epoch": 0.19610046506437356,
      "grad_norm": 8.385056495666504,
      "learning_rate": 5.712902344445731e-06,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8362686946988106,
      "num_tokens": 76083932.0,
      "step": 63260
    },
    {
      "entropy": 1.8576195508241653,
      "epoch": 0.19613146418942326,
      "grad_norm": 9.054718017578125,
      "learning_rate": 5.712450849392791e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8375715285539627,
      "num_tokens": 76096111.0,
      "step": 63270
    },
    {
      "entropy": 1.9397896856069565,
      "epoch": 0.19616246331447296,
      "grad_norm": 10.121234893798828,
      "learning_rate": 5.711999461368935e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.834457078576088,
      "num_tokens": 76107744.0,
      "step": 63280
    },
    {
      "entropy": 1.8755070850253106,
      "epoch": 0.19619346243952265,
      "grad_norm": 9.817523956298828,
      "learning_rate": 5.711548180331882e-06,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8419964432716369,
      "num_tokens": 76119347.0,
      "step": 63290
    },
    {
      "entropy": 1.915390780568123,
      "epoch": 0.19622446156457235,
      "grad_norm": 7.848161697387695,
      "learning_rate": 5.711097006239378e-06,
      "loss": 0.56,
      "mean_token_accuracy": 0.8367161169648171,
      "num_tokens": 76130270.0,
      "step": 63300
    },
    {
      "entropy": 1.9267538145184517,
      "epoch": 0.19625546068962205,
      "grad_norm": 8.433647155761719,
      "learning_rate": 5.710645939049189e-06,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8365942299365997,
      "num_tokens": 76141625.0,
      "step": 63310
    },
    {
      "entropy": 1.9020789548754693,
      "epoch": 0.19628645981467174,
      "grad_norm": 7.910606861114502,
      "learning_rate": 5.710194978719105e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.838070061802864,
      "num_tokens": 76153615.0,
      "step": 63320
    },
    {
      "entropy": 1.8399833947420121,
      "epoch": 0.19631745893972144,
      "grad_norm": 9.023702621459961,
      "learning_rate": 5.70974412520694e-06,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8374461650848388,
      "num_tokens": 76166808.0,
      "step": 63330
    },
    {
      "entropy": 1.8543090134859086,
      "epoch": 0.1963484580647711,
      "grad_norm": 4.479185581207275,
      "learning_rate": 5.709293378470532e-06,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8332849636673927,
      "num_tokens": 76180101.0,
      "step": 63340
    },
    {
      "entropy": 1.9013725534081458,
      "epoch": 0.1963794571898208,
      "grad_norm": 9.714118003845215,
      "learning_rate": 5.70884273846774e-06,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8363057851791382,
      "num_tokens": 76191945.0,
      "step": 63350
    },
    {
      "entropy": 1.764454497396946,
      "epoch": 0.1964104563148705,
      "grad_norm": 9.242158889770508,
      "learning_rate": 5.7083922051564485e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8551331639289856,
      "num_tokens": 76205433.0,
      "step": 63360
    },
    {
      "entropy": 1.8029524132609367,
      "epoch": 0.1964414554399202,
      "grad_norm": 9.247615814208984,
      "learning_rate": 5.707941778494567e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8530771881341934,
      "num_tokens": 76218558.0,
      "step": 63370
    },
    {
      "entropy": 1.8929936558008194,
      "epoch": 0.1964724545649699,
      "grad_norm": 8.096500396728516,
      "learning_rate": 5.707491458440021e-06,
      "loss": 0.538,
      "mean_token_accuracy": 0.8339035794138908,
      "num_tokens": 76229934.0,
      "step": 63380
    },
    {
      "entropy": 1.92166768014431,
      "epoch": 0.1965034536900196,
      "grad_norm": 5.686055660247803,
      "learning_rate": 5.707041244950767e-06,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8357293531298637,
      "num_tokens": 76241582.0,
      "step": 63390
    },
    {
      "entropy": 1.9319213822484016,
      "epoch": 0.1965344528150693,
      "grad_norm": 9.794820785522461,
      "learning_rate": 5.706591137984782e-06,
      "loss": 0.5742,
      "mean_token_accuracy": 0.8362261816859246,
      "num_tokens": 76253353.0,
      "step": 63400
    },
    {
      "entropy": 1.8578762322664262,
      "epoch": 0.19656545194011898,
      "grad_norm": 7.473123550415039,
      "learning_rate": 5.706141137500063e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8495231464505195,
      "num_tokens": 76266223.0,
      "step": 63410
    },
    {
      "entropy": 1.897558230161667,
      "epoch": 0.19659645106516868,
      "grad_norm": 8.56474494934082,
      "learning_rate": 5.705691243454636e-06,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8442750841379165,
      "num_tokens": 76278256.0,
      "step": 63420
    },
    {
      "entropy": 1.8700624421238898,
      "epoch": 0.19662745019021838,
      "grad_norm": 4.685993194580078,
      "learning_rate": 5.705241455806546e-06,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8440624356269837,
      "num_tokens": 76290947.0,
      "step": 63430
    },
    {
      "entropy": 1.9605732291936875,
      "epoch": 0.19665844931526807,
      "grad_norm": 9.56114387512207,
      "learning_rate": 5.704791774513863e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8460546612739563,
      "num_tokens": 76302224.0,
      "step": 63440
    },
    {
      "entropy": 1.9243500038981438,
      "epoch": 0.19668944844031777,
      "grad_norm": 4.1643195152282715,
      "learning_rate": 5.704342199534677e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8379947647452355,
      "num_tokens": 76314456.0,
      "step": 63450
    },
    {
      "entropy": 1.9549641892313958,
      "epoch": 0.19672044756536747,
      "grad_norm": 10.69619083404541,
      "learning_rate": 5.703892730827107e-06,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8345342621207237,
      "num_tokens": 76325957.0,
      "step": 63460
    },
    {
      "entropy": 1.9176968559622765,
      "epoch": 0.19675144669041716,
      "grad_norm": 12.198366165161133,
      "learning_rate": 5.703443368349289e-06,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8376664772629738,
      "num_tokens": 76337257.0,
      "step": 63470
    },
    {
      "entropy": 1.8076652333140373,
      "epoch": 0.19678244581546686,
      "grad_norm": 9.941756248474121,
      "learning_rate": 5.7029941120593864e-06,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8417710468173027,
      "num_tokens": 76350461.0,
      "step": 63480
    },
    {
      "entropy": 1.8378579676151277,
      "epoch": 0.19681344494051656,
      "grad_norm": 8.623409271240234,
      "learning_rate": 5.702544961915585e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8429977357387543,
      "num_tokens": 76363284.0,
      "step": 63490
    },
    {
      "entropy": 1.9412892490625382,
      "epoch": 0.19684444406556625,
      "grad_norm": 10.340338706970215,
      "learning_rate": 5.702095917876089e-06,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8276393041014671,
      "num_tokens": 76374739.0,
      "step": 63500
    },
    {
      "entropy": 1.9820778489112854,
      "epoch": 0.19687544319061595,
      "grad_norm": 9.503996849060059,
      "learning_rate": 5.701646979899134e-06,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8296515002846718,
      "num_tokens": 76386166.0,
      "step": 63510
    },
    {
      "entropy": 1.8864339843392373,
      "epoch": 0.19690644231566565,
      "grad_norm": 8.490914344787598,
      "learning_rate": 5.70119814794297e-06,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8448032855987548,
      "num_tokens": 76399048.0,
      "step": 63520
    },
    {
      "entropy": 1.9335085853934288,
      "epoch": 0.19693744144071534,
      "grad_norm": 11.991941452026367,
      "learning_rate": 5.700749421965877e-06,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8254455342888832,
      "num_tokens": 76410513.0,
      "step": 63530
    },
    {
      "entropy": 1.9954243630170823,
      "epoch": 0.19696844056576504,
      "grad_norm": 9.641355514526367,
      "learning_rate": 5.700300801926151e-06,
      "loss": 0.5682,
      "mean_token_accuracy": 0.820628535747528,
      "num_tokens": 76422104.0,
      "step": 63540
    },
    {
      "entropy": 1.879958561062813,
      "epoch": 0.19699943969081474,
      "grad_norm": 4.384090423583984,
      "learning_rate": 5.6998522877821185e-06,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8393704131245613,
      "num_tokens": 76434124.0,
      "step": 63550
    },
    {
      "entropy": 1.8802961066365242,
      "epoch": 0.19703043881586443,
      "grad_norm": 6.248841762542725,
      "learning_rate": 5.699403879492125e-06,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8394171461462975,
      "num_tokens": 76446867.0,
      "step": 63560
    },
    {
      "entropy": 1.9431464537978171,
      "epoch": 0.19706143794091413,
      "grad_norm": 9.1453857421875,
      "learning_rate": 5.6989555770145386e-06,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8342863008379936,
      "num_tokens": 76458179.0,
      "step": 63570
    },
    {
      "entropy": 1.8758263066411018,
      "epoch": 0.19709243706596383,
      "grad_norm": 8.520499229431152,
      "learning_rate": 5.69850738030775e-06,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8425302848219871,
      "num_tokens": 76469611.0,
      "step": 63580
    },
    {
      "entropy": 1.9540351748466491,
      "epoch": 0.1971234361910135,
      "grad_norm": 8.32021713256836,
      "learning_rate": 5.698059289330175e-06,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8359291762113571,
      "num_tokens": 76481274.0,
      "step": 63590
    },
    {
      "entropy": 1.8006914272904395,
      "epoch": 0.1971544353160632,
      "grad_norm": 4.276872634887695,
      "learning_rate": 5.69761130404025e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8568015277385712,
      "num_tokens": 76494249.0,
      "step": 63600
    },
    {
      "entropy": 1.9028367415070533,
      "epoch": 0.1971854344411129,
      "grad_norm": 9.324864387512207,
      "learning_rate": 5.6971634243964366e-06,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8430081695318222,
      "num_tokens": 76506213.0,
      "step": 63610
    },
    {
      "entropy": 1.8858824104070664,
      "epoch": 0.19721643356616259,
      "grad_norm": 10.004111289978027,
      "learning_rate": 5.696715650357217e-06,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8416607335209847,
      "num_tokens": 76519093.0,
      "step": 63620
    },
    {
      "entropy": 1.8651671513915062,
      "epoch": 0.19724743269121228,
      "grad_norm": 8.351085662841797,
      "learning_rate": 5.696267981881099e-06,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8429739147424697,
      "num_tokens": 76531349.0,
      "step": 63630
    },
    {
      "entropy": 1.9040121346712113,
      "epoch": 0.19727843181626198,
      "grad_norm": 8.7483549118042,
      "learning_rate": 5.69582041892661e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8458465442061425,
      "num_tokens": 76542921.0,
      "step": 63640
    },
    {
      "entropy": 1.8771297112107277,
      "epoch": 0.19730943094131168,
      "grad_norm": 7.3089141845703125,
      "learning_rate": 5.695372961452301e-06,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8363812834024429,
      "num_tokens": 76555655.0,
      "step": 63650
    },
    {
      "entropy": 1.8038627937436105,
      "epoch": 0.19734043006636137,
      "grad_norm": 7.573980808258057,
      "learning_rate": 5.694925609416748e-06,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8495232254266739,
      "num_tokens": 76569165.0,
      "step": 63660
    },
    {
      "entropy": 1.8558023154735566,
      "epoch": 0.19737142919141107,
      "grad_norm": 8.455429077148438,
      "learning_rate": 5.694478362778547e-06,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8412219509482384,
      "num_tokens": 76581380.0,
      "step": 63670
    },
    {
      "entropy": 1.8561557665467263,
      "epoch": 0.19740242831646077,
      "grad_norm": 3.842336416244507,
      "learning_rate": 5.69403122149632e-06,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8478426292538643,
      "num_tokens": 76594529.0,
      "step": 63680
    },
    {
      "entropy": 1.85400443226099,
      "epoch": 0.19743342744151046,
      "grad_norm": 4.739361763000488,
      "learning_rate": 5.693584185528707e-06,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8424937948584557,
      "num_tokens": 76607350.0,
      "step": 63690
    },
    {
      "entropy": 1.8113030225038529,
      "epoch": 0.19746442656656016,
      "grad_norm": 8.30618953704834,
      "learning_rate": 5.693137254834375e-06,
      "loss": 0.398,
      "mean_token_accuracy": 0.8603976845741272,
      "num_tokens": 76620524.0,
      "step": 63700
    },
    {
      "entropy": 1.853132924437523,
      "epoch": 0.19749542569160985,
      "grad_norm": 9.655352592468262,
      "learning_rate": 5.692690429372012e-06,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8338514223694802,
      "num_tokens": 76633444.0,
      "step": 63710
    },
    {
      "entropy": 1.898238630592823,
      "epoch": 0.19752642481665955,
      "grad_norm": 4.058250904083252,
      "learning_rate": 5.692243709100329e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8403716444969177,
      "num_tokens": 76645664.0,
      "step": 63720
    },
    {
      "entropy": 1.882179079949856,
      "epoch": 0.19755742394170925,
      "grad_norm": 5.784068584442139,
      "learning_rate": 5.691797093978061e-06,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8447945788502693,
      "num_tokens": 76657600.0,
      "step": 63730
    },
    {
      "entropy": 1.8364885538816451,
      "epoch": 0.19758842306675894,
      "grad_norm": 5.507530689239502,
      "learning_rate": 5.69135058396396e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8419953018426896,
      "num_tokens": 76671201.0,
      "step": 63740
    },
    {
      "entropy": 1.8077232390642166,
      "epoch": 0.19761942219180864,
      "grad_norm": 8.423868179321289,
      "learning_rate": 5.690904179016809e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8521018475294113,
      "num_tokens": 76684611.0,
      "step": 63750
    },
    {
      "entropy": 1.8499022141098975,
      "epoch": 0.19765042131685834,
      "grad_norm": 5.460327625274658,
      "learning_rate": 5.690457879095407e-06,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8433385342359543,
      "num_tokens": 76697424.0,
      "step": 63760
    },
    {
      "entropy": 1.8565186977386474,
      "epoch": 0.19768142044190803,
      "grad_norm": 10.535595893859863,
      "learning_rate": 5.690011684158579e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8477386757731438,
      "num_tokens": 76710657.0,
      "step": 63770
    },
    {
      "entropy": 1.9264386996626854,
      "epoch": 0.19771241956695773,
      "grad_norm": 8.825230598449707,
      "learning_rate": 5.689565594165174e-06,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8357585594058037,
      "num_tokens": 76721904.0,
      "step": 63780
    },
    {
      "entropy": 1.9114050999283791,
      "epoch": 0.19774341869200743,
      "grad_norm": 10.269311904907227,
      "learning_rate": 5.689119609074057e-06,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8319689840078354,
      "num_tokens": 76734403.0,
      "step": 63790
    },
    {
      "entropy": 1.8821352809667586,
      "epoch": 0.19777441781705712,
      "grad_norm": 8.149957656860352,
      "learning_rate": 5.688673728844121e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8380433395504951,
      "num_tokens": 76745913.0,
      "step": 63800
    },
    {
      "entropy": 1.9270009264349937,
      "epoch": 0.19780541694210682,
      "grad_norm": 8.53276252746582,
      "learning_rate": 5.688227953434282e-06,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8341132417321205,
      "num_tokens": 76758099.0,
      "step": 63810
    },
    {
      "entropy": 1.8181735575199127,
      "epoch": 0.19783641606715652,
      "grad_norm": 7.680285930633545,
      "learning_rate": 5.687782282803477e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8435733437538147,
      "num_tokens": 76771266.0,
      "step": 63820
    },
    {
      "entropy": 1.8427308067679404,
      "epoch": 0.1978674151922062,
      "grad_norm": 3.6180124282836914,
      "learning_rate": 5.687336716910663e-06,
      "loss": 0.4316,
      "mean_token_accuracy": 0.8437779262661934,
      "num_tokens": 76783158.0,
      "step": 63830
    },
    {
      "entropy": 1.9142847254872322,
      "epoch": 0.19789841431725588,
      "grad_norm": 4.747416019439697,
      "learning_rate": 5.6868912557148245e-06,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8309885829687118,
      "num_tokens": 76795826.0,
      "step": 63840
    },
    {
      "entropy": 1.8509532332420349,
      "epoch": 0.19792941344230558,
      "grad_norm": 8.225378036499023,
      "learning_rate": 5.686445899174965e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8449263677001,
      "num_tokens": 76808003.0,
      "step": 63850
    },
    {
      "entropy": 1.9634561657905578,
      "epoch": 0.19796041256735528,
      "grad_norm": 8.663080215454102,
      "learning_rate": 5.686000647250109e-06,
      "loss": 0.583,
      "mean_token_accuracy": 0.8314743667840958,
      "num_tokens": 76819491.0,
      "step": 63860
    },
    {
      "entropy": 1.9227616682648658,
      "epoch": 0.19799141169240497,
      "grad_norm": 7.894753932952881,
      "learning_rate": 5.685555499899311e-06,
      "loss": 0.5717,
      "mean_token_accuracy": 0.831498672068119,
      "num_tokens": 76831514.0,
      "step": 63870
    },
    {
      "entropy": 1.969376365840435,
      "epoch": 0.19802241081745467,
      "grad_norm": 9.801876068115234,
      "learning_rate": 5.685110457081639e-06,
      "loss": 0.5539,
      "mean_token_accuracy": 0.8390384584665298,
      "num_tokens": 76842781.0,
      "step": 63880
    },
    {
      "entropy": 1.9508319780230523,
      "epoch": 0.19805340994250437,
      "grad_norm": 11.13720989227295,
      "learning_rate": 5.6846655187561874e-06,
      "loss": 0.5257,
      "mean_token_accuracy": 0.835049818456173,
      "num_tokens": 76854513.0,
      "step": 63890
    },
    {
      "entropy": 1.9810258507728578,
      "epoch": 0.19808440906755406,
      "grad_norm": 9.226847648620605,
      "learning_rate": 5.684220684882074e-06,
      "loss": 0.5617,
      "mean_token_accuracy": 0.8322012394666671,
      "num_tokens": 76865518.0,
      "step": 63900
    },
    {
      "entropy": 1.9172653019428254,
      "epoch": 0.19811540819260376,
      "grad_norm": 4.333598613739014,
      "learning_rate": 5.683775955418437e-06,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8240251198410988,
      "num_tokens": 76877792.0,
      "step": 63910
    },
    {
      "entropy": 1.843831568211317,
      "epoch": 0.19814640731765346,
      "grad_norm": 9.46835708618164,
      "learning_rate": 5.683331330324438e-06,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8351884678006172,
      "num_tokens": 76890123.0,
      "step": 63920
    },
    {
      "entropy": 1.7964137971401215,
      "epoch": 0.19817740644270315,
      "grad_norm": 8.710184097290039,
      "learning_rate": 5.682886809559261e-06,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8368506744503975,
      "num_tokens": 76902945.0,
      "step": 63930
    },
    {
      "entropy": 1.9574948653578759,
      "epoch": 0.19820840556775285,
      "grad_norm": 8.956100463867188,
      "learning_rate": 5.682442393082113e-06,
      "loss": 0.528,
      "mean_token_accuracy": 0.8361043930053711,
      "num_tokens": 76914195.0,
      "step": 63940
    },
    {
      "entropy": 1.8992867067456245,
      "epoch": 0.19823940469280255,
      "grad_norm": 5.019556045532227,
      "learning_rate": 5.681998080852219e-06,
      "loss": 0.5534,
      "mean_token_accuracy": 0.8368826553225517,
      "num_tokens": 76925863.0,
      "step": 63950
    },
    {
      "entropy": 1.9133247062563896,
      "epoch": 0.19827040381785224,
      "grad_norm": 4.297521591186523,
      "learning_rate": 5.681553872828835e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8392513215541839,
      "num_tokens": 76937635.0,
      "step": 63960
    },
    {
      "entropy": 1.9356270998716354,
      "epoch": 0.19830140294290194,
      "grad_norm": 7.358262062072754,
      "learning_rate": 5.681109768971228e-06,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8359444841742516,
      "num_tokens": 76948918.0,
      "step": 63970
    },
    {
      "entropy": 1.807618674635887,
      "epoch": 0.19833240206795164,
      "grad_norm": 8.462199211120605,
      "learning_rate": 5.680665769238698e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.844212980568409,
      "num_tokens": 76962029.0,
      "step": 63980
    },
    {
      "entropy": 1.8870945870876312,
      "epoch": 0.19836340119300133,
      "grad_norm": 9.050925254821777,
      "learning_rate": 5.68022187359056e-06,
      "loss": 0.4935,
      "mean_token_accuracy": 0.836103081703186,
      "num_tokens": 76973718.0,
      "step": 63990
    },
    {
      "entropy": 1.8385724887251853,
      "epoch": 0.19839440031805103,
      "grad_norm": 9.02990436553955,
      "learning_rate": 5.6797780819861545e-06,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8344948351383209,
      "num_tokens": 76986094.0,
      "step": 64000
    },
    {
      "entropy": 1.8950312197208405,
      "epoch": 0.19842539944310073,
      "grad_norm": 17.118181228637695,
      "learning_rate": 5.679334394384845e-06,
      "loss": 0.4611,
      "mean_token_accuracy": 0.846867561340332,
      "num_tokens": 76998477.0,
      "step": 64010
    },
    {
      "entropy": 1.9193229526281357,
      "epoch": 0.19845639856815042,
      "grad_norm": 10.047476768493652,
      "learning_rate": 5.6788908107460135e-06,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8332524642348289,
      "num_tokens": 77009850.0,
      "step": 64020
    },
    {
      "entropy": 1.8773248583078384,
      "epoch": 0.19848739769320012,
      "grad_norm": 8.258344650268555,
      "learning_rate": 5.678447331029068e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8483355283737183,
      "num_tokens": 77022417.0,
      "step": 64030
    },
    {
      "entropy": 1.9686390221118928,
      "epoch": 0.19851839681824981,
      "grad_norm": 9.272161483764648,
      "learning_rate": 5.678003955193437e-06,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8388251766562462,
      "num_tokens": 77033370.0,
      "step": 64040
    },
    {
      "entropy": 1.8505666583776474,
      "epoch": 0.1985493959432995,
      "grad_norm": 8.902565956115723,
      "learning_rate": 5.677560683198569e-06,
      "loss": 0.5142,
      "mean_token_accuracy": 0.840177808701992,
      "num_tokens": 77046098.0,
      "step": 64050
    },
    {
      "entropy": 1.9229148238897324,
      "epoch": 0.1985803950683492,
      "grad_norm": 7.94978141784668,
      "learning_rate": 5.677117515003942e-06,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8431189730763435,
      "num_tokens": 77058145.0,
      "step": 64060
    },
    {
      "entropy": 1.8256406679749488,
      "epoch": 0.1986113941933989,
      "grad_norm": 5.319147109985352,
      "learning_rate": 5.676674450569045e-06,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8440526649355888,
      "num_tokens": 77071645.0,
      "step": 64070
    },
    {
      "entropy": 1.8413712576031684,
      "epoch": 0.19864239331844857,
      "grad_norm": 7.502344131469727,
      "learning_rate": 5.6762314898534e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8615258172154426,
      "num_tokens": 77083570.0,
      "step": 64080
    },
    {
      "entropy": 1.8211113825440406,
      "epoch": 0.19867339244349827,
      "grad_norm": 4.350229740142822,
      "learning_rate": 5.675788632816544e-06,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8473315939307213,
      "num_tokens": 77096128.0,
      "step": 64090
    },
    {
      "entropy": 1.7832973524928093,
      "epoch": 0.19870439156854797,
      "grad_norm": 3.9118053913116455,
      "learning_rate": 5.67534587941804e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8490145295858383,
      "num_tokens": 77109619.0,
      "step": 64100
    },
    {
      "entropy": 1.8864678606390952,
      "epoch": 0.19873539069359766,
      "grad_norm": 4.381364345550537,
      "learning_rate": 5.674903229617469e-06,
      "loss": 0.5189,
      "mean_token_accuracy": 0.8414922654628754,
      "num_tokens": 77121495.0,
      "step": 64110
    },
    {
      "entropy": 1.8282597333192825,
      "epoch": 0.19876638981864736,
      "grad_norm": 8.134101867675781,
      "learning_rate": 5.67446068337444e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8438005477190018,
      "num_tokens": 77133762.0,
      "step": 64120
    },
    {
      "entropy": 1.891626113653183,
      "epoch": 0.19879738894369706,
      "grad_norm": 8.915387153625488,
      "learning_rate": 5.674018240648578e-06,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8532876297831535,
      "num_tokens": 77145322.0,
      "step": 64130
    },
    {
      "entropy": 1.8996186777949333,
      "epoch": 0.19882838806874675,
      "grad_norm": 9.293252944946289,
      "learning_rate": 5.673575901399533e-06,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8366784289479255,
      "num_tokens": 77157517.0,
      "step": 64140
    },
    {
      "entropy": 1.8965847790241241,
      "epoch": 0.19885938719379645,
      "grad_norm": 4.745813846588135,
      "learning_rate": 5.673133665586977e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8319694399833679,
      "num_tokens": 77169583.0,
      "step": 64150
    },
    {
      "entropy": 1.8250271826982498,
      "epoch": 0.19889038631884615,
      "grad_norm": 9.300322532653809,
      "learning_rate": 5.672691533170605e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8459231272339821,
      "num_tokens": 77182323.0,
      "step": 64160
    },
    {
      "entropy": 1.7872568264603614,
      "epoch": 0.19892138544389584,
      "grad_norm": 9.027938842773438,
      "learning_rate": 5.672249504110131e-06,
      "loss": 0.4101,
      "mean_token_accuracy": 0.8547685459256172,
      "num_tokens": 77196100.0,
      "step": 64170
    },
    {
      "entropy": 1.814260269701481,
      "epoch": 0.19895238456894554,
      "grad_norm": 8.810227394104004,
      "learning_rate": 5.671807578365294e-06,
      "loss": 0.4367,
      "mean_token_accuracy": 0.849295774102211,
      "num_tokens": 77208500.0,
      "step": 64180
    },
    {
      "entropy": 1.8780076310038567,
      "epoch": 0.19898338369399524,
      "grad_norm": 9.624731063842773,
      "learning_rate": 5.671365755895851e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8416895166039466,
      "num_tokens": 77220028.0,
      "step": 64190
    },
    {
      "entropy": 1.9330232828855514,
      "epoch": 0.19901438281904493,
      "grad_norm": 9.217514038085938,
      "learning_rate": 5.670924036661586e-06,
      "loss": 0.5643,
      "mean_token_accuracy": 0.8271647378802299,
      "num_tokens": 77231617.0,
      "step": 64200
    },
    {
      "entropy": 1.9576894700527192,
      "epoch": 0.19904538194409463,
      "grad_norm": 7.6577534675598145,
      "learning_rate": 5.670482420622302e-06,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8377937138080597,
      "num_tokens": 77243109.0,
      "step": 64210
    },
    {
      "entropy": 1.9337700963020326,
      "epoch": 0.19907638106914433,
      "grad_norm": 9.766473770141602,
      "learning_rate": 5.6700409077378235e-06,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8324208498001099,
      "num_tokens": 77253457.0,
      "step": 64220
    },
    {
      "entropy": 1.888157394528389,
      "epoch": 0.19910738019419402,
      "grad_norm": 8.229586601257324,
      "learning_rate": 5.6695994979679995e-06,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8483126178383827,
      "num_tokens": 77265767.0,
      "step": 64230
    },
    {
      "entropy": 1.795005053281784,
      "epoch": 0.19913837931924372,
      "grad_norm": 8.953361511230469,
      "learning_rate": 5.669158191272697e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8460436001420021,
      "num_tokens": 77278828.0,
      "step": 64240
    },
    {
      "entropy": 1.8841128557920457,
      "epoch": 0.19916937844429342,
      "grad_norm": 9.218796730041504,
      "learning_rate": 5.668716987611807e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8368963778018952,
      "num_tokens": 77291261.0,
      "step": 64250
    },
    {
      "entropy": 1.816078770160675,
      "epoch": 0.1992003775693431,
      "grad_norm": 7.860946178436279,
      "learning_rate": 5.668275886945246e-06,
      "loss": 0.4316,
      "mean_token_accuracy": 0.8497811511158944,
      "num_tokens": 77303890.0,
      "step": 64260
    },
    {
      "entropy": 1.9705949038267137,
      "epoch": 0.1992313766943928,
      "grad_norm": 7.762573719024658,
      "learning_rate": 5.667834889232945e-06,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8459170028567314,
      "num_tokens": 77315352.0,
      "step": 64270
    },
    {
      "entropy": 1.8340299040079118,
      "epoch": 0.1992623758194425,
      "grad_norm": 5.21130895614624,
      "learning_rate": 5.667393994434861e-06,
      "loss": 0.49,
      "mean_token_accuracy": 0.8440465152263641,
      "num_tokens": 77328106.0,
      "step": 64280
    },
    {
      "entropy": 1.872594805061817,
      "epoch": 0.1992933749444922,
      "grad_norm": 7.447837829589844,
      "learning_rate": 5.666953202510973e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8528081148862838,
      "num_tokens": 77340307.0,
      "step": 64290
    },
    {
      "entropy": 1.9547280743718147,
      "epoch": 0.1993243740695419,
      "grad_norm": 8.335238456726074,
      "learning_rate": 5.666512513421281e-06,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8396565094590187,
      "num_tokens": 77351792.0,
      "step": 64300
    },
    {
      "entropy": 1.8686283484101296,
      "epoch": 0.1993553731945916,
      "grad_norm": 10.21964168548584,
      "learning_rate": 5.666071927125806e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8395275220274925,
      "num_tokens": 77364455.0,
      "step": 64310
    },
    {
      "entropy": 1.9559142783284187,
      "epoch": 0.1993863723196413,
      "grad_norm": 6.589775085449219,
      "learning_rate": 5.665631443584593e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8342868536710739,
      "num_tokens": 77376312.0,
      "step": 64320
    },
    {
      "entropy": 1.8677933678030967,
      "epoch": 0.19941737144469096,
      "grad_norm": 4.30719518661499,
      "learning_rate": 5.665191062757705e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8398402318358421,
      "num_tokens": 77388805.0,
      "step": 64330
    },
    {
      "entropy": 2.0073965549468995,
      "epoch": 0.19944837056974066,
      "grad_norm": 8.787602424621582,
      "learning_rate": 5.664750784605233e-06,
      "loss": 0.5647,
      "mean_token_accuracy": 0.8316746145486832,
      "num_tokens": 77399105.0,
      "step": 64340
    },
    {
      "entropy": 1.9256497889757156,
      "epoch": 0.19947936969479035,
      "grad_norm": 4.221436023712158,
      "learning_rate": 5.664310609087283e-06,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8293071076273918,
      "num_tokens": 77411168.0,
      "step": 64350
    },
    {
      "entropy": 1.8341520741581916,
      "epoch": 0.19951036881984005,
      "grad_norm": 7.341567516326904,
      "learning_rate": 5.663870536163986e-06,
      "loss": 0.5056,
      "mean_token_accuracy": 0.835312868654728,
      "num_tokens": 77424558.0,
      "step": 64360
    },
    {
      "entropy": 1.9391889229416848,
      "epoch": 0.19954136794488975,
      "grad_norm": 10.012337684631348,
      "learning_rate": 5.663430565795495e-06,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8299921050667762,
      "num_tokens": 77436576.0,
      "step": 64370
    },
    {
      "entropy": 1.9387200742959976,
      "epoch": 0.19957236706993944,
      "grad_norm": 8.041949272155762,
      "learning_rate": 5.6629906979419805e-06,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8453411236405373,
      "num_tokens": 77448293.0,
      "step": 64380
    },
    {
      "entropy": 1.845164766907692,
      "epoch": 0.19960336619498914,
      "grad_norm": 9.091232299804688,
      "learning_rate": 5.662550932563643e-06,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8558371141552925,
      "num_tokens": 77461006.0,
      "step": 64390
    },
    {
      "entropy": 1.9543328523635863,
      "epoch": 0.19963436532003884,
      "grad_norm": 8.760300636291504,
      "learning_rate": 5.662111269620696e-06,
      "loss": 0.546,
      "mean_token_accuracy": 0.8322904482483864,
      "num_tokens": 77472575.0,
      "step": 64400
    },
    {
      "entropy": 1.8849376797676087,
      "epoch": 0.19966536444508853,
      "grad_norm": 8.26484489440918,
      "learning_rate": 5.661671709073379e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8436549201607704,
      "num_tokens": 77484546.0,
      "step": 64410
    },
    {
      "entropy": 1.9597482338547707,
      "epoch": 0.19969636357013823,
      "grad_norm": 7.358891487121582,
      "learning_rate": 5.661232250881952e-06,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8410298019647598,
      "num_tokens": 77496212.0,
      "step": 64420
    },
    {
      "entropy": 1.9519696533679962,
      "epoch": 0.19972736269518793,
      "grad_norm": 8.28138542175293,
      "learning_rate": 5.660792895006698e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8325450897216797,
      "num_tokens": 77507649.0,
      "step": 64430
    },
    {
      "entropy": 1.9554157480597496,
      "epoch": 0.19975836182023762,
      "grad_norm": 8.739425659179688,
      "learning_rate": 5.660353641407921e-06,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8404718577861786,
      "num_tokens": 77518910.0,
      "step": 64440
    },
    {
      "entropy": 1.9898914635181426,
      "epoch": 0.19978936094528732,
      "grad_norm": 11.325966835021973,
      "learning_rate": 5.659914490045944e-06,
      "loss": 0.5837,
      "mean_token_accuracy": 0.8286274582147598,
      "num_tokens": 77529561.0,
      "step": 64450
    },
    {
      "entropy": 1.8334274768829346,
      "epoch": 0.19982036007033702,
      "grad_norm": 7.993432998657227,
      "learning_rate": 5.659475440881115e-06,
      "loss": 0.4159,
      "mean_token_accuracy": 0.8571745663881302,
      "num_tokens": 77542142.0,
      "step": 64460
    },
    {
      "entropy": 1.864787982404232,
      "epoch": 0.1998513591953867,
      "grad_norm": 8.196157455444336,
      "learning_rate": 5.659036493873802e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8399261072278023,
      "num_tokens": 77554999.0,
      "step": 64470
    },
    {
      "entropy": 1.9031756028532982,
      "epoch": 0.1998823583204364,
      "grad_norm": 7.0196404457092285,
      "learning_rate": 5.658597648984394e-06,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8386317491531372,
      "num_tokens": 77567455.0,
      "step": 64480
    },
    {
      "entropy": 1.8681576654314995,
      "epoch": 0.1999133574454861,
      "grad_norm": 8.093351364135742,
      "learning_rate": 5.658158906173302e-06,
      "loss": 0.406,
      "mean_token_accuracy": 0.8515056669712067,
      "num_tokens": 77580142.0,
      "step": 64490
    },
    {
      "entropy": 1.9180188179016113,
      "epoch": 0.1999443565705358,
      "grad_norm": 9.29157543182373,
      "learning_rate": 5.657720265400961e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8398342624306678,
      "num_tokens": 77591543.0,
      "step": 64500
    },
    {
      "entropy": 1.7902460798621178,
      "epoch": 0.1999753556955855,
      "grad_norm": 3.0435149669647217,
      "learning_rate": 5.657281726627822e-06,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8577824845910073,
      "num_tokens": 77604945.0,
      "step": 64510
    },
    {
      "entropy": 1.9697556897997857,
      "epoch": 0.2000063548206352,
      "grad_norm": 8.26711654663086,
      "learning_rate": 5.656843289814361e-06,
      "loss": 0.6193,
      "mean_token_accuracy": 0.8271805793046951,
      "num_tokens": 77616932.0,
      "step": 64520
    },
    {
      "entropy": 1.90864490121603,
      "epoch": 0.2000373539456849,
      "grad_norm": 9.074986457824707,
      "learning_rate": 5.656404954921076e-06,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8418331027030945,
      "num_tokens": 77629419.0,
      "step": 64530
    },
    {
      "entropy": 1.9975577175617218,
      "epoch": 0.2000683530707346,
      "grad_norm": 9.11250114440918,
      "learning_rate": 5.655966721908486e-06,
      "loss": 0.5851,
      "mean_token_accuracy": 0.8237589776515961,
      "num_tokens": 77640305.0,
      "step": 64540
    },
    {
      "entropy": 1.9538366571068764,
      "epoch": 0.20009935219578429,
      "grad_norm": 8.259376525878906,
      "learning_rate": 5.65552859073713e-06,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8349790692329406,
      "num_tokens": 77651850.0,
      "step": 64550
    },
    {
      "entropy": 1.8756514206528663,
      "epoch": 0.20013035132083398,
      "grad_norm": 9.364764213562012,
      "learning_rate": 5.655090561367568e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8493126779794693,
      "num_tokens": 77664419.0,
      "step": 64560
    },
    {
      "entropy": 1.908227115869522,
      "epoch": 0.20016135044588365,
      "grad_norm": 8.66217041015625,
      "learning_rate": 5.654652633760384e-06,
      "loss": 0.5649,
      "mean_token_accuracy": 0.8257618814706802,
      "num_tokens": 77676594.0,
      "step": 64570
    },
    {
      "entropy": 1.9556605026125908,
      "epoch": 0.20019234957093335,
      "grad_norm": 3.5400590896606445,
      "learning_rate": 5.654214807876182e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8475314304232597,
      "num_tokens": 77687943.0,
      "step": 64580
    },
    {
      "entropy": 1.883988819271326,
      "epoch": 0.20022334869598304,
      "grad_norm": 9.24106216430664,
      "learning_rate": 5.653777083675587e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8436937823891639,
      "num_tokens": 77700896.0,
      "step": 64590
    },
    {
      "entropy": 1.857344676554203,
      "epoch": 0.20025434782103274,
      "grad_norm": 9.264991760253906,
      "learning_rate": 5.653339461119245e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8411024749279022,
      "num_tokens": 77713541.0,
      "step": 64600
    },
    {
      "entropy": 1.9974335208535194,
      "epoch": 0.20028534694608244,
      "grad_norm": 9.765063285827637,
      "learning_rate": 5.6529019401678256e-06,
      "loss": 0.5565,
      "mean_token_accuracy": 0.8282542005181313,
      "num_tokens": 77724948.0,
      "step": 64610
    },
    {
      "entropy": 1.9897297486662864,
      "epoch": 0.20031634607113213,
      "grad_norm": 10.479876518249512,
      "learning_rate": 5.652464520782016e-06,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8397626534104348,
      "num_tokens": 77736336.0,
      "step": 64620
    },
    {
      "entropy": 1.8929444566369056,
      "epoch": 0.20034734519618183,
      "grad_norm": 9.795084953308105,
      "learning_rate": 5.652027202922528e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8592626452445984,
      "num_tokens": 77748048.0,
      "step": 64630
    },
    {
      "entropy": 1.814023308455944,
      "epoch": 0.20037834432123153,
      "grad_norm": 7.460301399230957,
      "learning_rate": 5.651589986550092e-06,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8527435094118119,
      "num_tokens": 77761289.0,
      "step": 64640
    },
    {
      "entropy": 1.8586131647229194,
      "epoch": 0.20040934344628122,
      "grad_norm": 8.590618133544922,
      "learning_rate": 5.6511528716254636e-06,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8501928374171257,
      "num_tokens": 77773602.0,
      "step": 64650
    },
    {
      "entropy": 1.8472480833530427,
      "epoch": 0.20044034257133092,
      "grad_norm": 3.6471304893493652,
      "learning_rate": 5.650715858109416e-06,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8560655087232589,
      "num_tokens": 77785559.0,
      "step": 64660
    },
    {
      "entropy": 2.008794406056404,
      "epoch": 0.20047134169638062,
      "grad_norm": 10.51722240447998,
      "learning_rate": 5.650278945962744e-06,
      "loss": 0.5934,
      "mean_token_accuracy": 0.8334228664636611,
      "num_tokens": 77796525.0,
      "step": 64670
    },
    {
      "entropy": 1.9870349198579789,
      "epoch": 0.2005023408214303,
      "grad_norm": 8.583792686462402,
      "learning_rate": 5.649842135146264e-06,
      "loss": 0.6073,
      "mean_token_accuracy": 0.8145452126860618,
      "num_tokens": 77807477.0,
      "step": 64680
    },
    {
      "entropy": 1.8656486958265304,
      "epoch": 0.20053333994648,
      "grad_norm": 8.072990417480469,
      "learning_rate": 5.649405425620815e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8582053184509277,
      "num_tokens": 77819825.0,
      "step": 64690
    },
    {
      "entropy": 1.8668094590306281,
      "epoch": 0.2005643390715297,
      "grad_norm": 9.32240104675293,
      "learning_rate": 5.648968817347257e-06,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8417691692709923,
      "num_tokens": 77832238.0,
      "step": 64700
    },
    {
      "entropy": 1.9135265216231345,
      "epoch": 0.2005953381965794,
      "grad_norm": 8.36419677734375,
      "learning_rate": 5.648532310286469e-06,
      "loss": 0.507,
      "mean_token_accuracy": 0.8405087172985077,
      "num_tokens": 77844237.0,
      "step": 64710
    },
    {
      "entropy": 2.0349139839410784,
      "epoch": 0.2006263373216291,
      "grad_norm": 8.855172157287598,
      "learning_rate": 5.648095904399352e-06,
      "loss": 0.6111,
      "mean_token_accuracy": 0.8124338254332543,
      "num_tokens": 77854626.0,
      "step": 64720
    },
    {
      "entropy": 1.838395781815052,
      "epoch": 0.2006573364466788,
      "grad_norm": 5.492077350616455,
      "learning_rate": 5.64765959964683e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8443232774734497,
      "num_tokens": 77867844.0,
      "step": 64730
    },
    {
      "entropy": 1.9272115871310234,
      "epoch": 0.2006883355717285,
      "grad_norm": 8.975436210632324,
      "learning_rate": 5.647223395989846e-06,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8435587778687477,
      "num_tokens": 77880414.0,
      "step": 64740
    },
    {
      "entropy": 1.8668215185403825,
      "epoch": 0.2007193346967782,
      "grad_norm": 8.991979598999023,
      "learning_rate": 5.646787293389365e-06,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8380271717905998,
      "num_tokens": 77893230.0,
      "step": 64750
    },
    {
      "entropy": 1.9775980859994888,
      "epoch": 0.2007503338218279,
      "grad_norm": 9.368738174438477,
      "learning_rate": 5.646351291806372e-06,
      "loss": 0.5447,
      "mean_token_accuracy": 0.8423718631267547,
      "num_tokens": 77904698.0,
      "step": 64760
    },
    {
      "entropy": 1.880776160955429,
      "epoch": 0.20078133294687758,
      "grad_norm": 4.232335090637207,
      "learning_rate": 5.645915391201876e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8433832973241806,
      "num_tokens": 77918354.0,
      "step": 64770
    },
    {
      "entropy": 1.9753024756908417,
      "epoch": 0.20081233207192728,
      "grad_norm": 9.014158248901367,
      "learning_rate": 5.645479591536904e-06,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8433095306158066,
      "num_tokens": 77929796.0,
      "step": 64780
    },
    {
      "entropy": 1.9146323308348656,
      "epoch": 0.20084333119697698,
      "grad_norm": 8.354053497314453,
      "learning_rate": 5.645043892772506e-06,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8488031610846519,
      "num_tokens": 77941791.0,
      "step": 64790
    },
    {
      "entropy": 1.9045788452029229,
      "epoch": 0.20087433032202667,
      "grad_norm": 9.207497596740723,
      "learning_rate": 5.64460829486975e-06,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8390837252140045,
      "num_tokens": 77953907.0,
      "step": 64800
    },
    {
      "entropy": 1.9096369460225104,
      "epoch": 0.20090532944707637,
      "grad_norm": 3.9759466648101807,
      "learning_rate": 5.64417279778973e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.834852209687233,
      "num_tokens": 77966476.0,
      "step": 64810
    },
    {
      "entropy": 1.971180261671543,
      "epoch": 0.20093632857212604,
      "grad_norm": 7.077384948730469,
      "learning_rate": 5.643737401493556e-06,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8305558815598488,
      "num_tokens": 77978231.0,
      "step": 64820
    },
    {
      "entropy": 1.910263580083847,
      "epoch": 0.20096732769717573,
      "grad_norm": 5.62261438369751,
      "learning_rate": 5.643302105942363e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8370845153927803,
      "num_tokens": 77991195.0,
      "step": 64830
    },
    {
      "entropy": 1.8971102446317674,
      "epoch": 0.20099832682222543,
      "grad_norm": 14.074230194091797,
      "learning_rate": 5.6428669110973035e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8393953680992127,
      "num_tokens": 78003545.0,
      "step": 64840
    },
    {
      "entropy": 1.8818523928523063,
      "epoch": 0.20102932594727513,
      "grad_norm": 8.744156837463379,
      "learning_rate": 5.642431816919555e-06,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8389109507203102,
      "num_tokens": 78015484.0,
      "step": 64850
    },
    {
      "entropy": 1.8921169385313987,
      "epoch": 0.20106032507232482,
      "grad_norm": 6.810762882232666,
      "learning_rate": 5.6419968233703105e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8551272869110107,
      "num_tokens": 78027522.0,
      "step": 64860
    },
    {
      "entropy": 2.0100679606199265,
      "epoch": 0.20109132419737452,
      "grad_norm": 7.5746378898620605,
      "learning_rate": 5.641561930410791e-06,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8342107847332955,
      "num_tokens": 78038669.0,
      "step": 64870
    },
    {
      "entropy": 1.886759166419506,
      "epoch": 0.20112232332242422,
      "grad_norm": 8.69468879699707,
      "learning_rate": 5.64112713800223e-06,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8427166402339935,
      "num_tokens": 78051321.0,
      "step": 64880
    },
    {
      "entropy": 1.980288290977478,
      "epoch": 0.20115332244747391,
      "grad_norm": 9.738602638244629,
      "learning_rate": 5.64069244610589e-06,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8373025566339493,
      "num_tokens": 78062300.0,
      "step": 64890
    },
    {
      "entropy": 1.9037230789661408,
      "epoch": 0.2011843215725236,
      "grad_norm": 7.894903659820557,
      "learning_rate": 5.640257854683049e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8482183083891869,
      "num_tokens": 78074143.0,
      "step": 64900
    },
    {
      "entropy": 1.965438848733902,
      "epoch": 0.2012153206975733,
      "grad_norm": 10.353565216064453,
      "learning_rate": 5.639823363695008e-06,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8452044859528541,
      "num_tokens": 78084809.0,
      "step": 64910
    },
    {
      "entropy": 1.8660222113132476,
      "epoch": 0.201246319822623,
      "grad_norm": 8.646041870117188,
      "learning_rate": 5.639388973103089e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8471137225627899,
      "num_tokens": 78097733.0,
      "step": 64920
    },
    {
      "entropy": 1.92643845975399,
      "epoch": 0.2012773189476727,
      "grad_norm": 8.62920093536377,
      "learning_rate": 5.638954682868635e-06,
      "loss": 0.5631,
      "mean_token_accuracy": 0.8208620086312294,
      "num_tokens": 78110514.0,
      "step": 64930
    },
    {
      "entropy": 1.8348462983965874,
      "epoch": 0.2013083180727224,
      "grad_norm": 7.881258964538574,
      "learning_rate": 5.638520492953008e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8508792489767074,
      "num_tokens": 78123634.0,
      "step": 64940
    },
    {
      "entropy": 1.9531269282102586,
      "epoch": 0.2013393171977721,
      "grad_norm": 3.806800603866577,
      "learning_rate": 5.638086403317592e-06,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8420859977602959,
      "num_tokens": 78134835.0,
      "step": 64950
    },
    {
      "entropy": 1.9173472926020623,
      "epoch": 0.2013703163228218,
      "grad_norm": 9.862812995910645,
      "learning_rate": 5.637652413923792e-06,
      "loss": 0.5778,
      "mean_token_accuracy": 0.817606084048748,
      "num_tokens": 78146512.0,
      "step": 64960
    },
    {
      "entropy": 1.9328098177909852,
      "epoch": 0.2014013154478715,
      "grad_norm": 8.525472640991211,
      "learning_rate": 5.637218524733035e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8353380098938942,
      "num_tokens": 78158404.0,
      "step": 64970
    },
    {
      "entropy": 1.9535952508449554,
      "epoch": 0.20143231457292118,
      "grad_norm": 7.10647439956665,
      "learning_rate": 5.636784735706765e-06,
      "loss": 0.5739,
      "mean_token_accuracy": 0.8305754721164703,
      "num_tokens": 78169749.0,
      "step": 64980
    },
    {
      "entropy": 1.946779978275299,
      "epoch": 0.20146331369797088,
      "grad_norm": 8.665600776672363,
      "learning_rate": 5.636351046806451e-06,
      "loss": 0.5525,
      "mean_token_accuracy": 0.835592320561409,
      "num_tokens": 78180620.0,
      "step": 64990
    },
    {
      "entropy": 1.8874027088284493,
      "epoch": 0.20149431282302058,
      "grad_norm": 10.456498146057129,
      "learning_rate": 5.6359174579935805e-06,
      "loss": 0.5393,
      "mean_token_accuracy": 0.8345045700669289,
      "num_tokens": 78192494.0,
      "step": 65000
    },
    {
      "entropy": 1.9049117237329483,
      "epoch": 0.20152531194807027,
      "grad_norm": 8.67405891418457,
      "learning_rate": 5.635483969229662e-06,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8426391571760178,
      "num_tokens": 78204503.0,
      "step": 65010
    },
    {
      "entropy": 1.937725681066513,
      "epoch": 0.20155631107311997,
      "grad_norm": 8.417729377746582,
      "learning_rate": 5.635050580476227e-06,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8371980518102646,
      "num_tokens": 78215731.0,
      "step": 65020
    },
    {
      "entropy": 1.881471572816372,
      "epoch": 0.20158731019816967,
      "grad_norm": 8.803342819213867,
      "learning_rate": 5.6346172916948215e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8541508734226226,
      "num_tokens": 78228595.0,
      "step": 65030
    },
    {
      "entropy": 1.997860112786293,
      "epoch": 0.20161830932321936,
      "grad_norm": 8.877419471740723,
      "learning_rate": 5.634184102847018e-06,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8311259895563126,
      "num_tokens": 78239815.0,
      "step": 65040
    },
    {
      "entropy": 1.9192116037011147,
      "epoch": 0.20164930844826906,
      "grad_norm": 8.195626258850098,
      "learning_rate": 5.6337510138944094e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8414540275931358,
      "num_tokens": 78250962.0,
      "step": 65050
    },
    {
      "entropy": 1.9783461928367614,
      "epoch": 0.20168030757331876,
      "grad_norm": 9.960369110107422,
      "learning_rate": 5.633318024798608e-06,
      "loss": 0.5766,
      "mean_token_accuracy": 0.8294192060828209,
      "num_tokens": 78261583.0,
      "step": 65060
    },
    {
      "entropy": 1.999306383728981,
      "epoch": 0.20171130669836843,
      "grad_norm": 8.405972480773926,
      "learning_rate": 5.6328851355212445e-06,
      "loss": 0.5814,
      "mean_token_accuracy": 0.8255703672766685,
      "num_tokens": 78272701.0,
      "step": 65070
    },
    {
      "entropy": 1.9220364853739738,
      "epoch": 0.20174230582341812,
      "grad_norm": 8.912109375,
      "learning_rate": 5.632452346023972e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8397694423794746,
      "num_tokens": 78284100.0,
      "step": 65080
    },
    {
      "entropy": 1.8529915317893029,
      "epoch": 0.20177330494846782,
      "grad_norm": 4.4439697265625,
      "learning_rate": 5.6320196562684685e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8452705577015877,
      "num_tokens": 78296972.0,
      "step": 65090
    },
    {
      "entropy": 1.8957914143800736,
      "epoch": 0.20180430407351752,
      "grad_norm": 21.00503921508789,
      "learning_rate": 5.6315870662164244e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8400771632790566,
      "num_tokens": 78309482.0,
      "step": 65100
    },
    {
      "entropy": 1.9614157870411872,
      "epoch": 0.2018353031985672,
      "grad_norm": 9.07087230682373,
      "learning_rate": 5.631154575829556e-06,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8453190490603447,
      "num_tokens": 78321178.0,
      "step": 65110
    },
    {
      "entropy": 1.8932098597288132,
      "epoch": 0.2018663023236169,
      "grad_norm": 2.6463136672973633,
      "learning_rate": 5.6307221850696e-06,
      "loss": 0.5016,
      "mean_token_accuracy": 0.8434045687317848,
      "num_tokens": 78333684.0,
      "step": 65120
    },
    {
      "entropy": 1.8674306973814965,
      "epoch": 0.2018973014486666,
      "grad_norm": 8.476088523864746,
      "learning_rate": 5.630289893898313e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8461329773068428,
      "num_tokens": 78346646.0,
      "step": 65130
    },
    {
      "entropy": 1.9643313199281693,
      "epoch": 0.2019283005737163,
      "grad_norm": 9.779008865356445,
      "learning_rate": 5.629857702277471e-06,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8256931126117706,
      "num_tokens": 78358044.0,
      "step": 65140
    },
    {
      "entropy": 1.9168717786669731,
      "epoch": 0.201959299698766,
      "grad_norm": 8.545971870422363,
      "learning_rate": 5.629425610168872e-06,
      "loss": 0.5267,
      "mean_token_accuracy": 0.8337204769253731,
      "num_tokens": 78370065.0,
      "step": 65150
    },
    {
      "entropy": 1.8033166334033013,
      "epoch": 0.2019902988238157,
      "grad_norm": 4.0642991065979,
      "learning_rate": 5.628993617534335e-06,
      "loss": 0.3731,
      "mean_token_accuracy": 0.8547035023570061,
      "num_tokens": 78384325.0,
      "step": 65160
    },
    {
      "entropy": 1.8826155215501785,
      "epoch": 0.2020212979488654,
      "grad_norm": 4.962791442871094,
      "learning_rate": 5.628561724335695e-06,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8410342499613762,
      "num_tokens": 78396987.0,
      "step": 65170
    },
    {
      "entropy": 1.8694611176848412,
      "epoch": 0.2020522970739151,
      "grad_norm": 3.838369131088257,
      "learning_rate": 5.628129930534814e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.846546882390976,
      "num_tokens": 78410583.0,
      "step": 65180
    },
    {
      "entropy": 1.8311977073550225,
      "epoch": 0.20208329619896478,
      "grad_norm": 3.510143518447876,
      "learning_rate": 5.627698236093573e-06,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8387721896171569,
      "num_tokens": 78423729.0,
      "step": 65190
    },
    {
      "entropy": 1.9377717927098275,
      "epoch": 0.20211429532401448,
      "grad_norm": 8.662220001220703,
      "learning_rate": 5.627266640973867e-06,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8386742398142815,
      "num_tokens": 78435466.0,
      "step": 65200
    },
    {
      "entropy": 1.9007938578724861,
      "epoch": 0.20214529444906418,
      "grad_norm": 4.089953899383545,
      "learning_rate": 5.626835145137622e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8495527386665345,
      "num_tokens": 78447613.0,
      "step": 65210
    },
    {
      "entropy": 1.9075167581439019,
      "epoch": 0.20217629357411387,
      "grad_norm": 8.904956817626953,
      "learning_rate": 5.626403748546773e-06,
      "loss": 0.5229,
      "mean_token_accuracy": 0.8407324820756912,
      "num_tokens": 78459460.0,
      "step": 65220
    },
    {
      "entropy": 1.829827456176281,
      "epoch": 0.20220729269916357,
      "grad_norm": 8.076254844665527,
      "learning_rate": 5.625972451163285e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8538295239210129,
      "num_tokens": 78472309.0,
      "step": 65230
    },
    {
      "entropy": 1.964961776137352,
      "epoch": 0.20223829182421327,
      "grad_norm": 4.608421325683594,
      "learning_rate": 5.625541252949139e-06,
      "loss": 0.5394,
      "mean_token_accuracy": 0.8294176504015922,
      "num_tokens": 78483546.0,
      "step": 65240
    },
    {
      "entropy": 1.9537744015455245,
      "epoch": 0.20226929094926296,
      "grad_norm": 8.989550590515137,
      "learning_rate": 5.6251101538663364e-06,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8399158343672752,
      "num_tokens": 78493937.0,
      "step": 65250
    },
    {
      "entropy": 1.8486858293414117,
      "epoch": 0.20230029007431266,
      "grad_norm": 8.618539810180664,
      "learning_rate": 5.624679153876901e-06,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8347037017345429,
      "num_tokens": 78506495.0,
      "step": 65260
    },
    {
      "entropy": 1.8448414601385594,
      "epoch": 0.20233128919936236,
      "grad_norm": 1.566992998123169,
      "learning_rate": 5.624248252942874e-06,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8431580245494843,
      "num_tokens": 78519640.0,
      "step": 65270
    },
    {
      "entropy": 1.9085783809423447,
      "epoch": 0.20236228832441205,
      "grad_norm": 8.527741432189941,
      "learning_rate": 5.623817451026321e-06,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8311206966638565,
      "num_tokens": 78530969.0,
      "step": 65280
    },
    {
      "entropy": 1.9010276600718499,
      "epoch": 0.20239328744946175,
      "grad_norm": 8.006790161132812,
      "learning_rate": 5.623386748089322e-06,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8460407719016075,
      "num_tokens": 78542383.0,
      "step": 65290
    },
    {
      "entropy": 1.9630857422947883,
      "epoch": 0.20242428657451145,
      "grad_norm": 10.483144760131836,
      "learning_rate": 5.622956144093983e-06,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8222698003053666,
      "num_tokens": 78553514.0,
      "step": 65300
    },
    {
      "entropy": 1.9169240906834601,
      "epoch": 0.20245528569956112,
      "grad_norm": 7.206988334655762,
      "learning_rate": 5.622525639002427e-06,
      "loss": 0.5551,
      "mean_token_accuracy": 0.8317204505205155,
      "num_tokens": 78565113.0,
      "step": 65310
    },
    {
      "entropy": 2.0341182202100754,
      "epoch": 0.2024862848246108,
      "grad_norm": 9.495631217956543,
      "learning_rate": 5.6220952327768e-06,
      "loss": 0.5879,
      "mean_token_accuracy": 0.8261968463659286,
      "num_tokens": 78576309.0,
      "step": 65320
    },
    {
      "entropy": 1.8233429193496704,
      "epoch": 0.2025172839496605,
      "grad_norm": 7.392541885375977,
      "learning_rate": 5.6216649253792645e-06,
      "loss": 0.485,
      "mean_token_accuracy": 0.8485372483730316,
      "num_tokens": 78589699.0,
      "step": 65330
    },
    {
      "entropy": 1.9578388512134552,
      "epoch": 0.2025482830747102,
      "grad_norm": 9.04162883758545,
      "learning_rate": 5.6212347167720085e-06,
      "loss": 0.575,
      "mean_token_accuracy": 0.8360649108886719,
      "num_tokens": 78601015.0,
      "step": 65340
    },
    {
      "entropy": 1.9982189297676087,
      "epoch": 0.2025792821997599,
      "grad_norm": 8.363797187805176,
      "learning_rate": 5.620804606917233e-06,
      "loss": 0.5966,
      "mean_token_accuracy": 0.8217172041535378,
      "num_tokens": 78611845.0,
      "step": 65350
    },
    {
      "entropy": 1.954882425069809,
      "epoch": 0.2026102813248096,
      "grad_norm": 3.7723734378814697,
      "learning_rate": 5.620374595777169e-06,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8342722833156586,
      "num_tokens": 78622857.0,
      "step": 65360
    },
    {
      "entropy": 1.9132612109184266,
      "epoch": 0.2026412804498593,
      "grad_norm": 8.916566848754883,
      "learning_rate": 5.619944683314056e-06,
      "loss": 0.545,
      "mean_token_accuracy": 0.8364124834537506,
      "num_tokens": 78635064.0,
      "step": 65370
    },
    {
      "entropy": 1.9422496438026429,
      "epoch": 0.202672279574909,
      "grad_norm": 9.716279983520508,
      "learning_rate": 5.619514869490165e-06,
      "loss": 0.5646,
      "mean_token_accuracy": 0.8320507362484932,
      "num_tokens": 78646048.0,
      "step": 65380
    },
    {
      "entropy": 1.8317863315343856,
      "epoch": 0.2027032786999587,
      "grad_norm": 9.566564559936523,
      "learning_rate": 5.619085154267778e-06,
      "loss": 0.4479,
      "mean_token_accuracy": 0.8540365427732468,
      "num_tokens": 78658841.0,
      "step": 65390
    },
    {
      "entropy": 1.8700613364577294,
      "epoch": 0.20273427782500839,
      "grad_norm": 8.530824661254883,
      "learning_rate": 5.618655537609205e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8470278188586235,
      "num_tokens": 78671464.0,
      "step": 65400
    },
    {
      "entropy": 1.9383513778448105,
      "epoch": 0.20276527695005808,
      "grad_norm": 11.235833168029785,
      "learning_rate": 5.61822601947677e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8409254685044288,
      "num_tokens": 78683244.0,
      "step": 65410
    },
    {
      "entropy": 1.8878853350877762,
      "epoch": 0.20279627607510778,
      "grad_norm": 4.765620231628418,
      "learning_rate": 5.617796599832821e-06,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8444775015115737,
      "num_tokens": 78695252.0,
      "step": 65420
    },
    {
      "entropy": 1.9394584164023398,
      "epoch": 0.20282727520015748,
      "grad_norm": 9.20919132232666,
      "learning_rate": 5.617367278639724e-06,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8318523272871972,
      "num_tokens": 78707014.0,
      "step": 65430
    },
    {
      "entropy": 1.856240051984787,
      "epoch": 0.20285827432520717,
      "grad_norm": 8.130507469177246,
      "learning_rate": 5.6169380558598655e-06,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8484870880842209,
      "num_tokens": 78719124.0,
      "step": 65440
    },
    {
      "entropy": 1.9641065716743469,
      "epoch": 0.20288927345025687,
      "grad_norm": 11.00446891784668,
      "learning_rate": 5.616508931455653e-06,
      "loss": 0.6064,
      "mean_token_accuracy": 0.8264747887849808,
      "num_tokens": 78730740.0,
      "step": 65450
    },
    {
      "entropy": 1.906115210056305,
      "epoch": 0.20292027257530657,
      "grad_norm": 9.167335510253906,
      "learning_rate": 5.616079905389513e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8373008102178574,
      "num_tokens": 78742187.0,
      "step": 65460
    },
    {
      "entropy": 1.8616785183548927,
      "epoch": 0.20295127170035626,
      "grad_norm": 4.638375282287598,
      "learning_rate": 5.6156509776238955e-06,
      "loss": 0.4359,
      "mean_token_accuracy": 0.851144377887249,
      "num_tokens": 78754915.0,
      "step": 65470
    },
    {
      "entropy": 1.9143847838044166,
      "epoch": 0.20298227082540596,
      "grad_norm": 9.230887413024902,
      "learning_rate": 5.615222148121263e-06,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8377783179283143,
      "num_tokens": 78767603.0,
      "step": 65480
    },
    {
      "entropy": 1.9339503601193428,
      "epoch": 0.20301326995045565,
      "grad_norm": 7.70396614074707,
      "learning_rate": 5.614793416844106e-06,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8370099574327469,
      "num_tokens": 78779568.0,
      "step": 65490
    },
    {
      "entropy": 1.9379954680800437,
      "epoch": 0.20304426907550535,
      "grad_norm": 9.033950805664062,
      "learning_rate": 5.614364783754932e-06,
      "loss": 0.5227,
      "mean_token_accuracy": 0.8270322412252427,
      "num_tokens": 78791706.0,
      "step": 65500
    },
    {
      "entropy": 1.8664806455373764,
      "epoch": 0.20307526820055505,
      "grad_norm": 2.266970634460449,
      "learning_rate": 5.613936248816266e-06,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8406695753335953,
      "num_tokens": 78804244.0,
      "step": 65510
    },
    {
      "entropy": 1.8240263767540454,
      "epoch": 0.20310626732560474,
      "grad_norm": 9.05511474609375,
      "learning_rate": 5.613507811990659e-06,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8478545337915421,
      "num_tokens": 78817721.0,
      "step": 65520
    },
    {
      "entropy": 1.8720962792634963,
      "epoch": 0.20313726645065444,
      "grad_norm": 4.106721878051758,
      "learning_rate": 5.613079473240674e-06,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8406256809830666,
      "num_tokens": 78829445.0,
      "step": 65530
    },
    {
      "entropy": 1.9249654412269592,
      "epoch": 0.20316826557570414,
      "grad_norm": 7.825181007385254,
      "learning_rate": 5.612651232528903e-06,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8406661361455917,
      "num_tokens": 78841287.0,
      "step": 65540
    },
    {
      "entropy": 1.9155500084161758,
      "epoch": 0.20319926470075383,
      "grad_norm": 8.039549827575684,
      "learning_rate": 5.61222308981795e-06,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8323755100369453,
      "num_tokens": 78853504.0,
      "step": 65550
    },
    {
      "entropy": 1.9643807247281075,
      "epoch": 0.2032302638258035,
      "grad_norm": 7.621034622192383,
      "learning_rate": 5.611795045070444e-06,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8504145249724389,
      "num_tokens": 78864902.0,
      "step": 65560
    },
    {
      "entropy": 1.8879443630576134,
      "epoch": 0.2032612629508532,
      "grad_norm": 10.218156814575195,
      "learning_rate": 5.611367098249031e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8361433282494545,
      "num_tokens": 78877179.0,
      "step": 65570
    },
    {
      "entropy": 1.863333511352539,
      "epoch": 0.2032922620759029,
      "grad_norm": 8.165308952331543,
      "learning_rate": 5.61093924931638e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.836507086455822,
      "num_tokens": 78889655.0,
      "step": 65580
    },
    {
      "entropy": 1.8848236933350564,
      "epoch": 0.2033232612009526,
      "grad_norm": 4.619635581970215,
      "learning_rate": 5.610511498235176e-06,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8372624054551124,
      "num_tokens": 78901982.0,
      "step": 65590
    },
    {
      "entropy": 1.9587459236383438,
      "epoch": 0.2033542603260023,
      "grad_norm": 9.958028793334961,
      "learning_rate": 5.610083844968128e-06,
      "loss": 0.5397,
      "mean_token_accuracy": 0.8413488537073135,
      "num_tokens": 78912795.0,
      "step": 65600
    },
    {
      "entropy": 1.973938637971878,
      "epoch": 0.203385259451052,
      "grad_norm": 8.87924861907959,
      "learning_rate": 5.609656289477961e-06,
      "loss": 0.5612,
      "mean_token_accuracy": 0.828423522412777,
      "num_tokens": 78923643.0,
      "step": 65610
    },
    {
      "entropy": 1.9414395466446877,
      "epoch": 0.20341625857610168,
      "grad_norm": 8.88390064239502,
      "learning_rate": 5.609228831727426e-06,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8232395872473717,
      "num_tokens": 78934550.0,
      "step": 65620
    },
    {
      "entropy": 1.9254730343818665,
      "epoch": 0.20344725770115138,
      "grad_norm": 9.035659790039062,
      "learning_rate": 5.6088014716792835e-06,
      "loss": 0.5016,
      "mean_token_accuracy": 0.8413706630468368,
      "num_tokens": 78946328.0,
      "step": 65630
    },
    {
      "entropy": 1.8993248373270035,
      "epoch": 0.20347825682620108,
      "grad_norm": 8.599936485290527,
      "learning_rate": 5.6083742092963255e-06,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8389776006340981,
      "num_tokens": 78958262.0,
      "step": 65640
    },
    {
      "entropy": 1.8646668374538422,
      "epoch": 0.20350925595125077,
      "grad_norm": 9.56519603729248,
      "learning_rate": 5.607947044541355e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8369742676615715,
      "num_tokens": 78971613.0,
      "step": 65650
    },
    {
      "entropy": 1.9807878568768502,
      "epoch": 0.20354025507630047,
      "grad_norm": 7.997987747192383,
      "learning_rate": 5.607519977377199e-06,
      "loss": 0.535,
      "mean_token_accuracy": 0.8315387591719627,
      "num_tokens": 78983244.0,
      "step": 65660
    },
    {
      "entropy": 1.8589694291353225,
      "epoch": 0.20357125420135017,
      "grad_norm": 8.80921459197998,
      "learning_rate": 5.607093007766705e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8467966377735138,
      "num_tokens": 78997102.0,
      "step": 65670
    },
    {
      "entropy": 1.8324778914451598,
      "epoch": 0.20360225332639986,
      "grad_norm": 4.456134796142578,
      "learning_rate": 5.606666135672738e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8528119862079621,
      "num_tokens": 79011102.0,
      "step": 65680
    },
    {
      "entropy": 1.94354690015316,
      "epoch": 0.20363325245144956,
      "grad_norm": 8.482820510864258,
      "learning_rate": 5.6062393610581824e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8464238449931145,
      "num_tokens": 79022653.0,
      "step": 65690
    },
    {
      "entropy": 1.8365743920207023,
      "epoch": 0.20366425157649926,
      "grad_norm": 3.94012188911438,
      "learning_rate": 5.605812683885945e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8460432440042496,
      "num_tokens": 79035953.0,
      "step": 65700
    },
    {
      "entropy": 1.9237422615289688,
      "epoch": 0.20369525070154895,
      "grad_norm": 8.296760559082031,
      "learning_rate": 5.6053861041189515e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8385673075914383,
      "num_tokens": 79049233.0,
      "step": 65710
    },
    {
      "entropy": 1.9227632522583007,
      "epoch": 0.20372624982659865,
      "grad_norm": 4.054368019104004,
      "learning_rate": 5.604959621720145e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8334377571940422,
      "num_tokens": 79061362.0,
      "step": 65720
    },
    {
      "entropy": 1.9160942152142524,
      "epoch": 0.20375724895164835,
      "grad_norm": 3.290419340133667,
      "learning_rate": 5.604533236652492e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.840754111111164,
      "num_tokens": 79073746.0,
      "step": 65730
    },
    {
      "entropy": 1.9141247898340226,
      "epoch": 0.20378824807669804,
      "grad_norm": 8.813379287719727,
      "learning_rate": 5.604106948878974e-06,
      "loss": 0.4842,
      "mean_token_accuracy": 0.846279302239418,
      "num_tokens": 79086069.0,
      "step": 65740
    },
    {
      "entropy": 1.8690850347280503,
      "epoch": 0.20381924720174774,
      "grad_norm": 3.4831085205078125,
      "learning_rate": 5.603680758362599e-06,
      "loss": 0.4491,
      "mean_token_accuracy": 0.844321160018444,
      "num_tokens": 79098143.0,
      "step": 65750
    },
    {
      "entropy": 1.9780056357383728,
      "epoch": 0.20385024632679744,
      "grad_norm": 8.82219409942627,
      "learning_rate": 5.603254665066387e-06,
      "loss": 0.5811,
      "mean_token_accuracy": 0.8181054502725601,
      "num_tokens": 79109063.0,
      "step": 65760
    },
    {
      "entropy": 1.9268645226955414,
      "epoch": 0.20388124545184713,
      "grad_norm": 7.107331275939941,
      "learning_rate": 5.602828668953384e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8484486505389214,
      "num_tokens": 79120352.0,
      "step": 65770
    },
    {
      "entropy": 1.9007486268877982,
      "epoch": 0.20391224457689683,
      "grad_norm": 8.633251190185547,
      "learning_rate": 5.602402769986652e-06,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8409877792000771,
      "num_tokens": 79131530.0,
      "step": 65780
    },
    {
      "entropy": 1.873155763745308,
      "epoch": 0.20394324370194652,
      "grad_norm": 4.75627326965332,
      "learning_rate": 5.601976968129274e-06,
      "loss": 0.4344,
      "mean_token_accuracy": 0.8587671235203743,
      "num_tokens": 79143461.0,
      "step": 65790
    },
    {
      "entropy": 1.895801869034767,
      "epoch": 0.20397424282699622,
      "grad_norm": 6.479430675506592,
      "learning_rate": 5.6015512633443526e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8410785019397735,
      "num_tokens": 79155115.0,
      "step": 65800
    },
    {
      "entropy": 1.8817868903279305,
      "epoch": 0.2040052419520459,
      "grad_norm": 7.321788787841797,
      "learning_rate": 5.60112565559501e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8473842695355416,
      "num_tokens": 79168062.0,
      "step": 65810
    },
    {
      "entropy": 1.9196514919400216,
      "epoch": 0.2040362410770956,
      "grad_norm": 4.99709415435791,
      "learning_rate": 5.600700144844387e-06,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8285318657755851,
      "num_tokens": 79180238.0,
      "step": 65820
    },
    {
      "entropy": 1.880447769165039,
      "epoch": 0.20406724020214528,
      "grad_norm": 10.029952049255371,
      "learning_rate": 5.600274731055645e-06,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8361893489956855,
      "num_tokens": 79193071.0,
      "step": 65830
    },
    {
      "entropy": 1.9641305297613143,
      "epoch": 0.20409823932719498,
      "grad_norm": 8.848322868347168,
      "learning_rate": 5.599849414191965e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8405619546771049,
      "num_tokens": 79204235.0,
      "step": 65840
    },
    {
      "entropy": 1.896867436170578,
      "epoch": 0.20412923845224468,
      "grad_norm": 9.602326393127441,
      "learning_rate": 5.599424194216547e-06,
      "loss": 0.492,
      "mean_token_accuracy": 0.8406928956508637,
      "num_tokens": 79216424.0,
      "step": 65850
    },
    {
      "entropy": 1.9198857069015502,
      "epoch": 0.20416023757729437,
      "grad_norm": 10.032066345214844,
      "learning_rate": 5.598999071092613e-06,
      "loss": 0.4794,
      "mean_token_accuracy": 0.8465869203209877,
      "num_tokens": 79227957.0,
      "step": 65860
    },
    {
      "entropy": 1.8796373263001442,
      "epoch": 0.20419123670234407,
      "grad_norm": 4.095681190490723,
      "learning_rate": 5.598574044783399e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8457603216171264,
      "num_tokens": 79241020.0,
      "step": 65870
    },
    {
      "entropy": 1.890330995619297,
      "epoch": 0.20422223582739377,
      "grad_norm": 8.238532066345215,
      "learning_rate": 5.598149115252166e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8491918399930001,
      "num_tokens": 79254022.0,
      "step": 65880
    },
    {
      "entropy": 1.960501140356064,
      "epoch": 0.20425323495244346,
      "grad_norm": 9.54404354095459,
      "learning_rate": 5.597724282462193e-06,
      "loss": 0.5882,
      "mean_token_accuracy": 0.8301891297101974,
      "num_tokens": 79265375.0,
      "step": 65890
    },
    {
      "entropy": 1.8832382291555405,
      "epoch": 0.20428423407749316,
      "grad_norm": 10.813145637512207,
      "learning_rate": 5.597299546376778e-06,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8315230071544647,
      "num_tokens": 79277983.0,
      "step": 65900
    },
    {
      "entropy": 1.9647821724414825,
      "epoch": 0.20431523320254286,
      "grad_norm": 8.538866996765137,
      "learning_rate": 5.596874906959238e-06,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8316359147429466,
      "num_tokens": 79289116.0,
      "step": 65910
    },
    {
      "entropy": 1.8251458272337913,
      "epoch": 0.20434623232759255,
      "grad_norm": 8.377923965454102,
      "learning_rate": 5.596450364172909e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.86036896109581,
      "num_tokens": 79301925.0,
      "step": 65920
    },
    {
      "entropy": 1.9117643371224404,
      "epoch": 0.20437723145264225,
      "grad_norm": 8.313860893249512,
      "learning_rate": 5.596025917981147e-06,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8350561752915382,
      "num_tokens": 79313308.0,
      "step": 65930
    },
    {
      "entropy": 1.9248536437749864,
      "epoch": 0.20440823057769195,
      "grad_norm": 7.247808933258057,
      "learning_rate": 5.595601568347332e-06,
      "loss": 0.5714,
      "mean_token_accuracy": 0.8272919088602066,
      "num_tokens": 79325093.0,
      "step": 65940
    },
    {
      "entropy": 1.9249488562345505,
      "epoch": 0.20443922970274164,
      "grad_norm": 4.61583137512207,
      "learning_rate": 5.5951773152348545e-06,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8503524556756019,
      "num_tokens": 79336001.0,
      "step": 65950
    },
    {
      "entropy": 1.8732093065977096,
      "epoch": 0.20447022882779134,
      "grad_norm": 8.46263313293457,
      "learning_rate": 5.594753158607133e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8438627734780312,
      "num_tokens": 79347710.0,
      "step": 65960
    },
    {
      "entropy": 1.8675259336829186,
      "epoch": 0.20450122795284104,
      "grad_norm": 9.65857219696045,
      "learning_rate": 5.594329098427599e-06,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8433469668030739,
      "num_tokens": 79359872.0,
      "step": 65970
    },
    {
      "entropy": 1.8642318069934845,
      "epoch": 0.20453222707789073,
      "grad_norm": 3.6823036670684814,
      "learning_rate": 5.5939051346597075e-06,
      "loss": 0.4386,
      "mean_token_accuracy": 0.853775417804718,
      "num_tokens": 79372088.0,
      "step": 65980
    },
    {
      "entropy": 1.7361238494515419,
      "epoch": 0.20456322620294043,
      "grad_norm": 3.253519296646118,
      "learning_rate": 5.59348126726693e-06,
      "loss": 0.4458,
      "mean_token_accuracy": 0.8488474145531655,
      "num_tokens": 79386133.0,
      "step": 65990
    },
    {
      "entropy": 1.9137817591428756,
      "epoch": 0.20459422532799013,
      "grad_norm": 9.194711685180664,
      "learning_rate": 5.593057496212762e-06,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8341950923204422,
      "num_tokens": 79397436.0,
      "step": 66000
    },
    {
      "entropy": 1.8905989840626716,
      "epoch": 0.20462522445303982,
      "grad_norm": 9.221059799194336,
      "learning_rate": 5.592633821460712e-06,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8302080318331718,
      "num_tokens": 79409957.0,
      "step": 66010
    },
    {
      "entropy": 1.9771162524819375,
      "epoch": 0.20465622357808952,
      "grad_norm": 8.793451309204102,
      "learning_rate": 5.592210242974312e-06,
      "loss": 0.5649,
      "mean_token_accuracy": 0.8244690522551537,
      "num_tokens": 79421218.0,
      "step": 66020
    },
    {
      "entropy": 1.9063599698245526,
      "epoch": 0.20468722270313922,
      "grad_norm": 8.209443092346191,
      "learning_rate": 5.5917867607171115e-06,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8356138646602631,
      "num_tokens": 79432683.0,
      "step": 66030
    },
    {
      "entropy": 1.9731371477246284,
      "epoch": 0.2047182218281889,
      "grad_norm": 8.800919532775879,
      "learning_rate": 5.5913633746526845e-06,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8384016111493111,
      "num_tokens": 79443817.0,
      "step": 66040
    },
    {
      "entropy": 1.9695898175239563,
      "epoch": 0.2047492209532386,
      "grad_norm": 5.338846206665039,
      "learning_rate": 5.590940084744614e-06,
      "loss": 0.556,
      "mean_token_accuracy": 0.8361503854393959,
      "num_tokens": 79455128.0,
      "step": 66050
    },
    {
      "entropy": 1.9504301056265831,
      "epoch": 0.20478022007828828,
      "grad_norm": 6.454843997955322,
      "learning_rate": 5.590516890956512e-06,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8300734832882881,
      "num_tokens": 79466545.0,
      "step": 66060
    },
    {
      "entropy": 1.916282233595848,
      "epoch": 0.20481121920333797,
      "grad_norm": 10.762775421142578,
      "learning_rate": 5.590093793252005e-06,
      "loss": 0.5018,
      "mean_token_accuracy": 0.850112085044384,
      "num_tokens": 79477559.0,
      "step": 66070
    },
    {
      "entropy": 1.932596181333065,
      "epoch": 0.20484221832838767,
      "grad_norm": 8.139491081237793,
      "learning_rate": 5.5896707915947404e-06,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8297857508063317,
      "num_tokens": 79489215.0,
      "step": 66080
    },
    {
      "entropy": 1.7636524528264999,
      "epoch": 0.20487321745343737,
      "grad_norm": 3.8502655029296875,
      "learning_rate": 5.5892478859483836e-06,
      "loss": 0.4123,
      "mean_token_accuracy": 0.8482681661844254,
      "num_tokens": 79503087.0,
      "step": 66090
    },
    {
      "entropy": 1.9150663495063782,
      "epoch": 0.20490421657848706,
      "grad_norm": 8.005017280578613,
      "learning_rate": 5.588825076276619e-06,
      "loss": 0.5327,
      "mean_token_accuracy": 0.8414066791534424,
      "num_tokens": 79514977.0,
      "step": 66100
    },
    {
      "entropy": 1.884339025616646,
      "epoch": 0.20493521570353676,
      "grad_norm": 12.069209098815918,
      "learning_rate": 5.5884023625431536e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8346866846084595,
      "num_tokens": 79527912.0,
      "step": 66110
    },
    {
      "entropy": 1.9233988001942635,
      "epoch": 0.20496621482858646,
      "grad_norm": 7.263315200805664,
      "learning_rate": 5.58797974471171e-06,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8440830901265144,
      "num_tokens": 79539486.0,
      "step": 66120
    },
    {
      "entropy": 1.9001603171229362,
      "epoch": 0.20499721395363615,
      "grad_norm": 8.054461479187012,
      "learning_rate": 5.587557222746031e-06,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8448784217238426,
      "num_tokens": 79551307.0,
      "step": 66130
    },
    {
      "entropy": 1.966906487941742,
      "epoch": 0.20502821307868585,
      "grad_norm": 7.843715190887451,
      "learning_rate": 5.587134796609878e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8330809384584427,
      "num_tokens": 79563106.0,
      "step": 66140
    },
    {
      "entropy": 1.9580473832786083,
      "epoch": 0.20505921220373555,
      "grad_norm": 8.565571784973145,
      "learning_rate": 5.586712466267033e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8391212552785874,
      "num_tokens": 79574679.0,
      "step": 66150
    },
    {
      "entropy": 1.9535856321454048,
      "epoch": 0.20509021132878524,
      "grad_norm": 7.142053604125977,
      "learning_rate": 5.586290231681297e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8499238818883896,
      "num_tokens": 79586243.0,
      "step": 66160
    },
    {
      "entropy": 1.832335925102234,
      "epoch": 0.20512121045383494,
      "grad_norm": 4.120788097381592,
      "learning_rate": 5.58586809281649e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8358307898044586,
      "num_tokens": 79599760.0,
      "step": 66170
    },
    {
      "entropy": 2.0018374592065813,
      "epoch": 0.20515220957888464,
      "grad_norm": 9.204635620117188,
      "learning_rate": 5.585446049636449e-06,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8410599693655968,
      "num_tokens": 79610302.0,
      "step": 66180
    },
    {
      "entropy": 1.914768175780773,
      "epoch": 0.20518320870393433,
      "grad_norm": 9.314388275146484,
      "learning_rate": 5.585024102105034e-06,
      "loss": 0.5387,
      "mean_token_accuracy": 0.8398109778761864,
      "num_tokens": 79621486.0,
      "step": 66190
    },
    {
      "entropy": 1.8635256588459015,
      "epoch": 0.20521420782898403,
      "grad_norm": 8.0902738571167,
      "learning_rate": 5.5846022501861204e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.8411575838923454,
      "num_tokens": 79634109.0,
      "step": 66200
    },
    {
      "entropy": 1.9430198609828948,
      "epoch": 0.20524520695403373,
      "grad_norm": 9.754541397094727,
      "learning_rate": 5.584180493843605e-06,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8356720983982087,
      "num_tokens": 79645040.0,
      "step": 66210
    },
    {
      "entropy": 1.9141881585121154,
      "epoch": 0.20527620607908342,
      "grad_norm": 4.26265287399292,
      "learning_rate": 5.583758833041404e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8484968930482865,
      "num_tokens": 79656732.0,
      "step": 66220
    },
    {
      "entropy": 1.9716264665126801,
      "epoch": 0.20530720520413312,
      "grad_norm": 8.999820709228516,
      "learning_rate": 5.583337267743449e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8347686797380447,
      "num_tokens": 79667904.0,
      "step": 66230
    },
    {
      "entropy": 1.9396614864468575,
      "epoch": 0.20533820432918282,
      "grad_norm": 8.992365837097168,
      "learning_rate": 5.582915797913695e-06,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8403194323182106,
      "num_tokens": 79679499.0,
      "step": 66240
    },
    {
      "entropy": 1.9468914091587066,
      "epoch": 0.2053692034542325,
      "grad_norm": 3.764883279800415,
      "learning_rate": 5.582494423516115e-06,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8384948015213013,
      "num_tokens": 79690852.0,
      "step": 66250
    },
    {
      "entropy": 1.9295331597328187,
      "epoch": 0.2054002025792822,
      "grad_norm": 7.999744415283203,
      "learning_rate": 5.582073144514698e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8378969594836235,
      "num_tokens": 79701855.0,
      "step": 66260
    },
    {
      "entropy": 1.9138659819960595,
      "epoch": 0.2054312017043319,
      "grad_norm": 7.380083084106445,
      "learning_rate": 5.5816519608734575e-06,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8312686800956726,
      "num_tokens": 79713471.0,
      "step": 66270
    },
    {
      "entropy": 1.8375635713338851,
      "epoch": 0.2054622008293816,
      "grad_norm": 6.883208751678467,
      "learning_rate": 5.58123087255642e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8480865985155106,
      "num_tokens": 79726767.0,
      "step": 66280
    },
    {
      "entropy": 1.8265572801232337,
      "epoch": 0.2054931999544313,
      "grad_norm": 10.417244911193848,
      "learning_rate": 5.580809879527636e-06,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8491293832659721,
      "num_tokens": 79739638.0,
      "step": 66290
    },
    {
      "entropy": 1.8894486725330353,
      "epoch": 0.20552419907948097,
      "grad_norm": 6.471380710601807,
      "learning_rate": 5.580388981751174e-06,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8371762230992317,
      "num_tokens": 79752223.0,
      "step": 66300
    },
    {
      "entropy": 1.7594513550400734,
      "epoch": 0.20555519820453066,
      "grad_norm": 3.132899522781372,
      "learning_rate": 5.579968179191117e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8502689436078071,
      "num_tokens": 79766341.0,
      "step": 66310
    },
    {
      "entropy": 1.7812666043639183,
      "epoch": 0.20558619732958036,
      "grad_norm": 3.3629016876220703,
      "learning_rate": 5.579547471811571e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8472658976912498,
      "num_tokens": 79779942.0,
      "step": 66320
    },
    {
      "entropy": 1.908031238615513,
      "epoch": 0.20561719645463006,
      "grad_norm": 9.27261734008789,
      "learning_rate": 5.579126859576662e-06,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8606662392616272,
      "num_tokens": 79791530.0,
      "step": 66330
    },
    {
      "entropy": 1.912331785261631,
      "epoch": 0.20564819557967975,
      "grad_norm": 6.716431140899658,
      "learning_rate": 5.578706342450532e-06,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8352584257721901,
      "num_tokens": 79804036.0,
      "step": 66340
    },
    {
      "entropy": 1.886544433236122,
      "epoch": 0.20567919470472945,
      "grad_norm": 7.406538963317871,
      "learning_rate": 5.578285920397344e-06,
      "loss": 0.469,
      "mean_token_accuracy": 0.839650048315525,
      "num_tokens": 79816050.0,
      "step": 66350
    },
    {
      "entropy": 1.9246870696544647,
      "epoch": 0.20571019382977915,
      "grad_norm": 9.786649703979492,
      "learning_rate": 5.577865593381278e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8357265755534172,
      "num_tokens": 79827947.0,
      "step": 66360
    },
    {
      "entropy": 1.9001474544405936,
      "epoch": 0.20574119295482884,
      "grad_norm": 5.080446243286133,
      "learning_rate": 5.577445361366534e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.8487439841032028,
      "num_tokens": 79839935.0,
      "step": 66370
    },
    {
      "entropy": 1.9511905193328858,
      "epoch": 0.20577219207987854,
      "grad_norm": 9.604957580566406,
      "learning_rate": 5.577025224317332e-06,
      "loss": 0.536,
      "mean_token_accuracy": 0.8385162249207496,
      "num_tokens": 79851151.0,
      "step": 66380
    },
    {
      "entropy": 1.9839515179395675,
      "epoch": 0.20580319120492824,
      "grad_norm": 8.998725891113281,
      "learning_rate": 5.576605182197907e-06,
      "loss": 0.5523,
      "mean_token_accuracy": 0.8333553358912468,
      "num_tokens": 79862474.0,
      "step": 66390
    },
    {
      "entropy": 2.031079703569412,
      "epoch": 0.20583419032997793,
      "grad_norm": 9.277356147766113,
      "learning_rate": 5.57618523497252e-06,
      "loss": 0.5982,
      "mean_token_accuracy": 0.822320033609867,
      "num_tokens": 79873184.0,
      "step": 66400
    },
    {
      "entropy": 1.959704938530922,
      "epoch": 0.20586518945502763,
      "grad_norm": 9.140758514404297,
      "learning_rate": 5.575765382605441e-06,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8415651440620422,
      "num_tokens": 79884713.0,
      "step": 66410
    },
    {
      "entropy": 1.9794253259897232,
      "epoch": 0.20589618858007733,
      "grad_norm": 10.414609909057617,
      "learning_rate": 5.575345625060967e-06,
      "loss": 0.5002,
      "mean_token_accuracy": 0.844949309527874,
      "num_tokens": 79895411.0,
      "step": 66420
    },
    {
      "entropy": 1.9424155369400977,
      "epoch": 0.20592718770512702,
      "grad_norm": 7.093960762023926,
      "learning_rate": 5.574925962303411e-06,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8516281858086586,
      "num_tokens": 79906896.0,
      "step": 66430
    },
    {
      "entropy": 1.966086308658123,
      "epoch": 0.20595818683017672,
      "grad_norm": 9.383204460144043,
      "learning_rate": 5.574506394297104e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8342742830514908,
      "num_tokens": 79918533.0,
      "step": 66440
    },
    {
      "entropy": 1.9549981564283372,
      "epoch": 0.20598918595522642,
      "grad_norm": 7.823967456817627,
      "learning_rate": 5.574086921006398e-06,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8452614173293114,
      "num_tokens": 79930153.0,
      "step": 66450
    },
    {
      "entropy": 1.9196519732475281,
      "epoch": 0.2060201850802761,
      "grad_norm": 10.632098197937012,
      "learning_rate": 5.57366754239566e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8417575985193253,
      "num_tokens": 79941908.0,
      "step": 66460
    },
    {
      "entropy": 1.9567938655614854,
      "epoch": 0.2060511842053258,
      "grad_norm": 7.4412150382995605,
      "learning_rate": 5.57324825842928e-06,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8311595633625984,
      "num_tokens": 79953034.0,
      "step": 66470
    },
    {
      "entropy": 1.9112473011016846,
      "epoch": 0.2060821833303755,
      "grad_norm": 10.366971015930176,
      "learning_rate": 5.572829069071665e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8476077273488045,
      "num_tokens": 79964657.0,
      "step": 66480
    },
    {
      "entropy": 1.9210296481847764,
      "epoch": 0.2061131824554252,
      "grad_norm": 9.77088451385498,
      "learning_rate": 5.572409974287238e-06,
      "loss": 0.525,
      "mean_token_accuracy": 0.8393374785780907,
      "num_tokens": 79976836.0,
      "step": 66490
    },
    {
      "entropy": 1.9088515147566796,
      "epoch": 0.2061441815804749,
      "grad_norm": 8.623429298400879,
      "learning_rate": 5.5719909740404465e-06,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8390779867768288,
      "num_tokens": 79988995.0,
      "step": 66500
    },
    {
      "entropy": 1.8673071622848512,
      "epoch": 0.2061751807055246,
      "grad_norm": 4.687022686004639,
      "learning_rate": 5.571572068295751e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.8348147094249725,
      "num_tokens": 80001572.0,
      "step": 66510
    },
    {
      "entropy": 1.9063550010323524,
      "epoch": 0.2062061798305743,
      "grad_norm": 7.440145015716553,
      "learning_rate": 5.571153257017634e-06,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8437089160084724,
      "num_tokens": 80013203.0,
      "step": 66520
    },
    {
      "entropy": 1.8806561693549155,
      "epoch": 0.206237178955624,
      "grad_norm": 3.6028239727020264,
      "learning_rate": 5.570734540170597e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8492167726159096,
      "num_tokens": 80025389.0,
      "step": 66530
    },
    {
      "entropy": 2.0046221882104875,
      "epoch": 0.2062681780806737,
      "grad_norm": 10.231752395629883,
      "learning_rate": 5.570315917719158e-06,
      "loss": 0.5636,
      "mean_token_accuracy": 0.8313092350959778,
      "num_tokens": 80036574.0,
      "step": 66540
    },
    {
      "entropy": 1.9903879404067992,
      "epoch": 0.20629917720572336,
      "grad_norm": 9.108248710632324,
      "learning_rate": 5.569897389627855e-06,
      "loss": 0.6043,
      "mean_token_accuracy": 0.8205847591161728,
      "num_tokens": 80048154.0,
      "step": 66550
    },
    {
      "entropy": 1.9472715437412262,
      "epoch": 0.20633017633077305,
      "grad_norm": 8.83901596069336,
      "learning_rate": 5.569478955861244e-06,
      "loss": 0.5332,
      "mean_token_accuracy": 0.8277617216110229,
      "num_tokens": 80060061.0,
      "step": 66560
    },
    {
      "entropy": 1.9148873046040535,
      "epoch": 0.20636117545582275,
      "grad_norm": 8.391643524169922,
      "learning_rate": 5.5690606163839e-06,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8314318493008613,
      "num_tokens": 80071978.0,
      "step": 66570
    },
    {
      "entropy": 1.9878364622592926,
      "epoch": 0.20639217458087245,
      "grad_norm": 9.095196723937988,
      "learning_rate": 5.568642371160418e-06,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8388659507036209,
      "num_tokens": 80083500.0,
      "step": 66580
    },
    {
      "entropy": 1.941962979733944,
      "epoch": 0.20642317370592214,
      "grad_norm": 10.697854995727539,
      "learning_rate": 5.568224220155408e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.833092825114727,
      "num_tokens": 80095609.0,
      "step": 66590
    },
    {
      "entropy": 1.8618491291999817,
      "epoch": 0.20645417283097184,
      "grad_norm": 3.9840543270111084,
      "learning_rate": 5.567806163333503e-06,
      "loss": 0.4091,
      "mean_token_accuracy": 0.8554950803518295,
      "num_tokens": 80108181.0,
      "step": 66600
    },
    {
      "entropy": 1.8243538379669189,
      "epoch": 0.20648517195602153,
      "grad_norm": 3.919654130935669,
      "learning_rate": 5.5673882006593514e-06,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8437583655118942,
      "num_tokens": 80121901.0,
      "step": 66610
    },
    {
      "entropy": 1.950028759241104,
      "epoch": 0.20651617108107123,
      "grad_norm": 9.593600273132324,
      "learning_rate": 5.566970332097621e-06,
      "loss": 0.5336,
      "mean_token_accuracy": 0.8381760567426682,
      "num_tokens": 80132818.0,
      "step": 66620
    },
    {
      "entropy": 1.9005480587482453,
      "epoch": 0.20654717020612093,
      "grad_norm": 8.083364486694336,
      "learning_rate": 5.5665525576129985e-06,
      "loss": 0.5259,
      "mean_token_accuracy": 0.837388077378273,
      "num_tokens": 80145296.0,
      "step": 66630
    },
    {
      "entropy": 1.9510286539793014,
      "epoch": 0.20657816933117062,
      "grad_norm": 9.924178123474121,
      "learning_rate": 5.566134877170189e-06,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8308894857764244,
      "num_tokens": 80156187.0,
      "step": 66640
    },
    {
      "entropy": 1.9527115762233733,
      "epoch": 0.20660916845622032,
      "grad_norm": 8.096648216247559,
      "learning_rate": 5.565717290733918e-06,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8337409555912018,
      "num_tokens": 80167814.0,
      "step": 66650
    },
    {
      "entropy": 2.028332456946373,
      "epoch": 0.20664016758127002,
      "grad_norm": 8.106879234313965,
      "learning_rate": 5.565299798268925e-06,
      "loss": 0.5553,
      "mean_token_accuracy": 0.8302205935120582,
      "num_tokens": 80178500.0,
      "step": 66660
    },
    {
      "entropy": 2.0136495500802996,
      "epoch": 0.20667116670631971,
      "grad_norm": 10.762166023254395,
      "learning_rate": 5.5648823997399714e-06,
      "loss": 0.5889,
      "mean_token_accuracy": 0.8241096153855324,
      "num_tokens": 80189113.0,
      "step": 66670
    },
    {
      "entropy": 1.8538544356822968,
      "epoch": 0.2067021658313694,
      "grad_norm": 9.980530738830566,
      "learning_rate": 5.564465095111836e-06,
      "loss": 0.4405,
      "mean_token_accuracy": 0.854385307431221,
      "num_tokens": 80201832.0,
      "step": 66680
    },
    {
      "entropy": 1.9106570437550545,
      "epoch": 0.2067331649564191,
      "grad_norm": 8.356368064880371,
      "learning_rate": 5.564047884349318e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8357516199350357,
      "num_tokens": 80213809.0,
      "step": 66690
    },
    {
      "entropy": 1.9118531972169877,
      "epoch": 0.2067641640814688,
      "grad_norm": 7.672004699707031,
      "learning_rate": 5.563630767417233e-06,
      "loss": 0.4976,
      "mean_token_accuracy": 0.8388531014323235,
      "num_tokens": 80226123.0,
      "step": 66700
    },
    {
      "entropy": 1.8637908115983008,
      "epoch": 0.2067951632065185,
      "grad_norm": 4.457185745239258,
      "learning_rate": 5.563213744280416e-06,
      "loss": 0.4235,
      "mean_token_accuracy": 0.849481725692749,
      "num_tokens": 80238328.0,
      "step": 66710
    },
    {
      "entropy": 2.0010482162237166,
      "epoch": 0.2068261623315682,
      "grad_norm": 10.969298362731934,
      "learning_rate": 5.562796814903717e-06,
      "loss": 0.5952,
      "mean_token_accuracy": 0.8272617772221565,
      "num_tokens": 80249249.0,
      "step": 66720
    },
    {
      "entropy": 1.8944606065750123,
      "epoch": 0.2068571614566179,
      "grad_norm": 8.652742385864258,
      "learning_rate": 5.562379979252011e-06,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8307804465293884,
      "num_tokens": 80262484.0,
      "step": 66730
    },
    {
      "entropy": 1.8551345482468604,
      "epoch": 0.2068881605816676,
      "grad_norm": 9.101856231689453,
      "learning_rate": 5.5619632372901865e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8466040581464768,
      "num_tokens": 80275457.0,
      "step": 66740
    },
    {
      "entropy": 1.8982827827334403,
      "epoch": 0.2069191597067173,
      "grad_norm": 7.065288543701172,
      "learning_rate": 5.561546588983153e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.851127202808857,
      "num_tokens": 80287822.0,
      "step": 66750
    },
    {
      "entropy": 1.9204147845506667,
      "epoch": 0.20695015883176698,
      "grad_norm": 3.726187229156494,
      "learning_rate": 5.561130034295834e-06,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8382663875818253,
      "num_tokens": 80299151.0,
      "step": 66760
    },
    {
      "entropy": 1.8413376584649086,
      "epoch": 0.20698115795681668,
      "grad_norm": 3.96382737159729,
      "learning_rate": 5.560713573193179e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8472777932882309,
      "num_tokens": 80311811.0,
      "step": 66770
    },
    {
      "entropy": 1.8229566425085069,
      "epoch": 0.20701215708186638,
      "grad_norm": 9.275156021118164,
      "learning_rate": 5.560297205640148e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8521854087710381,
      "num_tokens": 80324515.0,
      "step": 66780
    },
    {
      "entropy": 1.9010821655392647,
      "epoch": 0.20704315620691607,
      "grad_norm": 3.5451276302337646,
      "learning_rate": 5.559880931601726e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8351982876658439,
      "num_tokens": 80337231.0,
      "step": 66790
    },
    {
      "entropy": 1.9652031093835831,
      "epoch": 0.20707415533196574,
      "grad_norm": 9.789170265197754,
      "learning_rate": 5.559464751042909e-06,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8287955492734909,
      "num_tokens": 80348415.0,
      "step": 66800
    },
    {
      "entropy": 1.9459122210741042,
      "epoch": 0.20710515445701544,
      "grad_norm": 8.680326461791992,
      "learning_rate": 5.559048663928719e-06,
      "loss": 0.5252,
      "mean_token_accuracy": 0.837482500076294,
      "num_tokens": 80360317.0,
      "step": 66810
    },
    {
      "entropy": 1.882257905602455,
      "epoch": 0.20713615358206514,
      "grad_norm": 9.22433090209961,
      "learning_rate": 5.558632670224192e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8451365128159523,
      "num_tokens": 80372302.0,
      "step": 66820
    },
    {
      "entropy": 1.9027293175458908,
      "epoch": 0.20716715270711483,
      "grad_norm": 10.963581085205078,
      "learning_rate": 5.558216769894383e-06,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8382361069321632,
      "num_tokens": 80384407.0,
      "step": 66830
    },
    {
      "entropy": 1.9402805969119072,
      "epoch": 0.20719815183216453,
      "grad_norm": 10.329010009765625,
      "learning_rate": 5.557800962904364e-06,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8304101303219795,
      "num_tokens": 80396463.0,
      "step": 66840
    },
    {
      "entropy": 1.9431497290730477,
      "epoch": 0.20722915095721423,
      "grad_norm": 6.116969108581543,
      "learning_rate": 5.557385249219228e-06,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8325564071536065,
      "num_tokens": 80407622.0,
      "step": 66850
    },
    {
      "entropy": 1.9404701754450797,
      "epoch": 0.20726015008226392,
      "grad_norm": 8.709739685058594,
      "learning_rate": 5.5569696288040865e-06,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8362139865756035,
      "num_tokens": 80418746.0,
      "step": 66860
    },
    {
      "entropy": 1.833868359029293,
      "epoch": 0.20729114920731362,
      "grad_norm": 3.70329213142395,
      "learning_rate": 5.5565541016240665e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8538075044751168,
      "num_tokens": 80431962.0,
      "step": 66870
    },
    {
      "entropy": 1.9629965022206306,
      "epoch": 0.20732214833236332,
      "grad_norm": 7.854671478271484,
      "learning_rate": 5.556138667644313e-06,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8370448827743531,
      "num_tokens": 80443654.0,
      "step": 66880
    },
    {
      "entropy": 1.8818984359502793,
      "epoch": 0.207353147457413,
      "grad_norm": 9.3052396774292,
      "learning_rate": 5.5557233268299925e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8465585172176361,
      "num_tokens": 80456312.0,
      "step": 66890
    },
    {
      "entropy": 1.925614383816719,
      "epoch": 0.2073841465824627,
      "grad_norm": 8.081002235412598,
      "learning_rate": 5.555308079146288e-06,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8237112745642662,
      "num_tokens": 80468324.0,
      "step": 66900
    },
    {
      "entropy": 1.9605086162686347,
      "epoch": 0.2074151457075124,
      "grad_norm": 9.929429054260254,
      "learning_rate": 5.554892924558401e-06,
      "loss": 0.5574,
      "mean_token_accuracy": 0.838320504128933,
      "num_tokens": 80480134.0,
      "step": 66910
    },
    {
      "entropy": 1.9550656154751778,
      "epoch": 0.2074461448325621,
      "grad_norm": 9.695679664611816,
      "learning_rate": 5.554477863031548e-06,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8479965060949326,
      "num_tokens": 80491369.0,
      "step": 66920
    },
    {
      "entropy": 1.7863646060228349,
      "epoch": 0.2074771439576118,
      "grad_norm": 8.903475761413574,
      "learning_rate": 5.55406289453097e-06,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8561421141028405,
      "num_tokens": 80504627.0,
      "step": 66930
    },
    {
      "entropy": 1.9312346950173378,
      "epoch": 0.2075081430826615,
      "grad_norm": 8.315851211547852,
      "learning_rate": 5.553648019021922e-06,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8424394950270653,
      "num_tokens": 80515858.0,
      "step": 66940
    },
    {
      "entropy": 1.9585540309548377,
      "epoch": 0.2075391422077112,
      "grad_norm": 8.037333488464355,
      "learning_rate": 5.553233236469678e-06,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8312124446034431,
      "num_tokens": 80527650.0,
      "step": 66950
    },
    {
      "entropy": 1.885070489346981,
      "epoch": 0.2075701413327609,
      "grad_norm": 3.955148458480835,
      "learning_rate": 5.552818546839529e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8447909742593765,
      "num_tokens": 80539892.0,
      "step": 66960
    },
    {
      "entropy": 1.9014841109514236,
      "epoch": 0.20760114045781058,
      "grad_norm": 8.854087829589844,
      "learning_rate": 5.552403950096787e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8488608747720718,
      "num_tokens": 80551351.0,
      "step": 66970
    },
    {
      "entropy": 1.9420479103922843,
      "epoch": 0.20763213958286028,
      "grad_norm": 8.27338981628418,
      "learning_rate": 5.551989446206778e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8347111612558364,
      "num_tokens": 80563303.0,
      "step": 66980
    },
    {
      "entropy": 1.830499567091465,
      "epoch": 0.20766313870790998,
      "grad_norm": 8.519128799438477,
      "learning_rate": 5.55157503513485e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8499986290931701,
      "num_tokens": 80575761.0,
      "step": 66990
    },
    {
      "entropy": 1.8168928176164627,
      "epoch": 0.20769413783295967,
      "grad_norm": 3.4133358001708984,
      "learning_rate": 5.551160716846368e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8549173071980476,
      "num_tokens": 80588988.0,
      "step": 67000
    },
    {
      "entropy": 1.7608814522624017,
      "epoch": 0.20772513695800937,
      "grad_norm": 9.151144981384277,
      "learning_rate": 5.550746491306713e-06,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8523845434188843,
      "num_tokens": 80602468.0,
      "step": 67010
    },
    {
      "entropy": 1.8375635772943497,
      "epoch": 0.20775613608305907,
      "grad_norm": 9.777922630310059,
      "learning_rate": 5.5503323584812866e-06,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8414626881480217,
      "num_tokens": 80615406.0,
      "step": 67020
    },
    {
      "entropy": 1.7990097239613534,
      "epoch": 0.20778713520810876,
      "grad_norm": 5.260679244995117,
      "learning_rate": 5.549918318335509e-06,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8451185315847397,
      "num_tokens": 80628306.0,
      "step": 67030
    },
    {
      "entropy": 1.8393166303634643,
      "epoch": 0.20781813433315843,
      "grad_norm": 9.089051246643066,
      "learning_rate": 5.549504370834814e-06,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8503473341464997,
      "num_tokens": 80640534.0,
      "step": 67040
    },
    {
      "entropy": 1.9743037328124047,
      "epoch": 0.20784913345820813,
      "grad_norm": 7.8871684074401855,
      "learning_rate": 5.54909051594466e-06,
      "loss": 0.5803,
      "mean_token_accuracy": 0.8254914477467536,
      "num_tokens": 80652287.0,
      "step": 67050
    },
    {
      "entropy": 1.9053780257701873,
      "epoch": 0.20788013258325783,
      "grad_norm": 9.635024070739746,
      "learning_rate": 5.5486767536305175e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8367680624127388,
      "num_tokens": 80664802.0,
      "step": 67060
    },
    {
      "entropy": 1.9582928448915482,
      "epoch": 0.20791113170830752,
      "grad_norm": 9.71713924407959,
      "learning_rate": 5.548263083857879e-06,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8391944229602813,
      "num_tokens": 80676644.0,
      "step": 67070
    },
    {
      "entropy": 1.914687879383564,
      "epoch": 0.20794213083335722,
      "grad_norm": 9.875877380371094,
      "learning_rate": 5.547849506592251e-06,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8417564824223518,
      "num_tokens": 80688640.0,
      "step": 67080
    },
    {
      "entropy": 1.883899413049221,
      "epoch": 0.20797312995840692,
      "grad_norm": 4.134374618530273,
      "learning_rate": 5.547436021799163e-06,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8355391338467598,
      "num_tokens": 80701658.0,
      "step": 67090
    },
    {
      "entropy": 1.892640021443367,
      "epoch": 0.2080041290834566,
      "grad_norm": 7.394375801086426,
      "learning_rate": 5.54702262944416e-06,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8335898667573929,
      "num_tokens": 80713327.0,
      "step": 67100
    },
    {
      "entropy": 1.8861460164189339,
      "epoch": 0.2080351282085063,
      "grad_norm": 7.844824314117432,
      "learning_rate": 5.546609329492804e-06,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8451134487986565,
      "num_tokens": 80724788.0,
      "step": 67110
    },
    {
      "entropy": 1.9601202994585036,
      "epoch": 0.208066127333556,
      "grad_norm": 9.103307723999023,
      "learning_rate": 5.546196121910674e-06,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8401788577437401,
      "num_tokens": 80735373.0,
      "step": 67120
    },
    {
      "entropy": 1.9305698484182359,
      "epoch": 0.2080971264586057,
      "grad_norm": 10.515917778015137,
      "learning_rate": 5.545783006663372e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.8503312930464745,
      "num_tokens": 80745599.0,
      "step": 67130
    },
    {
      "entropy": 1.8946173369884491,
      "epoch": 0.2081281255836554,
      "grad_norm": 8.005970001220703,
      "learning_rate": 5.545369983716514e-06,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8377565070986748,
      "num_tokens": 80757413.0,
      "step": 67140
    },
    {
      "entropy": 1.9489420622587204,
      "epoch": 0.2081591247087051,
      "grad_norm": 9.35682487487793,
      "learning_rate": 5.544957053035733e-06,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8386912405490875,
      "num_tokens": 80767952.0,
      "step": 67150
    },
    {
      "entropy": 1.8574773401021958,
      "epoch": 0.2081901238337548,
      "grad_norm": 4.570268154144287,
      "learning_rate": 5.5445442145866835e-06,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8455320358276367,
      "num_tokens": 80780046.0,
      "step": 67160
    },
    {
      "entropy": 1.871639384329319,
      "epoch": 0.2082211229588045,
      "grad_norm": 7.216861724853516,
      "learning_rate": 5.544131468335036e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.846157830953598,
      "num_tokens": 80792621.0,
      "step": 67170
    },
    {
      "entropy": 1.84230377972126,
      "epoch": 0.20825212208385419,
      "grad_norm": 3.1845858097076416,
      "learning_rate": 5.543718814246477e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8477763459086418,
      "num_tokens": 80805344.0,
      "step": 67180
    },
    {
      "entropy": 1.8911578252911567,
      "epoch": 0.20828312120890388,
      "grad_norm": 10.065200805664062,
      "learning_rate": 5.543306252286714e-06,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8389113351702691,
      "num_tokens": 80817758.0,
      "step": 67190
    },
    {
      "entropy": 1.9382609099149704,
      "epoch": 0.20831412033395358,
      "grad_norm": 9.62885570526123,
      "learning_rate": 5.542893782421471e-06,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8365870863199234,
      "num_tokens": 80828548.0,
      "step": 67200
    },
    {
      "entropy": 1.9892377644777297,
      "epoch": 0.20834511945900328,
      "grad_norm": 7.691858768463135,
      "learning_rate": 5.54248140461649e-06,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8110725492238998,
      "num_tokens": 80840477.0,
      "step": 67210
    },
    {
      "entropy": 1.9411058470606803,
      "epoch": 0.20837611858405297,
      "grad_norm": 12.743127822875977,
      "learning_rate": 5.54206911883753e-06,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8270158648490906,
      "num_tokens": 80851801.0,
      "step": 67220
    },
    {
      "entropy": 1.9056400299072265,
      "epoch": 0.20840711770910267,
      "grad_norm": 8.924535751342773,
      "learning_rate": 5.541656925050371e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8432095557451248,
      "num_tokens": 80864251.0,
      "step": 67230
    },
    {
      "entropy": 1.8577035881578923,
      "epoch": 0.20843811683415236,
      "grad_norm": 6.489585876464844,
      "learning_rate": 5.541244823220805e-06,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8534018874168396,
      "num_tokens": 80877496.0,
      "step": 67240
    },
    {
      "entropy": 1.7855018101632596,
      "epoch": 0.20846911595920206,
      "grad_norm": 2.9368503093719482,
      "learning_rate": 5.540832813314648e-06,
      "loss": 0.3625,
      "mean_token_accuracy": 0.8677582457661629,
      "num_tokens": 80891030.0,
      "step": 67250
    },
    {
      "entropy": 1.9163383916020393,
      "epoch": 0.20850011508425176,
      "grad_norm": 8.91242504119873,
      "learning_rate": 5.54042089529773e-06,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8421859487891197,
      "num_tokens": 80902336.0,
      "step": 67260
    },
    {
      "entropy": 1.9107780367136002,
      "epoch": 0.20853111420930145,
      "grad_norm": 8.577934265136719,
      "learning_rate": 5.5400090691359e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.837343692779541,
      "num_tokens": 80912858.0,
      "step": 67270
    },
    {
      "entropy": 1.868037761747837,
      "epoch": 0.20856211333435115,
      "grad_norm": 9.165842056274414,
      "learning_rate": 5.539597334795024e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8567280262708664,
      "num_tokens": 80924871.0,
      "step": 67280
    },
    {
      "entropy": 1.897133542597294,
      "epoch": 0.20859311245940082,
      "grad_norm": 3.6761436462402344,
      "learning_rate": 5.539185692240987e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8450244545936585,
      "num_tokens": 80936908.0,
      "step": 67290
    },
    {
      "entropy": 1.852649959921837,
      "epoch": 0.20862411158445052,
      "grad_norm": 5.1975483894348145,
      "learning_rate": 5.538774141439691e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8447863146662712,
      "num_tokens": 80949637.0,
      "step": 67300
    },
    {
      "entropy": 1.9235988169908524,
      "epoch": 0.2086551107095002,
      "grad_norm": 9.723345756530762,
      "learning_rate": 5.538362682357055e-06,
      "loss": 0.5537,
      "mean_token_accuracy": 0.8258920639753342,
      "num_tokens": 80960775.0,
      "step": 67310
    },
    {
      "entropy": 1.9941146105527878,
      "epoch": 0.2086861098345499,
      "grad_norm": 7.907135963439941,
      "learning_rate": 5.537951314959018e-06,
      "loss": 0.5285,
      "mean_token_accuracy": 0.8398841202259064,
      "num_tokens": 80971994.0,
      "step": 67320
    },
    {
      "entropy": 1.8793514996767045,
      "epoch": 0.2087171089595996,
      "grad_norm": 8.207256317138672,
      "learning_rate": 5.537540039211534e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.8463095262646675,
      "num_tokens": 80984309.0,
      "step": 67330
    },
    {
      "entropy": 1.798450830578804,
      "epoch": 0.2087481080846493,
      "grad_norm": 2.7568373680114746,
      "learning_rate": 5.537128855080577e-06,
      "loss": 0.4334,
      "mean_token_accuracy": 0.8558945968747139,
      "num_tokens": 80997276.0,
      "step": 67340
    },
    {
      "entropy": 1.8368099942803382,
      "epoch": 0.208779107209699,
      "grad_norm": 8.600449562072754,
      "learning_rate": 5.5367177625321355e-06,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8459367111325264,
      "num_tokens": 81010351.0,
      "step": 67350
    },
    {
      "entropy": 1.7894717290997506,
      "epoch": 0.2088101063347487,
      "grad_norm": 2.465794086456299,
      "learning_rate": 5.5363067615322206e-06,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8549363747239113,
      "num_tokens": 81024645.0,
      "step": 67360
    },
    {
      "entropy": 1.80710818618536,
      "epoch": 0.2088411054597984,
      "grad_norm": 5.430171012878418,
      "learning_rate": 5.535895852046857e-06,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8437926679849624,
      "num_tokens": 81038048.0,
      "step": 67370
    },
    {
      "entropy": 1.9083049342036247,
      "epoch": 0.2088721045848481,
      "grad_norm": 9.347772598266602,
      "learning_rate": 5.535485034042086e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.833581855893135,
      "num_tokens": 81050463.0,
      "step": 67380
    },
    {
      "entropy": 1.9086160019040108,
      "epoch": 0.20890310370989779,
      "grad_norm": 8.73170280456543,
      "learning_rate": 5.535074307483974e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8540397524833679,
      "num_tokens": 81062467.0,
      "step": 67390
    },
    {
      "entropy": 1.8362308278679849,
      "epoch": 0.20893410283494748,
      "grad_norm": 6.898866176605225,
      "learning_rate": 5.534663672338595e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8507218867540359,
      "num_tokens": 81075477.0,
      "step": 67400
    },
    {
      "entropy": 1.964953315258026,
      "epoch": 0.20896510195999718,
      "grad_norm": 10.953664779663086,
      "learning_rate": 5.534253128572048e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.845398873090744,
      "num_tokens": 81086245.0,
      "step": 67410
    },
    {
      "entropy": 1.8964574694633485,
      "epoch": 0.20899610108504688,
      "grad_norm": 10.055764198303223,
      "learning_rate": 5.533842676150446e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8452073886990548,
      "num_tokens": 81098248.0,
      "step": 67420
    },
    {
      "entropy": 1.9074363961815834,
      "epoch": 0.20902710021009657,
      "grad_norm": 8.339896202087402,
      "learning_rate": 5.533432315039921e-06,
      "loss": 0.481,
      "mean_token_accuracy": 0.8502192839980125,
      "num_tokens": 81109121.0,
      "step": 67430
    },
    {
      "entropy": 1.9613432750105857,
      "epoch": 0.20905809933514627,
      "grad_norm": 10.081904411315918,
      "learning_rate": 5.533022045206623e-06,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8488780468702316,
      "num_tokens": 81120269.0,
      "step": 67440
    },
    {
      "entropy": 1.951788181066513,
      "epoch": 0.20908909846019597,
      "grad_norm": 8.335789680480957,
      "learning_rate": 5.532611866616719e-06,
      "loss": 0.5574,
      "mean_token_accuracy": 0.8319823101162911,
      "num_tokens": 81131197.0,
      "step": 67450
    },
    {
      "entropy": 1.8895439878106117,
      "epoch": 0.20912009758524566,
      "grad_norm": 9.064699172973633,
      "learning_rate": 5.53220177923639e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8434992015361786,
      "num_tokens": 81142524.0,
      "step": 67460
    },
    {
      "entropy": 1.8257933855056763,
      "epoch": 0.20915109671029536,
      "grad_norm": 8.583352088928223,
      "learning_rate": 5.531791783031842e-06,
      "loss": 0.4277,
      "mean_token_accuracy": 0.850779265165329,
      "num_tokens": 81154674.0,
      "step": 67470
    },
    {
      "entropy": 1.7744839206337928,
      "epoch": 0.20918209583534506,
      "grad_norm": 7.863668918609619,
      "learning_rate": 5.5313818779692915e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8497899547219276,
      "num_tokens": 81168137.0,
      "step": 67480
    },
    {
      "entropy": 1.9520211279392243,
      "epoch": 0.20921309496039475,
      "grad_norm": 12.411447525024414,
      "learning_rate": 5.5309720640149785e-06,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8483351454138756,
      "num_tokens": 81178936.0,
      "step": 67490
    },
    {
      "entropy": 1.8984515577554704,
      "epoch": 0.20924409408544445,
      "grad_norm": 7.144548416137695,
      "learning_rate": 5.530562341135155e-06,
      "loss": 0.5423,
      "mean_token_accuracy": 0.837235240638256,
      "num_tokens": 81191279.0,
      "step": 67500
    },
    {
      "entropy": 1.8845224693417548,
      "epoch": 0.20927509321049415,
      "grad_norm": 9.875595092773438,
      "learning_rate": 5.5301527092960925e-06,
      "loss": 0.4426,
      "mean_token_accuracy": 0.8476980268955231,
      "num_tokens": 81203433.0,
      "step": 67510
    },
    {
      "entropy": 1.9049016535282135,
      "epoch": 0.20930609233554384,
      "grad_norm": 7.915565490722656,
      "learning_rate": 5.529743168464083e-06,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8503067553043365,
      "num_tokens": 81215534.0,
      "step": 67520
    },
    {
      "entropy": 1.8581149563193322,
      "epoch": 0.20933709146059354,
      "grad_norm": 3.4274792671203613,
      "learning_rate": 5.5293337186054315e-06,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8372601106762886,
      "num_tokens": 81227450.0,
      "step": 67530
    },
    {
      "entropy": 1.8535752549767495,
      "epoch": 0.2093680905856432,
      "grad_norm": 10.018072128295898,
      "learning_rate": 5.528924359686464e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.845684327185154,
      "num_tokens": 81239765.0,
      "step": 67540
    },
    {
      "entropy": 1.8562732204794883,
      "epoch": 0.2093990897106929,
      "grad_norm": 7.851308822631836,
      "learning_rate": 5.528515091673519e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8401390001177788,
      "num_tokens": 81252593.0,
      "step": 67550
    },
    {
      "entropy": 1.9241983875632287,
      "epoch": 0.2094300888357426,
      "grad_norm": 9.050371170043945,
      "learning_rate": 5.5281059145329605e-06,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8380835622549057,
      "num_tokens": 81264640.0,
      "step": 67560
    },
    {
      "entropy": 1.9892011910676957,
      "epoch": 0.2094610879607923,
      "grad_norm": 7.706739902496338,
      "learning_rate": 5.527696828231161e-06,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8230189695954323,
      "num_tokens": 81275373.0,
      "step": 67570
    },
    {
      "entropy": 1.9130545258522034,
      "epoch": 0.209492087085842,
      "grad_norm": 8.775412559509277,
      "learning_rate": 5.527287832734517e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8499484866857528,
      "num_tokens": 81287424.0,
      "step": 67580
    },
    {
      "entropy": 1.9063115805387496,
      "epoch": 0.2095230862108917,
      "grad_norm": 9.373629570007324,
      "learning_rate": 5.526878928009438e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8434836745262146,
      "num_tokens": 81300644.0,
      "step": 67590
    },
    {
      "entropy": 2.0209019780158997,
      "epoch": 0.2095540853359414,
      "grad_norm": 8.37852954864502,
      "learning_rate": 5.526470114022357e-06,
      "loss": 0.5427,
      "mean_token_accuracy": 0.830120287835598,
      "num_tokens": 81311779.0,
      "step": 67600
    },
    {
      "entropy": 1.944344201683998,
      "epoch": 0.20958508446099108,
      "grad_norm": 8.276811599731445,
      "learning_rate": 5.526061390739714e-06,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8304149687290192,
      "num_tokens": 81323352.0,
      "step": 67610
    },
    {
      "entropy": 1.9719750136137009,
      "epoch": 0.20961608358604078,
      "grad_norm": 7.648212909698486,
      "learning_rate": 5.5256527581279785e-06,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8441421076655388,
      "num_tokens": 81335024.0,
      "step": 67620
    },
    {
      "entropy": 1.9229772925376891,
      "epoch": 0.20964708271109048,
      "grad_norm": 4.550539970397949,
      "learning_rate": 5.5252442161536276e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8464221879839897,
      "num_tokens": 81346392.0,
      "step": 67630
    },
    {
      "entropy": 1.9195913657546044,
      "epoch": 0.20967808183614017,
      "grad_norm": 5.1315999031066895,
      "learning_rate": 5.524835764783162e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8396693095564842,
      "num_tokens": 81358537.0,
      "step": 67640
    },
    {
      "entropy": 1.7798305720090866,
      "epoch": 0.20970908096118987,
      "grad_norm": 8.729194641113281,
      "learning_rate": 5.524427403983096e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.8598864421248436,
      "num_tokens": 81372176.0,
      "step": 67650
    },
    {
      "entropy": 1.8792505249381066,
      "epoch": 0.20974008008623957,
      "grad_norm": 2.838498592376709,
      "learning_rate": 5.524019133719963e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8509874641895294,
      "num_tokens": 81384748.0,
      "step": 67660
    },
    {
      "entropy": 1.9109538584947585,
      "epoch": 0.20977107921128926,
      "grad_norm": 9.723373413085938,
      "learning_rate": 5.5236109539603145e-06,
      "loss": 0.502,
      "mean_token_accuracy": 0.8392190888524056,
      "num_tokens": 81396571.0,
      "step": 67670
    },
    {
      "entropy": 1.9938452377915383,
      "epoch": 0.20980207833633896,
      "grad_norm": 9.84363079071045,
      "learning_rate": 5.523202864670717e-06,
      "loss": 0.5627,
      "mean_token_accuracy": 0.8240502581000329,
      "num_tokens": 81408124.0,
      "step": 67680
    },
    {
      "entropy": 1.944431021809578,
      "epoch": 0.20983307746138866,
      "grad_norm": 7.82313871383667,
      "learning_rate": 5.522794865817755e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8497413724660874,
      "num_tokens": 81418940.0,
      "step": 67690
    },
    {
      "entropy": 1.866161908209324,
      "epoch": 0.20986407658643835,
      "grad_norm": 4.241913318634033,
      "learning_rate": 5.52238695736803e-06,
      "loss": 0.5066,
      "mean_token_accuracy": 0.838020247220993,
      "num_tokens": 81431920.0,
      "step": 67700
    },
    {
      "entropy": 1.9558416873216629,
      "epoch": 0.20989507571148805,
      "grad_norm": 6.742011547088623,
      "learning_rate": 5.521979139288163e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8325184598565102,
      "num_tokens": 81443316.0,
      "step": 67710
    },
    {
      "entropy": 1.9829859271645547,
      "epoch": 0.20992607483653775,
      "grad_norm": 7.284268856048584,
      "learning_rate": 5.521571411544792e-06,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8363759860396385,
      "num_tokens": 81454728.0,
      "step": 67720
    },
    {
      "entropy": 1.9214450731873511,
      "epoch": 0.20995707396158744,
      "grad_norm": 8.096468925476074,
      "learning_rate": 5.521163774104567e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.83761525452137,
      "num_tokens": 81467284.0,
      "step": 67730
    },
    {
      "entropy": 1.8891347169876098,
      "epoch": 0.20998807308663714,
      "grad_norm": 8.036181449890137,
      "learning_rate": 5.520756226934162e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8447176307439804,
      "num_tokens": 81480629.0,
      "step": 67740
    },
    {
      "entropy": 1.8999253660440445,
      "epoch": 0.21001907221168684,
      "grad_norm": 10.522799491882324,
      "learning_rate": 5.520348770000264e-06,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8358193069696427,
      "num_tokens": 81492287.0,
      "step": 67750
    },
    {
      "entropy": 1.8602621525526046,
      "epoch": 0.21005007133673653,
      "grad_norm": 3.4755728244781494,
      "learning_rate": 5.519941403269581e-06,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8537157669663429,
      "num_tokens": 81505125.0,
      "step": 67760
    },
    {
      "entropy": 1.9202613174915313,
      "epoch": 0.21008107046178623,
      "grad_norm": 8.444540977478027,
      "learning_rate": 5.519534126708833e-06,
      "loss": 0.4978,
      "mean_token_accuracy": 0.842594002187252,
      "num_tokens": 81517173.0,
      "step": 67770
    },
    {
      "entropy": 1.8747940637171268,
      "epoch": 0.2101120695868359,
      "grad_norm": 2.7482457160949707,
      "learning_rate": 5.519126940284762e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.847461374104023,
      "num_tokens": 81530328.0,
      "step": 67780
    },
    {
      "entropy": 1.904065564274788,
      "epoch": 0.2101430687118856,
      "grad_norm": 4.310403347015381,
      "learning_rate": 5.518719843964123e-06,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8386267200112343,
      "num_tokens": 81542313.0,
      "step": 67790
    },
    {
      "entropy": 1.9362899020314217,
      "epoch": 0.2101740678369353,
      "grad_norm": 8.252490043640137,
      "learning_rate": 5.518312837713692e-06,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8409799665212632,
      "num_tokens": 81554270.0,
      "step": 67800
    },
    {
      "entropy": 1.9585260882973672,
      "epoch": 0.210205066961985,
      "grad_norm": 9.089271545410156,
      "learning_rate": 5.517905921500259e-06,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8473243802785874,
      "num_tokens": 81565971.0,
      "step": 67810
    },
    {
      "entropy": 1.9076432511210442,
      "epoch": 0.21023606608703468,
      "grad_norm": 8.114571571350098,
      "learning_rate": 5.517499095290636e-06,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8429806470870972,
      "num_tokens": 81578195.0,
      "step": 67820
    },
    {
      "entropy": 1.876829606294632,
      "epoch": 0.21026706521208438,
      "grad_norm": 10.027071952819824,
      "learning_rate": 5.5170923590516444e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8376459792256356,
      "num_tokens": 81591671.0,
      "step": 67830
    },
    {
      "entropy": 1.9631253123283385,
      "epoch": 0.21029806433713408,
      "grad_norm": 6.608702659606934,
      "learning_rate": 5.51668571275013e-06,
      "loss": 0.5715,
      "mean_token_accuracy": 0.8352919310331345,
      "num_tokens": 81603027.0,
      "step": 67840
    },
    {
      "entropy": 1.9352620527148248,
      "epoch": 0.21032906346218377,
      "grad_norm": 3.764373302459717,
      "learning_rate": 5.51627915635295e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8549694180488586,
      "num_tokens": 81615161.0,
      "step": 67850
    },
    {
      "entropy": 1.9944947317242623,
      "epoch": 0.21036006258723347,
      "grad_norm": 9.673376083374023,
      "learning_rate": 5.515872689826984e-06,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8339143067598342,
      "num_tokens": 81627143.0,
      "step": 67860
    },
    {
      "entropy": 2.035477635264397,
      "epoch": 0.21039106171228317,
      "grad_norm": 9.949082374572754,
      "learning_rate": 5.515466313139126e-06,
      "loss": 0.5937,
      "mean_token_accuracy": 0.8266064122319221,
      "num_tokens": 81637833.0,
      "step": 67870
    },
    {
      "entropy": 1.9482318013906479,
      "epoch": 0.21042206083733286,
      "grad_norm": 8.834207534790039,
      "learning_rate": 5.5150600262562855e-06,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8446396380662918,
      "num_tokens": 81649773.0,
      "step": 67880
    },
    {
      "entropy": 2.023282551765442,
      "epoch": 0.21045305996238256,
      "grad_norm": 8.027750015258789,
      "learning_rate": 5.514653829145392e-06,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8356869459152222,
      "num_tokens": 81660767.0,
      "step": 67890
    },
    {
      "entropy": 1.8931564077734948,
      "epoch": 0.21048405908743226,
      "grad_norm": 4.419806003570557,
      "learning_rate": 5.5142477217733905e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8490284129977226,
      "num_tokens": 81673309.0,
      "step": 67900
    },
    {
      "entropy": 1.9538974300026895,
      "epoch": 0.21051505821248195,
      "grad_norm": 9.901854515075684,
      "learning_rate": 5.513841704107242e-06,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8569442689418793,
      "num_tokens": 81684556.0,
      "step": 67910
    },
    {
      "entropy": 1.9160713866353034,
      "epoch": 0.21054605733753165,
      "grad_norm": 7.857386112213135,
      "learning_rate": 5.513435776113929e-06,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8500816449522972,
      "num_tokens": 81696905.0,
      "step": 67920
    },
    {
      "entropy": 1.864570914208889,
      "epoch": 0.21057705646258135,
      "grad_norm": 9.086345672607422,
      "learning_rate": 5.513029937760446e-06,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8444380149245262,
      "num_tokens": 81709959.0,
      "step": 67930
    },
    {
      "entropy": 2.0148605525493624,
      "epoch": 0.21060805558763104,
      "grad_norm": 8.19638442993164,
      "learning_rate": 5.512624189013806e-06,
      "loss": 0.549,
      "mean_token_accuracy": 0.8306930497288704,
      "num_tokens": 81721588.0,
      "step": 67940
    },
    {
      "entropy": 1.9208003774285316,
      "epoch": 0.21063905471268074,
      "grad_norm": 8.312312126159668,
      "learning_rate": 5.512218529841038e-06,
      "loss": 0.4823,
      "mean_token_accuracy": 0.8352445542812348,
      "num_tokens": 81733933.0,
      "step": 67950
    },
    {
      "entropy": 1.8529336631298066,
      "epoch": 0.21067005383773044,
      "grad_norm": 3.6399710178375244,
      "learning_rate": 5.511812960209193e-06,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8411158919334412,
      "num_tokens": 81746778.0,
      "step": 67960
    },
    {
      "entropy": 1.9720004379749299,
      "epoch": 0.21070105296278013,
      "grad_norm": 8.460240364074707,
      "learning_rate": 5.511407480085334e-06,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8386137381196022,
      "num_tokens": 81758395.0,
      "step": 67970
    },
    {
      "entropy": 2.007098397612572,
      "epoch": 0.21073205208782983,
      "grad_norm": 9.23631763458252,
      "learning_rate": 5.51100208943654e-06,
      "loss": 0.5553,
      "mean_token_accuracy": 0.835183584690094,
      "num_tokens": 81769530.0,
      "step": 67980
    },
    {
      "entropy": 1.8748371377587318,
      "epoch": 0.21076305121287953,
      "grad_norm": 8.400230407714844,
      "learning_rate": 5.510596788229912e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8293806165456772,
      "num_tokens": 81782455.0,
      "step": 67990
    },
    {
      "entropy": 1.9910854011774064,
      "epoch": 0.21079405033792922,
      "grad_norm": 8.415918350219727,
      "learning_rate": 5.510191576432563e-06,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8300189360976219,
      "num_tokens": 81793212.0,
      "step": 68000
    },
    {
      "entropy": 1.8886552080512047,
      "epoch": 0.21082504946297892,
      "grad_norm": 7.983949661254883,
      "learning_rate": 5.509786454011627e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8474817335605621,
      "num_tokens": 81804680.0,
      "step": 68010
    },
    {
      "entropy": 1.929812017083168,
      "epoch": 0.21085604858802862,
      "grad_norm": 7.43207311630249,
      "learning_rate": 5.509381420934252e-06,
      "loss": 0.5576,
      "mean_token_accuracy": 0.8385087087750435,
      "num_tokens": 81816705.0,
      "step": 68020
    },
    {
      "entropy": 1.9719702288508416,
      "epoch": 0.21088704771307829,
      "grad_norm": 7.93623685836792,
      "learning_rate": 5.5089764771676035e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8436658188700676,
      "num_tokens": 81828323.0,
      "step": 68030
    },
    {
      "entropy": 1.9078006014227866,
      "epoch": 0.21091804683812798,
      "grad_norm": 10.205698013305664,
      "learning_rate": 5.508571622678865e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8465578466653824,
      "num_tokens": 81840696.0,
      "step": 68040
    },
    {
      "entropy": 1.9035088881850242,
      "epoch": 0.21094904596317768,
      "grad_norm": 8.092962265014648,
      "learning_rate": 5.5081668574352364e-06,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8459865510463714,
      "num_tokens": 81853045.0,
      "step": 68050
    },
    {
      "entropy": 2.0051578521728515,
      "epoch": 0.21098004508822737,
      "grad_norm": 7.893548488616943,
      "learning_rate": 5.507762181403934e-06,
      "loss": 0.5796,
      "mean_token_accuracy": 0.8307501718401908,
      "num_tokens": 81863610.0,
      "step": 68060
    },
    {
      "entropy": 1.862001748383045,
      "epoch": 0.21101104421327707,
      "grad_norm": 9.222723960876465,
      "learning_rate": 5.507357594552191e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8515830934047699,
      "num_tokens": 81876722.0,
      "step": 68070
    },
    {
      "entropy": 1.9945030003786086,
      "epoch": 0.21104204333832677,
      "grad_norm": 9.973723411560059,
      "learning_rate": 5.5069530968472575e-06,
      "loss": 0.5548,
      "mean_token_accuracy": 0.830809174478054,
      "num_tokens": 81887391.0,
      "step": 68080
    },
    {
      "entropy": 1.9193598270416259,
      "epoch": 0.21107304246337646,
      "grad_norm": 8.859430313110352,
      "learning_rate": 5.506548688256401e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.851340101659298,
      "num_tokens": 81899366.0,
      "step": 68090
    },
    {
      "entropy": 1.9339387387037277,
      "epoch": 0.21110404158842616,
      "grad_norm": 4.928318500518799,
      "learning_rate": 5.506144368746905e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8325812682509423,
      "num_tokens": 81911145.0,
      "step": 68100
    },
    {
      "entropy": 1.9344223082065581,
      "epoch": 0.21113504071347586,
      "grad_norm": 2.5606257915496826,
      "learning_rate": 5.505740138286071e-06,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8293151870369911,
      "num_tokens": 81923550.0,
      "step": 68110
    },
    {
      "entropy": 1.9258287683129311,
      "epoch": 0.21116603983852555,
      "grad_norm": 7.782974720001221,
      "learning_rate": 5.505335996841215e-06,
      "loss": 0.489,
      "mean_token_accuracy": 0.8488619342446327,
      "num_tokens": 81935832.0,
      "step": 68120
    },
    {
      "entropy": 1.8819763243198395,
      "epoch": 0.21119703896357525,
      "grad_norm": 2.672853708267212,
      "learning_rate": 5.504931944379673e-06,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8511589944362641,
      "num_tokens": 81949301.0,
      "step": 68130
    },
    {
      "entropy": 2.001712107658386,
      "epoch": 0.21122803808862495,
      "grad_norm": 9.238136291503906,
      "learning_rate": 5.504527980868795e-06,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8439046397805214,
      "num_tokens": 81960491.0,
      "step": 68140
    },
    {
      "entropy": 1.9117625072598456,
      "epoch": 0.21125903721367464,
      "grad_norm": 8.495841026306152,
      "learning_rate": 5.504124106275948e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8380966350436211,
      "num_tokens": 81972833.0,
      "step": 68150
    },
    {
      "entropy": 1.9697169050574304,
      "epoch": 0.21129003633872434,
      "grad_norm": 9.759078025817871,
      "learning_rate": 5.5037203205685196e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8402854591608048,
      "num_tokens": 81984232.0,
      "step": 68160
    },
    {
      "entropy": 1.9297288402915,
      "epoch": 0.21132103546377404,
      "grad_norm": 4.200081825256348,
      "learning_rate": 5.503316623713908e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8306440845131874,
      "num_tokens": 81995830.0,
      "step": 68170
    },
    {
      "entropy": 1.8650834277272224,
      "epoch": 0.21135203458882373,
      "grad_norm": 7.745209693908691,
      "learning_rate": 5.502913015679533e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8491313457489014,
      "num_tokens": 82008180.0,
      "step": 68180
    },
    {
      "entropy": 1.9364105448126794,
      "epoch": 0.21138303371387343,
      "grad_norm": 9.66666030883789,
      "learning_rate": 5.502509496432829e-06,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8386658191680908,
      "num_tokens": 82019778.0,
      "step": 68190
    },
    {
      "entropy": 1.9314407289028168,
      "epoch": 0.21141403283892313,
      "grad_norm": 9.512344360351562,
      "learning_rate": 5.502106065941247e-06,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8394078284502029,
      "num_tokens": 82031972.0,
      "step": 68200
    },
    {
      "entropy": 1.9816076889634133,
      "epoch": 0.21144503196397282,
      "grad_norm": 7.756664276123047,
      "learning_rate": 5.501702724172256e-06,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8252545759081841,
      "num_tokens": 82043777.0,
      "step": 68210
    },
    {
      "entropy": 1.9324266403913497,
      "epoch": 0.21147603108902252,
      "grad_norm": 4.227046966552734,
      "learning_rate": 5.501299471093341e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8418770223855973,
      "num_tokens": 82056125.0,
      "step": 68220
    },
    {
      "entropy": 1.8676659151911736,
      "epoch": 0.21150703021407222,
      "grad_norm": 9.944682121276855,
      "learning_rate": 5.500896306672003e-06,
      "loss": 0.448,
      "mean_token_accuracy": 0.8422658532857895,
      "num_tokens": 82068630.0,
      "step": 68230
    },
    {
      "entropy": 1.9347829192876815,
      "epoch": 0.2115380293391219,
      "grad_norm": 8.809101104736328,
      "learning_rate": 5.500493230875762e-06,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8409867346286773,
      "num_tokens": 82080910.0,
      "step": 68240
    },
    {
      "entropy": 1.9162681803107262,
      "epoch": 0.2115690284641716,
      "grad_norm": 8.236538887023926,
      "learning_rate": 5.500090243672151e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8516815707087517,
      "num_tokens": 82092398.0,
      "step": 68250
    },
    {
      "entropy": 1.9335265710949898,
      "epoch": 0.2116000275892213,
      "grad_norm": 8.672457695007324,
      "learning_rate": 5.4996873450287205e-06,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8431479528546333,
      "num_tokens": 82104602.0,
      "step": 68260
    },
    {
      "entropy": 1.8349942237138748,
      "epoch": 0.211631026714271,
      "grad_norm": 7.4427924156188965,
      "learning_rate": 5.4992845349130406e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.843446071445942,
      "num_tokens": 82117625.0,
      "step": 68270
    },
    {
      "entropy": 1.9474259793758393,
      "epoch": 0.21166202583932067,
      "grad_norm": 4.666007041931152,
      "learning_rate": 5.498881813292697e-06,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8499509572982789,
      "num_tokens": 82129574.0,
      "step": 68280
    },
    {
      "entropy": 1.8501393646001816,
      "epoch": 0.21169302496437037,
      "grad_norm": 8.344110488891602,
      "learning_rate": 5.498479180135289e-06,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8498427078127861,
      "num_tokens": 82143041.0,
      "step": 68290
    },
    {
      "entropy": 1.8913217276334762,
      "epoch": 0.21172402408942007,
      "grad_norm": 11.081100463867188,
      "learning_rate": 5.498076635408436e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8393001139163971,
      "num_tokens": 82155572.0,
      "step": 68300
    },
    {
      "entropy": 1.8870702683925629,
      "epoch": 0.21175502321446976,
      "grad_norm": 3.9212942123413086,
      "learning_rate": 5.497674179079771e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8416898652911187,
      "num_tokens": 82167720.0,
      "step": 68310
    },
    {
      "entropy": 1.8655384734272957,
      "epoch": 0.21178602233951946,
      "grad_norm": 9.342215538024902,
      "learning_rate": 5.497271811116948e-06,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8492528066039086,
      "num_tokens": 82180629.0,
      "step": 68320
    },
    {
      "entropy": 1.8181511342525483,
      "epoch": 0.21181702146456916,
      "grad_norm": 7.617997169494629,
      "learning_rate": 5.496869531487634e-06,
      "loss": 0.426,
      "mean_token_accuracy": 0.8530010193586349,
      "num_tokens": 82194518.0,
      "step": 68330
    },
    {
      "entropy": 1.939797979593277,
      "epoch": 0.21184802058961885,
      "grad_norm": 6.760346412658691,
      "learning_rate": 5.496467340159511e-06,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8504227936267853,
      "num_tokens": 82205732.0,
      "step": 68340
    },
    {
      "entropy": 1.9736777380108834,
      "epoch": 0.21187901971466855,
      "grad_norm": 10.192756652832031,
      "learning_rate": 5.496065237100283e-06,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8265336826443672,
      "num_tokens": 82217338.0,
      "step": 68350
    },
    {
      "entropy": 1.9774869233369827,
      "epoch": 0.21191001883971824,
      "grad_norm": 8.361988067626953,
      "learning_rate": 5.495663222277665e-06,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8427802279591561,
      "num_tokens": 82228528.0,
      "step": 68360
    },
    {
      "entropy": 1.941089576482773,
      "epoch": 0.21194101796476794,
      "grad_norm": 6.703628063201904,
      "learning_rate": 5.495261295659393e-06,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8476432636380196,
      "num_tokens": 82239902.0,
      "step": 68370
    },
    {
      "entropy": 1.9076963618397713,
      "epoch": 0.21197201708981764,
      "grad_norm": 7.75342321395874,
      "learning_rate": 5.494859457213216e-06,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8372241660952568,
      "num_tokens": 82251693.0,
      "step": 68380
    },
    {
      "entropy": 1.877877102792263,
      "epoch": 0.21200301621486733,
      "grad_norm": 4.743284225463867,
      "learning_rate": 5.494457706906901e-06,
      "loss": 0.4329,
      "mean_token_accuracy": 0.854452121257782,
      "num_tokens": 82264357.0,
      "step": 68390
    },
    {
      "entropy": 1.927871122956276,
      "epoch": 0.21203401533991703,
      "grad_norm": 8.312472343444824,
      "learning_rate": 5.494056044708233e-06,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8458049356937408,
      "num_tokens": 82275361.0,
      "step": 68400
    },
    {
      "entropy": 1.8819461211562156,
      "epoch": 0.21206501446496673,
      "grad_norm": 8.821391105651855,
      "learning_rate": 5.493654470585011e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8411281630396843,
      "num_tokens": 82287079.0,
      "step": 68410
    },
    {
      "entropy": 1.9309892505407333,
      "epoch": 0.21209601359001642,
      "grad_norm": 7.936349391937256,
      "learning_rate": 5.4932529845050494e-06,
      "loss": 0.5459,
      "mean_token_accuracy": 0.833111310005188,
      "num_tokens": 82298592.0,
      "step": 68420
    },
    {
      "entropy": 1.9115046963095665,
      "epoch": 0.21212701271506612,
      "grad_norm": 7.75383996963501,
      "learning_rate": 5.492851586436185e-06,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8484810963273048,
      "num_tokens": 82310414.0,
      "step": 68430
    },
    {
      "entropy": 1.8966820895671845,
      "epoch": 0.21215801184011582,
      "grad_norm": 9.623336791992188,
      "learning_rate": 5.492450276346264e-06,
      "loss": 0.5298,
      "mean_token_accuracy": 0.824579867720604,
      "num_tokens": 82323527.0,
      "step": 68440
    },
    {
      "entropy": 1.929227152466774,
      "epoch": 0.21218901096516551,
      "grad_norm": 8.437525749206543,
      "learning_rate": 5.4920490542031545e-06,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8451452344655991,
      "num_tokens": 82335454.0,
      "step": 68450
    },
    {
      "entropy": 1.8906142652034759,
      "epoch": 0.2122200100902152,
      "grad_norm": 8.768442153930664,
      "learning_rate": 5.4916479199747375e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8401165828108788,
      "num_tokens": 82347531.0,
      "step": 68460
    },
    {
      "entropy": 1.9358887538313865,
      "epoch": 0.2122510092152649,
      "grad_norm": 8.354040145874023,
      "learning_rate": 5.491246873628911e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8335384339094162,
      "num_tokens": 82358918.0,
      "step": 68470
    },
    {
      "entropy": 1.722612802684307,
      "epoch": 0.2122820083403146,
      "grad_norm": 4.17601203918457,
      "learning_rate": 5.490845915133592e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.852487288415432,
      "num_tokens": 82373282.0,
      "step": 68480
    },
    {
      "entropy": 1.9600424468517303,
      "epoch": 0.2123130074653643,
      "grad_norm": 8.103271484375,
      "learning_rate": 5.49044504445671e-06,
      "loss": 0.5425,
      "mean_token_accuracy": 0.8350730538368225,
      "num_tokens": 82384254.0,
      "step": 68490
    },
    {
      "entropy": 1.9285910725593567,
      "epoch": 0.212344006590414,
      "grad_norm": 7.515426158905029,
      "learning_rate": 5.490044261566214e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8370998069643975,
      "num_tokens": 82395748.0,
      "step": 68500
    },
    {
      "entropy": 1.7850094467401505,
      "epoch": 0.2123750057154637,
      "grad_norm": 2.562181234359741,
      "learning_rate": 5.489643566430068e-06,
      "loss": 0.3974,
      "mean_token_accuracy": 0.8605840176343917,
      "num_tokens": 82408632.0,
      "step": 68510
    },
    {
      "entropy": 1.8428252264857292,
      "epoch": 0.21240600484051336,
      "grad_norm": 10.207568168640137,
      "learning_rate": 5.489242959016253e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8491317898035049,
      "num_tokens": 82421317.0,
      "step": 68520
    },
    {
      "entropy": 1.9188319817185402,
      "epoch": 0.21243700396556306,
      "grad_norm": 8.114725112915039,
      "learning_rate": 5.488842439292764e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8487970292568207,
      "num_tokens": 82432817.0,
      "step": 68530
    },
    {
      "entropy": 1.8276264518499374,
      "epoch": 0.21246800309061276,
      "grad_norm": 3.9016194343566895,
      "learning_rate": 5.4884420072276175e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8536558151245117,
      "num_tokens": 82445275.0,
      "step": 68540
    },
    {
      "entropy": 1.8365425869822503,
      "epoch": 0.21249900221566245,
      "grad_norm": 10.570717811584473,
      "learning_rate": 5.48804166278884e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8519185066223145,
      "num_tokens": 82457785.0,
      "step": 68550
    },
    {
      "entropy": 1.9167245119810103,
      "epoch": 0.21253000134071215,
      "grad_norm": 6.744617938995361,
      "learning_rate": 5.487641405944478e-06,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8354932352900505,
      "num_tokens": 82469125.0,
      "step": 68560
    },
    {
      "entropy": 1.908539716899395,
      "epoch": 0.21256100046576185,
      "grad_norm": 8.550407409667969,
      "learning_rate": 5.487241236662596e-06,
      "loss": 0.456,
      "mean_token_accuracy": 0.8480324164032936,
      "num_tokens": 82481644.0,
      "step": 68570
    },
    {
      "entropy": 1.8927012383937836,
      "epoch": 0.21259199959081154,
      "grad_norm": 8.421663284301758,
      "learning_rate": 5.486841154911271e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8398829996585846,
      "num_tokens": 82493695.0,
      "step": 68580
    },
    {
      "entropy": 1.9059516102075578,
      "epoch": 0.21262299871586124,
      "grad_norm": 4.052833080291748,
      "learning_rate": 5.486441160658598e-06,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8456622689962388,
      "num_tokens": 82505268.0,
      "step": 68590
    },
    {
      "entropy": 1.9685225948691367,
      "epoch": 0.21265399784091094,
      "grad_norm": 12.194074630737305,
      "learning_rate": 5.486041253872687e-06,
      "loss": 0.5623,
      "mean_token_accuracy": 0.8278161734342575,
      "num_tokens": 82516596.0,
      "step": 68600
    },
    {
      "entropy": 1.84447330981493,
      "epoch": 0.21268499696596063,
      "grad_norm": 8.463014602661133,
      "learning_rate": 5.485641434521665e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8529325991868972,
      "num_tokens": 82529400.0,
      "step": 68610
    },
    {
      "entropy": 1.8710281908512116,
      "epoch": 0.21271599609101033,
      "grad_norm": 7.990627765655518,
      "learning_rate": 5.48524170257368e-06,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8378486022353172,
      "num_tokens": 82541818.0,
      "step": 68620
    },
    {
      "entropy": 1.8808648347854615,
      "epoch": 0.21274699521606003,
      "grad_norm": 4.200010776519775,
      "learning_rate": 5.484842057996887e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8450862690806389,
      "num_tokens": 82554516.0,
      "step": 68630
    },
    {
      "entropy": 1.8946150675415994,
      "epoch": 0.21277799434110972,
      "grad_norm": 7.040407657623291,
      "learning_rate": 5.484442500759464e-06,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8425398871302605,
      "num_tokens": 82566421.0,
      "step": 68640
    },
    {
      "entropy": 1.9813665717840194,
      "epoch": 0.21280899346615942,
      "grad_norm": 7.405858516693115,
      "learning_rate": 5.4840430308296035e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8323499038815498,
      "num_tokens": 82578195.0,
      "step": 68650
    },
    {
      "entropy": 1.9478058412671089,
      "epoch": 0.21283999259120912,
      "grad_norm": 4.5492329597473145,
      "learning_rate": 5.483643648175514e-06,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8383038908243179,
      "num_tokens": 82590022.0,
      "step": 68660
    },
    {
      "entropy": 1.838891714811325,
      "epoch": 0.2128709917162588,
      "grad_norm": 7.715846061706543,
      "learning_rate": 5.48324435276542e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8427498295903206,
      "num_tokens": 82603321.0,
      "step": 68670
    },
    {
      "entropy": 1.8843644648790359,
      "epoch": 0.2129019908413085,
      "grad_norm": 9.188892364501953,
      "learning_rate": 5.482845144567561e-06,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8428171291947365,
      "num_tokens": 82615976.0,
      "step": 68680
    },
    {
      "entropy": 1.9740795180201531,
      "epoch": 0.2129329899663582,
      "grad_norm": 4.1340765953063965,
      "learning_rate": 5.482446023550199e-06,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8349632441997528,
      "num_tokens": 82627312.0,
      "step": 68690
    },
    {
      "entropy": 1.9550517603754998,
      "epoch": 0.2129639890914079,
      "grad_norm": 8.036059379577637,
      "learning_rate": 5.482046989681602e-06,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8462382450699806,
      "num_tokens": 82638817.0,
      "step": 68700
    },
    {
      "entropy": 1.8374310091137886,
      "epoch": 0.2129949882164576,
      "grad_norm": 9.189522743225098,
      "learning_rate": 5.481648042930061e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.849276214838028,
      "num_tokens": 82651779.0,
      "step": 68710
    },
    {
      "entropy": 1.9555922120809555,
      "epoch": 0.2130259873415073,
      "grad_norm": 8.850881576538086,
      "learning_rate": 5.481249183263883e-06,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8374005898833274,
      "num_tokens": 82663394.0,
      "step": 68720
    },
    {
      "entropy": 1.9426356315612794,
      "epoch": 0.213056986466557,
      "grad_norm": 9.274955749511719,
      "learning_rate": 5.480850410651389e-06,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8367856085300446,
      "num_tokens": 82674791.0,
      "step": 68730
    },
    {
      "entropy": 1.815965899825096,
      "epoch": 0.2130879855916067,
      "grad_norm": 8.32228946685791,
      "learning_rate": 5.4804517250609165e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8415889486670494,
      "num_tokens": 82688146.0,
      "step": 68740
    },
    {
      "entropy": 1.8485229358077049,
      "epoch": 0.21311898471665638,
      "grad_norm": 4.761613845825195,
      "learning_rate": 5.4800531264608205e-06,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8439338058233261,
      "num_tokens": 82701247.0,
      "step": 68750
    },
    {
      "entropy": 1.9277915149927138,
      "epoch": 0.21314998384170608,
      "grad_norm": 9.579967498779297,
      "learning_rate": 5.47965461481947e-06,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8259765520691872,
      "num_tokens": 82713022.0,
      "step": 68760
    },
    {
      "entropy": 1.950098218023777,
      "epoch": 0.21318098296675575,
      "grad_norm": 10.044625282287598,
      "learning_rate": 5.47925619010525e-06,
      "loss": 0.5154,
      "mean_token_accuracy": 0.8383822947740555,
      "num_tokens": 82723810.0,
      "step": 68770
    },
    {
      "entropy": 1.9772588819265366,
      "epoch": 0.21321198209180545,
      "grad_norm": 3.088735342025757,
      "learning_rate": 5.478857852286567e-06,
      "loss": 0.5824,
      "mean_token_accuracy": 0.8151338204741478,
      "num_tokens": 82735793.0,
      "step": 68780
    },
    {
      "entropy": 1.8972803056240082,
      "epoch": 0.21324298121685514,
      "grad_norm": 7.994344711303711,
      "learning_rate": 5.478459601331835e-06,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8478777214884758,
      "num_tokens": 82747857.0,
      "step": 68790
    },
    {
      "entropy": 1.9425535961985587,
      "epoch": 0.21327398034190484,
      "grad_norm": 9.188566207885742,
      "learning_rate": 5.478061437209491e-06,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8370546743273735,
      "num_tokens": 82759441.0,
      "step": 68800
    },
    {
      "entropy": 1.9374151572585105,
      "epoch": 0.21330497946695454,
      "grad_norm": 9.73183822631836,
      "learning_rate": 5.477663359887986e-06,
      "loss": 0.5835,
      "mean_token_accuracy": 0.8232653871178627,
      "num_tokens": 82771750.0,
      "step": 68810
    },
    {
      "entropy": 1.8426042452454567,
      "epoch": 0.21333597859200423,
      "grad_norm": 8.94770622253418,
      "learning_rate": 5.4772653693357835e-06,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8506119534373283,
      "num_tokens": 82785172.0,
      "step": 68820
    },
    {
      "entropy": 1.9215306863188744,
      "epoch": 0.21336697771705393,
      "grad_norm": 8.768072128295898,
      "learning_rate": 5.476867465521369e-06,
      "loss": 0.485,
      "mean_token_accuracy": 0.8429464474320412,
      "num_tokens": 82796191.0,
      "step": 68830
    },
    {
      "entropy": 1.864244209229946,
      "epoch": 0.21339797684210363,
      "grad_norm": 8.412409782409668,
      "learning_rate": 5.4764696484132394e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8489756211638451,
      "num_tokens": 82808552.0,
      "step": 68840
    },
    {
      "entropy": 1.942490178346634,
      "epoch": 0.21342897596715332,
      "grad_norm": 7.980142116546631,
      "learning_rate": 5.47607191797991e-06,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8287267431616783,
      "num_tokens": 82820097.0,
      "step": 68850
    },
    {
      "entropy": 1.9282173097133637,
      "epoch": 0.21345997509220302,
      "grad_norm": 8.005396842956543,
      "learning_rate": 5.475674274189913e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8375453725457191,
      "num_tokens": 82831508.0,
      "step": 68860
    },
    {
      "entropy": 1.9350904941558837,
      "epoch": 0.21349097421725272,
      "grad_norm": 9.574058532714844,
      "learning_rate": 5.4752767170117924e-06,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8478696927428245,
      "num_tokens": 82842649.0,
      "step": 68870
    },
    {
      "entropy": 1.9676428467035294,
      "epoch": 0.2135219733423024,
      "grad_norm": 8.351263046264648,
      "learning_rate": 5.474879246414112e-06,
      "loss": 0.526,
      "mean_token_accuracy": 0.8345917791128159,
      "num_tokens": 82854006.0,
      "step": 68880
    },
    {
      "entropy": 1.8507450267672538,
      "epoch": 0.2135529724673521,
      "grad_norm": 9.128543853759766,
      "learning_rate": 5.47448186236545e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.843813742697239,
      "num_tokens": 82866178.0,
      "step": 68890
    },
    {
      "entropy": 1.9041711300611497,
      "epoch": 0.2135839715924018,
      "grad_norm": 8.370091438293457,
      "learning_rate": 5.474084564834402e-06,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8432056456804276,
      "num_tokens": 82878638.0,
      "step": 68900
    },
    {
      "entropy": 1.7404463455080985,
      "epoch": 0.2136149707174515,
      "grad_norm": 9.250161170959473,
      "learning_rate": 5.473687353789579e-06,
      "loss": 0.3617,
      "mean_token_accuracy": 0.8636480793356895,
      "num_tokens": 82892733.0,
      "step": 68910
    },
    {
      "entropy": 1.8433457240462303,
      "epoch": 0.2136459698425012,
      "grad_norm": 8.090189933776855,
      "learning_rate": 5.473290229199604e-06,
      "loss": 0.4374,
      "mean_token_accuracy": 0.8539926618337631,
      "num_tokens": 82905596.0,
      "step": 68920
    },
    {
      "entropy": 1.9230886220932006,
      "epoch": 0.2136769689675509,
      "grad_norm": 9.134618759155273,
      "learning_rate": 5.472893191033122e-06,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8437715753912925,
      "num_tokens": 82917687.0,
      "step": 68930
    },
    {
      "entropy": 1.9287979647517204,
      "epoch": 0.2137079680926006,
      "grad_norm": 9.821601867675781,
      "learning_rate": 5.472496239258791e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.839262530207634,
      "num_tokens": 82928862.0,
      "step": 68940
    },
    {
      "entropy": 1.8872993856668472,
      "epoch": 0.2137389672176503,
      "grad_norm": 8.534150123596191,
      "learning_rate": 5.472099373845285e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8436684221029281,
      "num_tokens": 82940484.0,
      "step": 68950
    },
    {
      "entropy": 1.8906041860580445,
      "epoch": 0.21376996634269999,
      "grad_norm": 10.668354988098145,
      "learning_rate": 5.471702594761294e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8390226155519486,
      "num_tokens": 82952870.0,
      "step": 68960
    },
    {
      "entropy": 1.808730572462082,
      "epoch": 0.21380096546774968,
      "grad_norm": 4.311106204986572,
      "learning_rate": 5.471305901975526e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8523236215114594,
      "num_tokens": 82965433.0,
      "step": 68970
    },
    {
      "entropy": 1.9497309267520904,
      "epoch": 0.21383196459279938,
      "grad_norm": 7.436934471130371,
      "learning_rate": 5.470909295456699e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.843020536005497,
      "num_tokens": 82976459.0,
      "step": 68980
    },
    {
      "entropy": 1.9538449883460998,
      "epoch": 0.21386296371784907,
      "grad_norm": 7.01023530960083,
      "learning_rate": 5.470512775173554e-06,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8403704330325127,
      "num_tokens": 82987717.0,
      "step": 68990
    },
    {
      "entropy": 1.879340572655201,
      "epoch": 0.21389396284289877,
      "grad_norm": 7.733237266540527,
      "learning_rate": 5.470116341094843e-06,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8418584942817688,
      "num_tokens": 83000323.0,
      "step": 69000
    },
    {
      "entropy": 1.9619764864444733,
      "epoch": 0.21392496196794847,
      "grad_norm": 9.5713472366333,
      "learning_rate": 5.469719993189336e-06,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8402963668107987,
      "num_tokens": 83011596.0,
      "step": 69010
    },
    {
      "entropy": 1.9356929183006286,
      "epoch": 0.21395596109299814,
      "grad_norm": 7.573556900024414,
      "learning_rate": 5.469323731425817e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.8439344793558121,
      "num_tokens": 83022881.0,
      "step": 69020
    },
    {
      "entropy": 1.8428806871175767,
      "epoch": 0.21398696021804783,
      "grad_norm": 4.41666841506958,
      "learning_rate": 5.46892755577309e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.845615790784359,
      "num_tokens": 83035756.0,
      "step": 69030
    },
    {
      "entropy": 1.8916326105594634,
      "epoch": 0.21401795934309753,
      "grad_norm": 8.813542366027832,
      "learning_rate": 5.46853146619997e-06,
      "loss": 0.485,
      "mean_token_accuracy": 0.8480862900614738,
      "num_tokens": 83046777.0,
      "step": 69040
    },
    {
      "entropy": 1.9366579607129097,
      "epoch": 0.21404895846814723,
      "grad_norm": 10.812383651733398,
      "learning_rate": 5.468135462675289e-06,
      "loss": 0.5598,
      "mean_token_accuracy": 0.8317944586277009,
      "num_tokens": 83058798.0,
      "step": 69050
    },
    {
      "entropy": 1.875098566710949,
      "epoch": 0.21407995759319692,
      "grad_norm": 7.257170677185059,
      "learning_rate": 5.467739545167898e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8402672663331032,
      "num_tokens": 83071399.0,
      "step": 69060
    },
    {
      "entropy": 1.9152754202485085,
      "epoch": 0.21411095671824662,
      "grad_norm": 8.708121299743652,
      "learning_rate": 5.467343713646658e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8351624384522438,
      "num_tokens": 83083904.0,
      "step": 69070
    },
    {
      "entropy": 1.938556207716465,
      "epoch": 0.21414195584329632,
      "grad_norm": 9.21479606628418,
      "learning_rate": 5.466947968080452e-06,
      "loss": 0.493,
      "mean_token_accuracy": 0.8361433446407318,
      "num_tokens": 83095844.0,
      "step": 69080
    },
    {
      "entropy": 1.9366083085536956,
      "epoch": 0.214172954968346,
      "grad_norm": 8.12380313873291,
      "learning_rate": 5.466552308438176e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8430330127477645,
      "num_tokens": 83107044.0,
      "step": 69090
    },
    {
      "entropy": 1.9676083683967591,
      "epoch": 0.2142039540933957,
      "grad_norm": 8.887046813964844,
      "learning_rate": 5.466156734688738e-06,
      "loss": 0.588,
      "mean_token_accuracy": 0.8293441072106361,
      "num_tokens": 83118055.0,
      "step": 69100
    },
    {
      "entropy": 1.8269154354929924,
      "epoch": 0.2142349532184454,
      "grad_norm": 4.919051647186279,
      "learning_rate": 5.4657612468010675e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8531086966395378,
      "num_tokens": 83130970.0,
      "step": 69110
    },
    {
      "entropy": 2.0073640048503876,
      "epoch": 0.2142659523434951,
      "grad_norm": 9.67685317993164,
      "learning_rate": 5.46536584474411e-06,
      "loss": 0.5912,
      "mean_token_accuracy": 0.8317721515893937,
      "num_tokens": 83142153.0,
      "step": 69120
    },
    {
      "entropy": 1.9201458156108857,
      "epoch": 0.2142969514685448,
      "grad_norm": 11.463045120239258,
      "learning_rate": 5.464970528486821e-06,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8264217585325241,
      "num_tokens": 83153550.0,
      "step": 69130
    },
    {
      "entropy": 2.0055788159370422,
      "epoch": 0.2143279505935945,
      "grad_norm": 9.573265075683594,
      "learning_rate": 5.464575297998177e-06,
      "loss": 0.5833,
      "mean_token_accuracy": 0.8264138385653496,
      "num_tokens": 83164548.0,
      "step": 69140
    },
    {
      "entropy": 1.8320914067327976,
      "epoch": 0.2143589497186442,
      "grad_norm": 7.352808952331543,
      "learning_rate": 5.464180153247167e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8488354295492172,
      "num_tokens": 83177984.0,
      "step": 69150
    },
    {
      "entropy": 1.8535144045948981,
      "epoch": 0.2143899488436939,
      "grad_norm": 8.093338012695312,
      "learning_rate": 5.463785094202798e-06,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8474163025617599,
      "num_tokens": 83190325.0,
      "step": 69160
    },
    {
      "entropy": 1.9603209257125855,
      "epoch": 0.21442094796874359,
      "grad_norm": 7.552434921264648,
      "learning_rate": 5.46339012083409e-06,
      "loss": 0.549,
      "mean_token_accuracy": 0.8280288457870484,
      "num_tokens": 83201141.0,
      "step": 69170
    },
    {
      "entropy": 1.9304186806082726,
      "epoch": 0.21445194709379328,
      "grad_norm": 4.334521770477295,
      "learning_rate": 5.462995233110082e-06,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8300809517502785,
      "num_tokens": 83212796.0,
      "step": 69180
    },
    {
      "entropy": 1.9205600872635842,
      "epoch": 0.21448294621884298,
      "grad_norm": 7.847184658050537,
      "learning_rate": 5.462600430999826e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8355577915906907,
      "num_tokens": 83224491.0,
      "step": 69190
    },
    {
      "entropy": 1.9560988396406174,
      "epoch": 0.21451394534389268,
      "grad_norm": 8.206576347351074,
      "learning_rate": 5.462205714472391e-06,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8278788521885871,
      "num_tokens": 83235792.0,
      "step": 69200
    },
    {
      "entropy": 1.8794623762369156,
      "epoch": 0.21454494446894237,
      "grad_norm": 8.750676155090332,
      "learning_rate": 5.4618110834968615e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8406390935182572,
      "num_tokens": 83247555.0,
      "step": 69210
    },
    {
      "entropy": 1.8585446387529374,
      "epoch": 0.21457594359399207,
      "grad_norm": 3.854595422744751,
      "learning_rate": 5.461416538042337e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8608892947435379,
      "num_tokens": 83259237.0,
      "step": 69220
    },
    {
      "entropy": 1.7895995572209358,
      "epoch": 0.21460694271904177,
      "grad_norm": 4.206018447875977,
      "learning_rate": 5.4610220780779335e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.8465156257152557,
      "num_tokens": 83272545.0,
      "step": 69230
    },
    {
      "entropy": 1.7522881373763084,
      "epoch": 0.21463794184409146,
      "grad_norm": 5.796809673309326,
      "learning_rate": 5.46062770357278e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8636647120118142,
      "num_tokens": 83285897.0,
      "step": 69240
    },
    {
      "entropy": 1.8799547150731086,
      "epoch": 0.21466894096914116,
      "grad_norm": 4.128879070281982,
      "learning_rate": 5.460233414496026e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8423994764685631,
      "num_tokens": 83297294.0,
      "step": 69250
    },
    {
      "entropy": 1.9117163017392158,
      "epoch": 0.21469994009419083,
      "grad_norm": 8.076613426208496,
      "learning_rate": 5.459839210816833e-06,
      "loss": 0.5012,
      "mean_token_accuracy": 0.8484107598662376,
      "num_tokens": 83308283.0,
      "step": 69260
    },
    {
      "entropy": 1.8744409173727035,
      "epoch": 0.21473093921924052,
      "grad_norm": 8.670746803283691,
      "learning_rate": 5.459445092504379e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.840450718998909,
      "num_tokens": 83319959.0,
      "step": 69270
    },
    {
      "entropy": 1.83160520195961,
      "epoch": 0.21476193834429022,
      "grad_norm": 9.35455322265625,
      "learning_rate": 5.459051059527855e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8386999174952507,
      "num_tokens": 83332935.0,
      "step": 69280
    },
    {
      "entropy": 1.9142531588673593,
      "epoch": 0.21479293746933992,
      "grad_norm": 9.220640182495117,
      "learning_rate": 5.458657111856474e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8410219177603722,
      "num_tokens": 83343839.0,
      "step": 69290
    },
    {
      "entropy": 1.859879168868065,
      "epoch": 0.21482393659438961,
      "grad_norm": 8.967870712280273,
      "learning_rate": 5.458263249459458e-06,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8425408810377121,
      "num_tokens": 83356149.0,
      "step": 69300
    },
    {
      "entropy": 1.9017312452197075,
      "epoch": 0.2148549357194393,
      "grad_norm": 8.242757797241211,
      "learning_rate": 5.4578694723060475e-06,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8415887534618378,
      "num_tokens": 83367770.0,
      "step": 69310
    },
    {
      "entropy": 1.8691633999347688,
      "epoch": 0.214885934844489,
      "grad_norm": 8.582977294921875,
      "learning_rate": 5.4574757803654994e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.850647197663784,
      "num_tokens": 83378887.0,
      "step": 69320
    },
    {
      "entropy": 1.908175478875637,
      "epoch": 0.2149169339695387,
      "grad_norm": 7.830326080322266,
      "learning_rate": 5.457082173607083e-06,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8378239721059799,
      "num_tokens": 83390350.0,
      "step": 69330
    },
    {
      "entropy": 1.9030898958444595,
      "epoch": 0.2149479330945884,
      "grad_norm": 6.693634510040283,
      "learning_rate": 5.456688652000087e-06,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8313490644097328,
      "num_tokens": 83402282.0,
      "step": 69340
    },
    {
      "entropy": 1.9227802157402039,
      "epoch": 0.2149789322196381,
      "grad_norm": 13.45190143585205,
      "learning_rate": 5.456295215513813e-06,
      "loss": 0.514,
      "mean_token_accuracy": 0.8335293993353844,
      "num_tokens": 83413989.0,
      "step": 69350
    },
    {
      "entropy": 1.957323005795479,
      "epoch": 0.2150099313446878,
      "grad_norm": 12.361275672912598,
      "learning_rate": 5.455901864117576e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.837789298593998,
      "num_tokens": 83424597.0,
      "step": 69360
    },
    {
      "entropy": 1.9297591596841812,
      "epoch": 0.2150409304697375,
      "grad_norm": 12.451163291931152,
      "learning_rate": 5.455508597780713e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.831664165854454,
      "num_tokens": 83436096.0,
      "step": 69370
    },
    {
      "entropy": 1.9417940333485604,
      "epoch": 0.2150719295947872,
      "grad_norm": 8.612032890319824,
      "learning_rate": 5.455115416472572e-06,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8476990014314651,
      "num_tokens": 83447255.0,
      "step": 69380
    },
    {
      "entropy": 1.9677758157253264,
      "epoch": 0.21510292871983688,
      "grad_norm": 7.140786170959473,
      "learning_rate": 5.454722320162514e-06,
      "loss": 0.5591,
      "mean_token_accuracy": 0.8349123015999794,
      "num_tokens": 83458140.0,
      "step": 69390
    },
    {
      "entropy": 1.8533436045050622,
      "epoch": 0.21513392784488658,
      "grad_norm": 8.395837783813477,
      "learning_rate": 5.454329308819922e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.835472346842289,
      "num_tokens": 83470694.0,
      "step": 69400
    },
    {
      "entropy": 1.9218518555164337,
      "epoch": 0.21516492696993628,
      "grad_norm": Infinity,
      "learning_rate": 5.45393638241419e-06,
      "loss": 0.518,
      "mean_token_accuracy": 0.8393091827630996,
      "num_tokens": 83482363.0,
      "step": 69410
    },
    {
      "entropy": 1.9040866911411285,
      "epoch": 0.21519592609498597,
      "grad_norm": 7.9769673347473145,
      "learning_rate": 5.4535435409147265e-06,
      "loss": 0.4728,
      "mean_token_accuracy": 0.842051412165165,
      "num_tokens": 83494631.0,
      "step": 69420
    },
    {
      "entropy": 1.9231672808527946,
      "epoch": 0.21522692522003567,
      "grad_norm": 9.245956420898438,
      "learning_rate": 5.45315078429096e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.847654415667057,
      "num_tokens": 83506941.0,
      "step": 69430
    },
    {
      "entropy": 1.9047624886035919,
      "epoch": 0.21525792434508537,
      "grad_norm": 9.397164344787598,
      "learning_rate": 5.452758112512331e-06,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8385005071759224,
      "num_tokens": 83518932.0,
      "step": 69440
    },
    {
      "entropy": 1.8809319645166398,
      "epoch": 0.21528892347013506,
      "grad_norm": 10.352240562438965,
      "learning_rate": 5.452365525548295e-06,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8362722516059875,
      "num_tokens": 83531350.0,
      "step": 69450
    },
    {
      "entropy": 1.8688334867358207,
      "epoch": 0.21531992259518476,
      "grad_norm": 4.476759433746338,
      "learning_rate": 5.451973023368325e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8403964400291443,
      "num_tokens": 83543584.0,
      "step": 69460
    },
    {
      "entropy": 1.9141950756311417,
      "epoch": 0.21535092172023446,
      "grad_norm": 8.979772567749023,
      "learning_rate": 5.451580605941909e-06,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8465558111667633,
      "num_tokens": 83555457.0,
      "step": 69470
    },
    {
      "entropy": 1.9037456959486008,
      "epoch": 0.21538192084528415,
      "grad_norm": 12.75918197631836,
      "learning_rate": 5.451188273238549e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8453539073467254,
      "num_tokens": 83567575.0,
      "step": 69480
    },
    {
      "entropy": 1.886037102341652,
      "epoch": 0.21541291997033385,
      "grad_norm": 2.314100503921509,
      "learning_rate": 5.450796025227764e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.8452310040593147,
      "num_tokens": 83580210.0,
      "step": 69490
    },
    {
      "entropy": 1.9223955109715463,
      "epoch": 0.21544391909538355,
      "grad_norm": 7.768220901489258,
      "learning_rate": 5.450403861879085e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8553515776991845,
      "num_tokens": 83592037.0,
      "step": 69500
    },
    {
      "entropy": 1.8389130905270576,
      "epoch": 0.21547491822043321,
      "grad_norm": 3.468709945678711,
      "learning_rate": 5.450011783162063e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8565122127532959,
      "num_tokens": 83604963.0,
      "step": 69510
    },
    {
      "entropy": 1.9391876861453057,
      "epoch": 0.2155059173454829,
      "grad_norm": 7.677327632904053,
      "learning_rate": 5.449619789046263e-06,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8431411743164062,
      "num_tokens": 83616136.0,
      "step": 69520
    },
    {
      "entropy": 1.923940536379814,
      "epoch": 0.2155369164705326,
      "grad_norm": 9.04365348815918,
      "learning_rate": 5.449227879501263e-06,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8554372638463974,
      "num_tokens": 83627773.0,
      "step": 69530
    },
    {
      "entropy": 1.8013005346059798,
      "epoch": 0.2155679155955823,
      "grad_norm": 7.584324359893799,
      "learning_rate": 5.448836054496658e-06,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8366676792502403,
      "num_tokens": 83642089.0,
      "step": 69540
    },
    {
      "entropy": 1.9511119529604912,
      "epoch": 0.215598914720632,
      "grad_norm": 8.205597877502441,
      "learning_rate": 5.448444314002058e-06,
      "loss": 0.561,
      "mean_token_accuracy": 0.8263240188360215,
      "num_tokens": 83653309.0,
      "step": 69550
    },
    {
      "entropy": 1.8489666640758515,
      "epoch": 0.2156299138456817,
      "grad_norm": 9.638209342956543,
      "learning_rate": 5.448052657987088e-06,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8453381776809692,
      "num_tokens": 83666503.0,
      "step": 69560
    },
    {
      "entropy": 1.9822421237826346,
      "epoch": 0.2156609129707314,
      "grad_norm": 7.235997676849365,
      "learning_rate": 5.4476610864213905e-06,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8368374884128571,
      "num_tokens": 83677549.0,
      "step": 69570
    },
    {
      "entropy": 1.9172227129340171,
      "epoch": 0.2156919120957811,
      "grad_norm": 9.961241722106934,
      "learning_rate": 5.447269599274621e-06,
      "loss": 0.5732,
      "mean_token_accuracy": 0.8320469737052918,
      "num_tokens": 83689740.0,
      "step": 69580
    },
    {
      "entropy": 1.8688659638166427,
      "epoch": 0.2157229112208308,
      "grad_norm": 4.677713394165039,
      "learning_rate": 5.446878196516448e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8490154132246971,
      "num_tokens": 83701923.0,
      "step": 69590
    },
    {
      "entropy": 1.9590770199894905,
      "epoch": 0.21575391034588048,
      "grad_norm": 4.404328346252441,
      "learning_rate": 5.446486878116561e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8431831479072571,
      "num_tokens": 83713981.0,
      "step": 69600
    },
    {
      "entropy": 1.9376450017094613,
      "epoch": 0.21578490947093018,
      "grad_norm": 9.207627296447754,
      "learning_rate": 5.446095644044661e-06,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8345242127776146,
      "num_tokens": 83725681.0,
      "step": 69610
    },
    {
      "entropy": 1.9425865828990936,
      "epoch": 0.21581590859597988,
      "grad_norm": 9.27120304107666,
      "learning_rate": 5.445704494270465e-06,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8417545929551125,
      "num_tokens": 83737057.0,
      "step": 69620
    },
    {
      "entropy": 1.794182512164116,
      "epoch": 0.21584690772102957,
      "grad_norm": 9.344676971435547,
      "learning_rate": 5.445313428763705e-06,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8520261868834496,
      "num_tokens": 83750518.0,
      "step": 69630
    },
    {
      "entropy": 1.8970528110861777,
      "epoch": 0.21587790684607927,
      "grad_norm": 10.588910102844238,
      "learning_rate": 5.444922447494128e-06,
      "loss": 0.4665,
      "mean_token_accuracy": 0.8415069609880448,
      "num_tokens": 83763123.0,
      "step": 69640
    },
    {
      "entropy": 1.9143267586827277,
      "epoch": 0.21590890597112897,
      "grad_norm": 8.116543769836426,
      "learning_rate": 5.444531550431497e-06,
      "loss": 0.487,
      "mean_token_accuracy": 0.8432750076055526,
      "num_tokens": 83774896.0,
      "step": 69650
    },
    {
      "entropy": 1.8820920512080193,
      "epoch": 0.21593990509617866,
      "grad_norm": 9.194504737854004,
      "learning_rate": 5.444140737545589e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8378720536828042,
      "num_tokens": 83786830.0,
      "step": 69660
    },
    {
      "entropy": 1.8950974389910697,
      "epoch": 0.21597090422122836,
      "grad_norm": 9.50033187866211,
      "learning_rate": 5.443750008806198e-06,
      "loss": 0.5127,
      "mean_token_accuracy": 0.840063288807869,
      "num_tokens": 83798450.0,
      "step": 69670
    },
    {
      "entropy": 1.9445108637213706,
      "epoch": 0.21600190334627806,
      "grad_norm": 7.2820892333984375,
      "learning_rate": 5.443359364183132e-06,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8457172334194183,
      "num_tokens": 83809795.0,
      "step": 69680
    },
    {
      "entropy": 1.8929253578186036,
      "epoch": 0.21603290247132775,
      "grad_norm": 8.967961311340332,
      "learning_rate": 5.442968803646214e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8517738983035088,
      "num_tokens": 83821954.0,
      "step": 69690
    },
    {
      "entropy": 1.8933438271284104,
      "epoch": 0.21606390159637745,
      "grad_norm": 8.333928108215332,
      "learning_rate": 5.4425783271652824e-06,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8390064716339112,
      "num_tokens": 83834653.0,
      "step": 69700
    },
    {
      "entropy": 1.8290221512317657,
      "epoch": 0.21609490072142715,
      "grad_norm": 4.229883670806885,
      "learning_rate": 5.44218793471019e-06,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8560205176472664,
      "num_tokens": 83847132.0,
      "step": 69710
    },
    {
      "entropy": 1.9552450522780418,
      "epoch": 0.21612589984647684,
      "grad_norm": 8.421517372131348,
      "learning_rate": 5.441797626250808e-06,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8295035645365715,
      "num_tokens": 83858171.0,
      "step": 69720
    },
    {
      "entropy": 1.912724708020687,
      "epoch": 0.21615689897152654,
      "grad_norm": 7.742305278778076,
      "learning_rate": 5.441407401757017e-06,
      "loss": 0.5841,
      "mean_token_accuracy": 0.8269262716174126,
      "num_tokens": 83870397.0,
      "step": 69730
    },
    {
      "entropy": 1.9444334492087365,
      "epoch": 0.21618789809657624,
      "grad_norm": 4.496399402618408,
      "learning_rate": 5.44101726119872e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8388503283262253,
      "num_tokens": 83882469.0,
      "step": 69740
    },
    {
      "entropy": 1.8968336150050162,
      "epoch": 0.21621889722162593,
      "grad_norm": 7.71511697769165,
      "learning_rate": 5.440627204545827e-06,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8507803946733474,
      "num_tokens": 83894493.0,
      "step": 69750
    },
    {
      "entropy": 1.9784729063510895,
      "epoch": 0.2162498963466756,
      "grad_norm": 10.393377304077148,
      "learning_rate": 5.440237231768271e-06,
      "loss": 0.5628,
      "mean_token_accuracy": 0.8280197530984879,
      "num_tokens": 83905959.0,
      "step": 69760
    },
    {
      "entropy": 1.9202745750546455,
      "epoch": 0.2162808954717253,
      "grad_norm": 7.838612079620361,
      "learning_rate": 5.439847342835992e-06,
      "loss": 0.517,
      "mean_token_accuracy": 0.8408389419317246,
      "num_tokens": 83917828.0,
      "step": 69770
    },
    {
      "entropy": 1.9422619119286537,
      "epoch": 0.216311894596775,
      "grad_norm": 3.9136011600494385,
      "learning_rate": 5.439457537718953e-06,
      "loss": 0.474,
      "mean_token_accuracy": 0.8431269928812981,
      "num_tokens": 83929517.0,
      "step": 69780
    },
    {
      "entropy": 2.040539500117302,
      "epoch": 0.2163428937218247,
      "grad_norm": 10.292856216430664,
      "learning_rate": 5.4390678163871265e-06,
      "loss": 0.585,
      "mean_token_accuracy": 0.8160558506846428,
      "num_tokens": 83940294.0,
      "step": 69790
    },
    {
      "entropy": 2.000251492857933,
      "epoch": 0.2163738928468744,
      "grad_norm": 7.672717571258545,
      "learning_rate": 5.438678178810503e-06,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8404903680086135,
      "num_tokens": 83951400.0,
      "step": 69800
    },
    {
      "entropy": 1.9795056134462357,
      "epoch": 0.21640489197192408,
      "grad_norm": 9.820237159729004,
      "learning_rate": 5.438288624959086e-06,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8368617355823517,
      "num_tokens": 83963042.0,
      "step": 69810
    },
    {
      "entropy": 1.9222991615533829,
      "epoch": 0.21643589109697378,
      "grad_norm": 9.5084810256958,
      "learning_rate": 5.437899154802895e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8351835623383522,
      "num_tokens": 83974409.0,
      "step": 69820
    },
    {
      "entropy": 1.945703822374344,
      "epoch": 0.21646689022202348,
      "grad_norm": 11.023735046386719,
      "learning_rate": 5.437509768311967e-06,
      "loss": 0.5189,
      "mean_token_accuracy": 0.8369563966989517,
      "num_tokens": 83985791.0,
      "step": 69830
    },
    {
      "entropy": 1.8587691098451615,
      "epoch": 0.21649788934707317,
      "grad_norm": 3.4747183322906494,
      "learning_rate": 5.437120465456348e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8473239108920098,
      "num_tokens": 83998427.0,
      "step": 69840
    },
    {
      "entropy": 1.891299197077751,
      "epoch": 0.21652888847212287,
      "grad_norm": 7.577733993530273,
      "learning_rate": 5.436731246206105e-06,
      "loss": 0.4721,
      "mean_token_accuracy": 0.846482339501381,
      "num_tokens": 84010579.0,
      "step": 69850
    },
    {
      "entropy": 1.9225863426923753,
      "epoch": 0.21655988759717257,
      "grad_norm": 6.7725324630737305,
      "learning_rate": 5.436342110531316e-06,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8383674845099449,
      "num_tokens": 84022493.0,
      "step": 69860
    },
    {
      "entropy": 1.955282860994339,
      "epoch": 0.21659088672222226,
      "grad_norm": 9.114567756652832,
      "learning_rate": 5.435953058402078e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8373220950365067,
      "num_tokens": 84033749.0,
      "step": 69870
    },
    {
      "entropy": 2.0027921319007875,
      "epoch": 0.21662188584727196,
      "grad_norm": 8.859573364257812,
      "learning_rate": 5.435564089788498e-06,
      "loss": 0.5998,
      "mean_token_accuracy": 0.8251512601971627,
      "num_tokens": 84044757.0,
      "step": 69880
    },
    {
      "entropy": 1.8767832040786743,
      "epoch": 0.21665288497232166,
      "grad_norm": 8.931615829467773,
      "learning_rate": 5.435175204660702e-06,
      "loss": 0.516,
      "mean_token_accuracy": 0.8378856316208839,
      "num_tokens": 84056798.0,
      "step": 69890
    },
    {
      "entropy": 1.8918719656765461,
      "epoch": 0.21668388409737135,
      "grad_norm": 9.630752563476562,
      "learning_rate": 5.4347864029888284e-06,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8510532274842262,
      "num_tokens": 84068455.0,
      "step": 69900
    },
    {
      "entropy": 1.9158081158995628,
      "epoch": 0.21671488322242105,
      "grad_norm": 4.970503807067871,
      "learning_rate": 5.434397684743034e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8565353244543076,
      "num_tokens": 84080433.0,
      "step": 69910
    },
    {
      "entropy": 1.8488872632384301,
      "epoch": 0.21674588234747075,
      "grad_norm": 8.853670120239258,
      "learning_rate": 5.434009049893485e-06,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8482698664069176,
      "num_tokens": 84093473.0,
      "step": 69920
    },
    {
      "entropy": 1.941627450287342,
      "epoch": 0.21677688147252044,
      "grad_norm": 8.593602180480957,
      "learning_rate": 5.433620498410368e-06,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8356386333703995,
      "num_tokens": 84105334.0,
      "step": 69930
    },
    {
      "entropy": 1.9388030782341956,
      "epoch": 0.21680788059757014,
      "grad_norm": 8.22298526763916,
      "learning_rate": 5.433232030263881e-06,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8294572129845619,
      "num_tokens": 84117279.0,
      "step": 69940
    },
    {
      "entropy": 1.844594794511795,
      "epoch": 0.21683887972261984,
      "grad_norm": 9.116217613220215,
      "learning_rate": 5.432843645424239e-06,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8563014373183251,
      "num_tokens": 84129513.0,
      "step": 69950
    },
    {
      "entropy": 1.860558421909809,
      "epoch": 0.21686987884766953,
      "grad_norm": 8.21207332611084,
      "learning_rate": 5.43245534386167e-06,
      "loss": 0.449,
      "mean_token_accuracy": 0.8442764699459075,
      "num_tokens": 84142071.0,
      "step": 69960
    },
    {
      "entropy": 1.9477815553545952,
      "epoch": 0.21690087797271923,
      "grad_norm": 9.044229507446289,
      "learning_rate": 5.432067125546419e-06,
      "loss": 0.507,
      "mean_token_accuracy": 0.8379213094711304,
      "num_tokens": 84153180.0,
      "step": 69970
    },
    {
      "entropy": 1.9056160926818848,
      "epoch": 0.21693187709776893,
      "grad_norm": 8.381840705871582,
      "learning_rate": 5.431678990448746e-06,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8412555903196335,
      "num_tokens": 84165277.0,
      "step": 69980
    },
    {
      "entropy": 1.9368507653474807,
      "epoch": 0.21696287622281862,
      "grad_norm": 9.226876258850098,
      "learning_rate": 5.431290938538921e-06,
      "loss": 0.5589,
      "mean_token_accuracy": 0.8354489773511886,
      "num_tokens": 84176763.0,
      "step": 69990
    },
    {
      "entropy": 1.9172518044710158,
      "epoch": 0.21699387534786832,
      "grad_norm": 4.2976603507995605,
      "learning_rate": 5.430902969787236e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8291812673211097,
      "num_tokens": 84189241.0,
      "step": 70000
    },
    {
      "entropy": 1.8809144780039788,
      "epoch": 0.217024874472918,
      "grad_norm": 7.371922016143799,
      "learning_rate": 5.430515084163993e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8422490224242211,
      "num_tokens": 84201347.0,
      "step": 70010
    },
    {
      "entropy": 1.951662066578865,
      "epoch": 0.21705587359796769,
      "grad_norm": 10.154802322387695,
      "learning_rate": 5.43012728163951e-06,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8398729935288429,
      "num_tokens": 84213095.0,
      "step": 70020
    },
    {
      "entropy": 1.8219065442681313,
      "epoch": 0.21708687272301738,
      "grad_norm": 8.22426700592041,
      "learning_rate": 5.429739562184121e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8590575277805328,
      "num_tokens": 84225978.0,
      "step": 70030
    },
    {
      "entropy": 1.8591913282871246,
      "epoch": 0.21711787184806708,
      "grad_norm": 9.920334815979004,
      "learning_rate": 5.429351925768173e-06,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8464103743433953,
      "num_tokens": 84239200.0,
      "step": 70040
    },
    {
      "entropy": 1.88853762447834,
      "epoch": 0.21714887097311678,
      "grad_norm": 4.088785171508789,
      "learning_rate": 5.428964372362031e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8548205360770226,
      "num_tokens": 84252272.0,
      "step": 70050
    },
    {
      "entropy": 1.8458364441990853,
      "epoch": 0.21717987009816647,
      "grad_norm": 9.360186576843262,
      "learning_rate": 5.428576901936069e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.852430385351181,
      "num_tokens": 84265230.0,
      "step": 70060
    },
    {
      "entropy": 1.9471904829144477,
      "epoch": 0.21721086922321617,
      "grad_norm": 3.6564393043518066,
      "learning_rate": 5.428189514460681e-06,
      "loss": 0.484,
      "mean_token_accuracy": 0.8501205831766129,
      "num_tokens": 84276446.0,
      "step": 70070
    },
    {
      "entropy": 1.8922279462218285,
      "epoch": 0.21724186834826587,
      "grad_norm": 3.798719644546509,
      "learning_rate": 5.427802209906274e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.838223971426487,
      "num_tokens": 84288449.0,
      "step": 70080
    },
    {
      "entropy": 1.9325299307703971,
      "epoch": 0.21727286747331556,
      "grad_norm": 9.839792251586914,
      "learning_rate": 5.427414988243273e-06,
      "loss": 0.5621,
      "mean_token_accuracy": 0.83007542937994,
      "num_tokens": 84299853.0,
      "step": 70090
    },
    {
      "entropy": 1.8988226309418679,
      "epoch": 0.21730386659836526,
      "grad_norm": 7.719738960266113,
      "learning_rate": 5.427027849442109e-06,
      "loss": 0.4806,
      "mean_token_accuracy": 0.8383999258279801,
      "num_tokens": 84312190.0,
      "step": 70100
    },
    {
      "entropy": 1.8617843106389045,
      "epoch": 0.21733486572341496,
      "grad_norm": 8.315159797668457,
      "learning_rate": 5.426640793473237e-06,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8574797883629799,
      "num_tokens": 84324677.0,
      "step": 70110
    },
    {
      "entropy": 1.9673965632915498,
      "epoch": 0.21736586484846465,
      "grad_norm": 8.235828399658203,
      "learning_rate": 5.426253820307122e-06,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8495687767863274,
      "num_tokens": 84335718.0,
      "step": 70120
    },
    {
      "entropy": 1.9207556203007699,
      "epoch": 0.21739686397351435,
      "grad_norm": 8.374682426452637,
      "learning_rate": 5.4258669299142465e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8351808488368988,
      "num_tokens": 84347566.0,
      "step": 70130
    },
    {
      "entropy": 1.9010922938585282,
      "epoch": 0.21742786309856404,
      "grad_norm": 9.063957214355469,
      "learning_rate": 5.425480122265106e-06,
      "loss": 0.501,
      "mean_token_accuracy": 0.8409019201993942,
      "num_tokens": 84360542.0,
      "step": 70140
    },
    {
      "entropy": 1.9091226264834404,
      "epoch": 0.21745886222361374,
      "grad_norm": 9.379876136779785,
      "learning_rate": 5.425093397330208e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8458316907286644,
      "num_tokens": 84372835.0,
      "step": 70150
    },
    {
      "entropy": 1.7553482845425605,
      "epoch": 0.21748986134866344,
      "grad_norm": 3.750493288040161,
      "learning_rate": 5.424706755080079e-06,
      "loss": 0.4027,
      "mean_token_accuracy": 0.8478049889206887,
      "num_tokens": 84387102.0,
      "step": 70160
    },
    {
      "entropy": 1.9269262328743935,
      "epoch": 0.21752086047371313,
      "grad_norm": 11.567874908447266,
      "learning_rate": 5.4243201954852605e-06,
      "loss": 0.5347,
      "mean_token_accuracy": 0.837294514477253,
      "num_tokens": 84398445.0,
      "step": 70170
    },
    {
      "entropy": 1.757249329984188,
      "epoch": 0.21755185959876283,
      "grad_norm": 2.541597604751587,
      "learning_rate": 5.423933718516307e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8519966840744019,
      "num_tokens": 84412057.0,
      "step": 70180
    },
    {
      "entropy": 1.9713240012526512,
      "epoch": 0.21758285872381253,
      "grad_norm": 9.577204704284668,
      "learning_rate": 5.423547324143784e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8448547378182412,
      "num_tokens": 84423064.0,
      "step": 70190
    },
    {
      "entropy": 1.9091683775186539,
      "epoch": 0.21761385784886222,
      "grad_norm": 7.392279624938965,
      "learning_rate": 5.423161012338279e-06,
      "loss": 0.494,
      "mean_token_accuracy": 0.8476917743682861,
      "num_tokens": 84435063.0,
      "step": 70200
    },
    {
      "entropy": 1.926077064871788,
      "epoch": 0.21764485697391192,
      "grad_norm": 8.991741180419922,
      "learning_rate": 5.422774783070391e-06,
      "loss": 0.542,
      "mean_token_accuracy": 0.8332406044006347,
      "num_tokens": 84446653.0,
      "step": 70210
    },
    {
      "entropy": 1.9285721063613892,
      "epoch": 0.21767585609896162,
      "grad_norm": 4.427052974700928,
      "learning_rate": 5.42238863631073e-06,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8307848066091538,
      "num_tokens": 84458387.0,
      "step": 70220
    },
    {
      "entropy": 1.8635422542691231,
      "epoch": 0.21770685522401131,
      "grad_norm": 8.436079025268555,
      "learning_rate": 5.422002572029925e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8484039083123207,
      "num_tokens": 84470516.0,
      "step": 70230
    },
    {
      "entropy": 1.9386792957782746,
      "epoch": 0.217737854349061,
      "grad_norm": 4.04444694519043,
      "learning_rate": 5.421616590198619e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8558227822184563,
      "num_tokens": 84481818.0,
      "step": 70240
    },
    {
      "entropy": 1.9428984984755515,
      "epoch": 0.21776885347411068,
      "grad_norm": 7.593319892883301,
      "learning_rate": 5.4212306907874705e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8404227659106255,
      "num_tokens": 84493867.0,
      "step": 70250
    },
    {
      "entropy": 1.854594275355339,
      "epoch": 0.21779985259916038,
      "grad_norm": 10.006421089172363,
      "learning_rate": 5.420844873767147e-06,
      "loss": 0.4851,
      "mean_token_accuracy": 0.8399219319224358,
      "num_tokens": 84506467.0,
      "step": 70260
    },
    {
      "entropy": 1.9580897375941277,
      "epoch": 0.21783085172421007,
      "grad_norm": 9.24285888671875,
      "learning_rate": 5.420459139108339e-06,
      "loss": 0.5512,
      "mean_token_accuracy": 0.8321612671017646,
      "num_tokens": 84517655.0,
      "step": 70270
    },
    {
      "entropy": 1.9946820080280303,
      "epoch": 0.21786185084925977,
      "grad_norm": 9.095592498779297,
      "learning_rate": 5.420073486781746e-06,
      "loss": 0.5713,
      "mean_token_accuracy": 0.8361142173409462,
      "num_tokens": 84527947.0,
      "step": 70280
    },
    {
      "entropy": 1.819739530980587,
      "epoch": 0.21789284997430947,
      "grad_norm": 9.337332725524902,
      "learning_rate": 5.419687916758083e-06,
      "loss": 0.414,
      "mean_token_accuracy": 0.857285912334919,
      "num_tokens": 84540966.0,
      "step": 70290
    },
    {
      "entropy": 1.959464368224144,
      "epoch": 0.21792384909935916,
      "grad_norm": 9.26564884185791,
      "learning_rate": 5.419302429008081e-06,
      "loss": 0.5639,
      "mean_token_accuracy": 0.8280451089143753,
      "num_tokens": 84552392.0,
      "step": 70300
    },
    {
      "entropy": 1.824575850367546,
      "epoch": 0.21795484822440886,
      "grad_norm": 8.046067237854004,
      "learning_rate": 5.418917023502482e-06,
      "loss": 0.496,
      "mean_token_accuracy": 0.8442411884665489,
      "num_tokens": 84565104.0,
      "step": 70310
    },
    {
      "entropy": 1.9438457012176513,
      "epoch": 0.21798584734945856,
      "grad_norm": 8.829744338989258,
      "learning_rate": 5.41853170021205e-06,
      "loss": 0.5719,
      "mean_token_accuracy": 0.8261488035321236,
      "num_tokens": 84576914.0,
      "step": 70320
    },
    {
      "entropy": 1.9732747316360473,
      "epoch": 0.21801684647450825,
      "grad_norm": 8.278334617614746,
      "learning_rate": 5.418146459107554e-06,
      "loss": 0.5788,
      "mean_token_accuracy": 0.8282134965062141,
      "num_tokens": 84588102.0,
      "step": 70330
    },
    {
      "entropy": 1.9039635524153709,
      "epoch": 0.21804784559955795,
      "grad_norm": 7.718066692352295,
      "learning_rate": 5.417761300159784e-06,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8389883413910866,
      "num_tokens": 84599955.0,
      "step": 70340
    },
    {
      "entropy": 1.910678581893444,
      "epoch": 0.21807884472460765,
      "grad_norm": 3.6842432022094727,
      "learning_rate": 5.417376223339544e-06,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8448718905448913,
      "num_tokens": 84611103.0,
      "step": 70350
    },
    {
      "entropy": 1.8811346575617791,
      "epoch": 0.21810984384965734,
      "grad_norm": 10.019588470458984,
      "learning_rate": 5.41699122861765e-06,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8373315662145615,
      "num_tokens": 84622901.0,
      "step": 70360
    },
    {
      "entropy": 1.751522246003151,
      "epoch": 0.21814084297470704,
      "grad_norm": 9.472254753112793,
      "learning_rate": 5.416606315964937e-06,
      "loss": 0.4021,
      "mean_token_accuracy": 0.8536264047026634,
      "num_tokens": 84636784.0,
      "step": 70370
    },
    {
      "entropy": 1.9128979295492172,
      "epoch": 0.21817184209975674,
      "grad_norm": 7.596695899963379,
      "learning_rate": 5.416221485352247e-06,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8282625824213028,
      "num_tokens": 84649019.0,
      "step": 70380
    },
    {
      "entropy": 1.90406776368618,
      "epoch": 0.21820284122480643,
      "grad_norm": 9.435015678405762,
      "learning_rate": 5.415836736750441e-06,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8421972319483757,
      "num_tokens": 84660281.0,
      "step": 70390
    },
    {
      "entropy": 1.77169189453125,
      "epoch": 0.21823384034985613,
      "grad_norm": 3.901792287826538,
      "learning_rate": 5.415452070130397e-06,
      "loss": 0.3818,
      "mean_token_accuracy": 0.8625829800963402,
      "num_tokens": 84673251.0,
      "step": 70400
    },
    {
      "entropy": 1.9100979268550873,
      "epoch": 0.21826483947490583,
      "grad_norm": 8.25220012664795,
      "learning_rate": 5.415067485463005e-06,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8386782616376877,
      "num_tokens": 84685476.0,
      "step": 70410
    },
    {
      "entropy": 1.8133963495492935,
      "epoch": 0.21829583859995552,
      "grad_norm": 3.5599453449249268,
      "learning_rate": 5.414682982719167e-06,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8472143262624741,
      "num_tokens": 84698468.0,
      "step": 70420
    },
    {
      "entropy": 1.9142759516835213,
      "epoch": 0.21832683772500522,
      "grad_norm": 7.450267314910889,
      "learning_rate": 5.414298561869803e-06,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8527796879410744,
      "num_tokens": 84709995.0,
      "step": 70430
    },
    {
      "entropy": 1.9114891082048415,
      "epoch": 0.21835783685005491,
      "grad_norm": 8.194131851196289,
      "learning_rate": 5.413914222885847e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.838103885948658,
      "num_tokens": 84720745.0,
      "step": 70440
    },
    {
      "entropy": 1.8565336361527442,
      "epoch": 0.2183888359751046,
      "grad_norm": 7.725749969482422,
      "learning_rate": 5.413529965738245e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8396166414022446,
      "num_tokens": 84732619.0,
      "step": 70450
    },
    {
      "entropy": 1.933957888185978,
      "epoch": 0.2184198351001543,
      "grad_norm": 9.328766822814941,
      "learning_rate": 5.41314579039796e-06,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8291135087609292,
      "num_tokens": 84744044.0,
      "step": 70460
    },
    {
      "entropy": 1.881012487411499,
      "epoch": 0.218450834225204,
      "grad_norm": 7.425388813018799,
      "learning_rate": 5.412761696835969e-06,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8454090863466263,
      "num_tokens": 84756091.0,
      "step": 70470
    },
    {
      "entropy": 1.8883946731686592,
      "epoch": 0.2184818333502537,
      "grad_norm": 7.838623523712158,
      "learning_rate": 5.412377685023262e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8344195529818534,
      "num_tokens": 84768181.0,
      "step": 70480
    },
    {
      "entropy": 1.8858201622962951,
      "epoch": 0.2185128324753034,
      "grad_norm": 8.257635116577148,
      "learning_rate": 5.411993754930844e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8470271736383438,
      "num_tokens": 84779613.0,
      "step": 70490
    },
    {
      "entropy": 1.87155821621418,
      "epoch": 0.21854383160035307,
      "grad_norm": 11.120357513427734,
      "learning_rate": 5.411609906529737e-06,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8390228837728501,
      "num_tokens": 84792236.0,
      "step": 70500
    },
    {
      "entropy": 1.836224192380905,
      "epoch": 0.21857483072540276,
      "grad_norm": 9.826521873474121,
      "learning_rate": 5.411226139790973e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8449719101190567,
      "num_tokens": 84805022.0,
      "step": 70510
    },
    {
      "entropy": 1.9778549373149872,
      "epoch": 0.21860582985045246,
      "grad_norm": 9.500571250915527,
      "learning_rate": 5.410842454685601e-06,
      "loss": 0.6753,
      "mean_token_accuracy": 0.8305225998163224,
      "num_tokens": 84816577.0,
      "step": 70520
    },
    {
      "entropy": 1.8446165218949317,
      "epoch": 0.21863682897550216,
      "grad_norm": 9.461227416992188,
      "learning_rate": 5.4104588511846846e-06,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8477078288793564,
      "num_tokens": 84829132.0,
      "step": 70530
    },
    {
      "entropy": 1.859433715045452,
      "epoch": 0.21866782810055185,
      "grad_norm": 7.800492286682129,
      "learning_rate": 5.410075329259299e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8472880437970162,
      "num_tokens": 84841353.0,
      "step": 70540
    },
    {
      "entropy": 1.9165452167391777,
      "epoch": 0.21869882722560155,
      "grad_norm": 8.318398475646973,
      "learning_rate": 5.4096918888805385e-06,
      "loss": 0.5168,
      "mean_token_accuracy": 0.8420958235859871,
      "num_tokens": 84852397.0,
      "step": 70550
    },
    {
      "entropy": 1.8914957597851754,
      "epoch": 0.21872982635065125,
      "grad_norm": 9.693846702575684,
      "learning_rate": 5.409308530019507e-06,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8405283436179161,
      "num_tokens": 84863491.0,
      "step": 70560
    },
    {
      "entropy": 1.945120519399643,
      "epoch": 0.21876082547570094,
      "grad_norm": 9.057845115661621,
      "learning_rate": 5.408925252647326e-06,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8449465945363045,
      "num_tokens": 84874898.0,
      "step": 70570
    },
    {
      "entropy": 1.9313878312706947,
      "epoch": 0.21879182460075064,
      "grad_norm": 8.234183311462402,
      "learning_rate": 5.408542056735129e-06,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8252523899078369,
      "num_tokens": 84886394.0,
      "step": 70580
    },
    {
      "entropy": 1.9243649192154408,
      "epoch": 0.21882282372580034,
      "grad_norm": 8.189180374145508,
      "learning_rate": 5.408158942254065e-06,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8330753937363624,
      "num_tokens": 84898510.0,
      "step": 70590
    },
    {
      "entropy": 1.8462709829211235,
      "epoch": 0.21885382285085003,
      "grad_norm": 7.798243522644043,
      "learning_rate": 5.407775909175298e-06,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8495460674166679,
      "num_tokens": 84910723.0,
      "step": 70600
    },
    {
      "entropy": 1.8084265619516373,
      "epoch": 0.21888482197589973,
      "grad_norm": 10.60255241394043,
      "learning_rate": 5.407392957470005e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8456479325890541,
      "num_tokens": 84923960.0,
      "step": 70610
    },
    {
      "entropy": 1.880794422328472,
      "epoch": 0.21891582110094943,
      "grad_norm": 4.523313045501709,
      "learning_rate": 5.4070100871093764e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8479678109288216,
      "num_tokens": 84935570.0,
      "step": 70620
    },
    {
      "entropy": 1.8770657986402512,
      "epoch": 0.21894682022599912,
      "grad_norm": 2.5397913455963135,
      "learning_rate": 5.406627298064622e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8388467326760292,
      "num_tokens": 84948444.0,
      "step": 70630
    },
    {
      "entropy": 1.946268130838871,
      "epoch": 0.21897781935104882,
      "grad_norm": 9.20574951171875,
      "learning_rate": 5.406244590306958e-06,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8426952421665191,
      "num_tokens": 84959437.0,
      "step": 70640
    },
    {
      "entropy": 1.9465517818927764,
      "epoch": 0.21900881847609852,
      "grad_norm": 9.563549995422363,
      "learning_rate": 5.405861963807622e-06,
      "loss": 0.5409,
      "mean_token_accuracy": 0.834696726500988,
      "num_tokens": 84970287.0,
      "step": 70650
    },
    {
      "entropy": 1.9270946726202964,
      "epoch": 0.2190398176011482,
      "grad_norm": 7.783606052398682,
      "learning_rate": 5.4054794185378615e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8351917177438736,
      "num_tokens": 84982065.0,
      "step": 70660
    },
    {
      "entropy": 1.8967857383191586,
      "epoch": 0.2190708167261979,
      "grad_norm": 8.703535079956055,
      "learning_rate": 5.405096954468938e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8460932150483131,
      "num_tokens": 84993902.0,
      "step": 70670
    },
    {
      "entropy": 1.9372851148247718,
      "epoch": 0.2191018158512476,
      "grad_norm": 4.806000232696533,
      "learning_rate": 5.4047145715721315e-06,
      "loss": 0.566,
      "mean_token_accuracy": 0.8275101691484451,
      "num_tokens": 85005220.0,
      "step": 70680
    },
    {
      "entropy": 1.88345315605402,
      "epoch": 0.2191328149762973,
      "grad_norm": 7.77454948425293,
      "learning_rate": 5.404332269818732e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8458261653780937,
      "num_tokens": 85017018.0,
      "step": 70690
    },
    {
      "entropy": 1.7957565858960152,
      "epoch": 0.219163814101347,
      "grad_norm": 8.677156448364258,
      "learning_rate": 5.403950049180046e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8442620366811753,
      "num_tokens": 85030620.0,
      "step": 70700
    },
    {
      "entropy": 1.8943561017513275,
      "epoch": 0.2191948132263967,
      "grad_norm": 7.602232933044434,
      "learning_rate": 5.403567909627393e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8428284287452698,
      "num_tokens": 85041835.0,
      "step": 70710
    },
    {
      "entropy": 1.9183341562747955,
      "epoch": 0.2192258123514464,
      "grad_norm": 8.18143367767334,
      "learning_rate": 5.4031858511321065e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8351936027407646,
      "num_tokens": 85053607.0,
      "step": 70720
    },
    {
      "entropy": 1.8503445282578468,
      "epoch": 0.2192568114764961,
      "grad_norm": 8.790541648864746,
      "learning_rate": 5.402803873665535e-06,
      "loss": 0.414,
      "mean_token_accuracy": 0.8500261098146439,
      "num_tokens": 85065898.0,
      "step": 70730
    },
    {
      "entropy": 1.869480137526989,
      "epoch": 0.21928781060154579,
      "grad_norm": 10.168290138244629,
      "learning_rate": 5.402421977199042e-06,
      "loss": 0.4884,
      "mean_token_accuracy": 0.838524155318737,
      "num_tokens": 85078033.0,
      "step": 70740
    },
    {
      "entropy": 1.9085437297821044,
      "epoch": 0.21931880972659545,
      "grad_norm": 8.224616050720215,
      "learning_rate": 5.402040161704004e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8398436903953552,
      "num_tokens": 85090009.0,
      "step": 70750
    },
    {
      "entropy": 1.8690147161483766,
      "epoch": 0.21934980885164515,
      "grad_norm": 7.7431416511535645,
      "learning_rate": 5.4016584271518116e-06,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8496010720729827,
      "num_tokens": 85102267.0,
      "step": 70760
    },
    {
      "entropy": 1.908894456923008,
      "epoch": 0.21938080797669485,
      "grad_norm": 9.895110130310059,
      "learning_rate": 5.401276773513869e-06,
      "loss": 0.5064,
      "mean_token_accuracy": 0.839865879714489,
      "num_tokens": 85113947.0,
      "step": 70770
    },
    {
      "entropy": 1.8951399207115174,
      "epoch": 0.21941180710174454,
      "grad_norm": 7.593353271484375,
      "learning_rate": 5.400895200761596e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8396402359008789,
      "num_tokens": 85125637.0,
      "step": 70780
    },
    {
      "entropy": 1.8372249327600003,
      "epoch": 0.21944280622679424,
      "grad_norm": 10.441277503967285,
      "learning_rate": 5.400513708866425e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8503782153129578,
      "num_tokens": 85138593.0,
      "step": 70790
    },
    {
      "entropy": 1.9151126846671105,
      "epoch": 0.21947380535184394,
      "grad_norm": 11.181161880493164,
      "learning_rate": 5.400132297799804e-06,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8408087193965912,
      "num_tokens": 85150130.0,
      "step": 70800
    },
    {
      "entropy": 1.9188938543200493,
      "epoch": 0.21950480447689363,
      "grad_norm": 7.681055545806885,
      "learning_rate": 5.399750967533195e-06,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8343793347477912,
      "num_tokens": 85161618.0,
      "step": 70810
    },
    {
      "entropy": 1.8500541925430298,
      "epoch": 0.21953580360194333,
      "grad_norm": 9.042118072509766,
      "learning_rate": 5.399369718038073e-06,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8330987498164177,
      "num_tokens": 85174263.0,
      "step": 70820
    },
    {
      "entropy": 1.8322535023093223,
      "epoch": 0.21956680272699303,
      "grad_norm": 4.178494930267334,
      "learning_rate": 5.398988549285927e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8345492288470269,
      "num_tokens": 85187028.0,
      "step": 70830
    },
    {
      "entropy": 1.8351117119193077,
      "epoch": 0.21959780185204272,
      "grad_norm": 4.005007743835449,
      "learning_rate": 5.398607461248263e-06,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8548778548836709,
      "num_tokens": 85199485.0,
      "step": 70840
    },
    {
      "entropy": 1.8513669028878212,
      "epoch": 0.21962880097709242,
      "grad_norm": 8.433499336242676,
      "learning_rate": 5.398226453896596e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8487704336643219,
      "num_tokens": 85211879.0,
      "step": 70850
    },
    {
      "entropy": 1.8096089884638786,
      "epoch": 0.21965980010214212,
      "grad_norm": 10.63721752166748,
      "learning_rate": 5.39784552720246e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.8548229023814201,
      "num_tokens": 85224815.0,
      "step": 70860
    },
    {
      "entropy": 1.8579236298799515,
      "epoch": 0.2196907992271918,
      "grad_norm": 8.522771835327148,
      "learning_rate": 5.3974646811373986e-06,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8453032687306404,
      "num_tokens": 85236804.0,
      "step": 70870
    },
    {
      "entropy": 1.8571562334895133,
      "epoch": 0.2197217983522415,
      "grad_norm": 8.076016426086426,
      "learning_rate": 5.397083915672975e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8438248604536056,
      "num_tokens": 85248947.0,
      "step": 70880
    },
    {
      "entropy": 1.851157009601593,
      "epoch": 0.2197527974772912,
      "grad_norm": 9.514842987060547,
      "learning_rate": 5.396703230780761e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8403349593281746,
      "num_tokens": 85261568.0,
      "step": 70890
    },
    {
      "entropy": 1.8390909820795058,
      "epoch": 0.2197837966023409,
      "grad_norm": 9.045916557312012,
      "learning_rate": 5.396322626432345e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8455731064081192,
      "num_tokens": 85274322.0,
      "step": 70900
    },
    {
      "entropy": 1.9245983332395553,
      "epoch": 0.2198147957273906,
      "grad_norm": 7.920559406280518,
      "learning_rate": 5.39594210259933e-06,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8428880482912063,
      "num_tokens": 85285617.0,
      "step": 70910
    },
    {
      "entropy": 1.8781443014740944,
      "epoch": 0.2198457948524403,
      "grad_norm": 4.677145481109619,
      "learning_rate": 5.395561659253331e-06,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8449257969856262,
      "num_tokens": 85297208.0,
      "step": 70920
    },
    {
      "entropy": 1.9284519299864769,
      "epoch": 0.21987679397749,
      "grad_norm": 3.986211061477661,
      "learning_rate": 5.395181296365979e-06,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8414773017168045,
      "num_tokens": 85308876.0,
      "step": 70930
    },
    {
      "entropy": 1.7243415489792824,
      "epoch": 0.2199077931025397,
      "grad_norm": 4.131589412689209,
      "learning_rate": 5.394801013908917e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.865143957734108,
      "num_tokens": 85323468.0,
      "step": 70940
    },
    {
      "entropy": 1.9777496784925461,
      "epoch": 0.21993879222758939,
      "grad_norm": 7.485686779022217,
      "learning_rate": 5.394420811853805e-06,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8345079109072685,
      "num_tokens": 85334331.0,
      "step": 70950
    },
    {
      "entropy": 1.9227191910147667,
      "epoch": 0.21996979135263908,
      "grad_norm": 4.045227527618408,
      "learning_rate": 5.394040690172313e-06,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8265855267643929,
      "num_tokens": 85346105.0,
      "step": 70960
    },
    {
      "entropy": 1.9031540900468826,
      "epoch": 0.22000079047768878,
      "grad_norm": 4.864028453826904,
      "learning_rate": 5.393660648836128e-06,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8400901973247528,
      "num_tokens": 85358132.0,
      "step": 70970
    },
    {
      "entropy": 1.98341506421566,
      "epoch": 0.22003178960273848,
      "grad_norm": 8.267914772033691,
      "learning_rate": 5.393280687816951e-06,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8373934581875802,
      "num_tokens": 85369306.0,
      "step": 70980
    },
    {
      "entropy": 1.9509702712297439,
      "epoch": 0.22006278872778814,
      "grad_norm": 8.570839881896973,
      "learning_rate": 5.392900807086495e-06,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8445903062820435,
      "num_tokens": 85380557.0,
      "step": 70990
    },
    {
      "entropy": 1.9639266401529312,
      "epoch": 0.22009378785283784,
      "grad_norm": 9.221185684204102,
      "learning_rate": 5.392521006616488e-06,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8395860701799392,
      "num_tokens": 85392420.0,
      "step": 71000
    },
    {
      "entropy": 1.9102492406964302,
      "epoch": 0.22012478697788754,
      "grad_norm": 3.7444820404052734,
      "learning_rate": 5.392141286378672e-06,
      "loss": 0.497,
      "mean_token_accuracy": 0.8374612465500831,
      "num_tokens": 85404213.0,
      "step": 71010
    },
    {
      "entropy": 1.8868770197033882,
      "epoch": 0.22015578610293723,
      "grad_norm": 2.842402219772339,
      "learning_rate": 5.391761646344802e-06,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8464625418186188,
      "num_tokens": 85416634.0,
      "step": 71020
    },
    {
      "entropy": 1.8559535294771194,
      "epoch": 0.22018678522798693,
      "grad_norm": 8.821431159973145,
      "learning_rate": 5.391382086486649e-06,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8445154055953026,
      "num_tokens": 85430022.0,
      "step": 71030
    },
    {
      "entropy": 1.9842773735523225,
      "epoch": 0.22021778435303663,
      "grad_norm": 8.328567504882812,
      "learning_rate": 5.391002606775996e-06,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8406598642468452,
      "num_tokens": 85441410.0,
      "step": 71040
    },
    {
      "entropy": 1.9358683452010155,
      "epoch": 0.22024878347808632,
      "grad_norm": 9.156634330749512,
      "learning_rate": 5.3906232071846385e-06,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8384227842092514,
      "num_tokens": 85453655.0,
      "step": 71050
    },
    {
      "entropy": 1.8562579780817032,
      "epoch": 0.22027978260313602,
      "grad_norm": 7.700805187225342,
      "learning_rate": 5.390243887684392e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.854043036699295,
      "num_tokens": 85466539.0,
      "step": 71060
    },
    {
      "entropy": 1.9701458364725113,
      "epoch": 0.22031078172818572,
      "grad_norm": 7.824542045593262,
      "learning_rate": 5.38986464824708e-06,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8268503859639168,
      "num_tokens": 85478191.0,
      "step": 71070
    },
    {
      "entropy": 1.9239497467875482,
      "epoch": 0.2203417808532354,
      "grad_norm": 10.349506378173828,
      "learning_rate": 5.3894854888445415e-06,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8427152633666992,
      "num_tokens": 85490546.0,
      "step": 71080
    },
    {
      "entropy": 1.9393352545797824,
      "epoch": 0.2203727799782851,
      "grad_norm": 8.5408935546875,
      "learning_rate": 5.389106409448628e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8323530539870262,
      "num_tokens": 85502425.0,
      "step": 71090
    },
    {
      "entropy": 1.997844734787941,
      "epoch": 0.2204037791033348,
      "grad_norm": 7.891308307647705,
      "learning_rate": 5.38872741003121e-06,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8281294673681259,
      "num_tokens": 85513168.0,
      "step": 71100
    },
    {
      "entropy": 1.9478082045912744,
      "epoch": 0.2204347782283845,
      "grad_norm": 8.2448091506958,
      "learning_rate": 5.388348490564164e-06,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8402855768799782,
      "num_tokens": 85524556.0,
      "step": 71110
    },
    {
      "entropy": 1.931041233241558,
      "epoch": 0.2204657773534342,
      "grad_norm": 9.231786727905273,
      "learning_rate": 5.387969651019387e-06,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8433810248970985,
      "num_tokens": 85536049.0,
      "step": 71120
    },
    {
      "entropy": 1.9640550106763839,
      "epoch": 0.2204967764784839,
      "grad_norm": 8.980378150939941,
      "learning_rate": 5.387590891368787e-06,
      "loss": 0.5885,
      "mean_token_accuracy": 0.8172830164432525,
      "num_tokens": 85547228.0,
      "step": 71130
    },
    {
      "entropy": 1.8748082131147386,
      "epoch": 0.2205277756035336,
      "grad_norm": 8.130640983581543,
      "learning_rate": 5.387212211584286e-06,
      "loss": 0.4647,
      "mean_token_accuracy": 0.838733246922493,
      "num_tokens": 85559551.0,
      "step": 71140
    },
    {
      "entropy": 1.8640730693936347,
      "epoch": 0.2205587747285833,
      "grad_norm": 7.803213596343994,
      "learning_rate": 5.386833611637822e-06,
      "loss": 0.5066,
      "mean_token_accuracy": 0.844725139439106,
      "num_tokens": 85571361.0,
      "step": 71150
    },
    {
      "entropy": 1.8213941514492036,
      "epoch": 0.220589773853633,
      "grad_norm": 9.419188499450684,
      "learning_rate": 5.386455091501342e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8451396659016609,
      "num_tokens": 85584550.0,
      "step": 71160
    },
    {
      "entropy": 1.8446722000837326,
      "epoch": 0.22062077297868268,
      "grad_norm": 9.770685195922852,
      "learning_rate": 5.3860766511468095e-06,
      "loss": 0.4354,
      "mean_token_accuracy": 0.849824532866478,
      "num_tokens": 85597173.0,
      "step": 71170
    },
    {
      "entropy": 1.8369273975491525,
      "epoch": 0.22065177210373238,
      "grad_norm": 7.802650451660156,
      "learning_rate": 5.385698290546205e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8467299669981003,
      "num_tokens": 85611218.0,
      "step": 71180
    },
    {
      "entropy": 1.9682567581534385,
      "epoch": 0.22068277122878208,
      "grad_norm": 8.1614408493042,
      "learning_rate": 5.3853200096715175e-06,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8387658104300499,
      "num_tokens": 85622538.0,
      "step": 71190
    },
    {
      "entropy": 1.8726427100598813,
      "epoch": 0.22071377035383177,
      "grad_norm": 7.950981616973877,
      "learning_rate": 5.384941808494753e-06,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8444047793745995,
      "num_tokens": 85635421.0,
      "step": 71200
    },
    {
      "entropy": 1.893759909272194,
      "epoch": 0.22074476947888147,
      "grad_norm": 8.878076553344727,
      "learning_rate": 5.384563686987928e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8406562879681587,
      "num_tokens": 85648124.0,
      "step": 71210
    },
    {
      "entropy": 1.9666777163743974,
      "epoch": 0.22077576860393117,
      "grad_norm": 7.653510093688965,
      "learning_rate": 5.384185645123078e-06,
      "loss": 0.5379,
      "mean_token_accuracy": 0.840550334751606,
      "num_tokens": 85659636.0,
      "step": 71220
    },
    {
      "entropy": 1.898690427839756,
      "epoch": 0.22080676772898086,
      "grad_norm": 4.228546619415283,
      "learning_rate": 5.383807682872247e-06,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8438500598073005,
      "num_tokens": 85671836.0,
      "step": 71230
    },
    {
      "entropy": 1.9357111945748329,
      "epoch": 0.22083776685403053,
      "grad_norm": 5.342586517333984,
      "learning_rate": 5.383429800207497e-06,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8345894768834115,
      "num_tokens": 85684340.0,
      "step": 71240
    },
    {
      "entropy": 1.9341301143169403,
      "epoch": 0.22086876597908023,
      "grad_norm": 8.754364013671875,
      "learning_rate": 5.3830519971009e-06,
      "loss": 0.5377,
      "mean_token_accuracy": 0.8313232839107514,
      "num_tokens": 85695420.0,
      "step": 71250
    },
    {
      "entropy": 1.9368458151817323,
      "epoch": 0.22089976510412992,
      "grad_norm": 8.349538803100586,
      "learning_rate": 5.382674273524544e-06,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8416445463895798,
      "num_tokens": 85707345.0,
      "step": 71260
    },
    {
      "entropy": 1.7997360065579415,
      "epoch": 0.22093076422917962,
      "grad_norm": 8.042038917541504,
      "learning_rate": 5.382296629450529e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.845326641201973,
      "num_tokens": 85720742.0,
      "step": 71270
    },
    {
      "entropy": 1.958139282464981,
      "epoch": 0.22096176335422932,
      "grad_norm": 10.129528045654297,
      "learning_rate": 5.3819190648509714e-06,
      "loss": 0.555,
      "mean_token_accuracy": 0.8313161879777908,
      "num_tokens": 85730990.0,
      "step": 71280
    },
    {
      "entropy": 1.959393960237503,
      "epoch": 0.22099276247927901,
      "grad_norm": 8.495170593261719,
      "learning_rate": 5.381541579697999e-06,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8422238364815712,
      "num_tokens": 85742428.0,
      "step": 71290
    },
    {
      "entropy": 1.851896020770073,
      "epoch": 0.2210237616043287,
      "grad_norm": 8.348380088806152,
      "learning_rate": 5.381164173963755e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8479127526283264,
      "num_tokens": 85755347.0,
      "step": 71300
    },
    {
      "entropy": 1.875650581717491,
      "epoch": 0.2210547607293784,
      "grad_norm": 3.1359446048736572,
      "learning_rate": 5.380786847620394e-06,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8393297314643859,
      "num_tokens": 85767657.0,
      "step": 71310
    },
    {
      "entropy": 1.8281921789050102,
      "epoch": 0.2210857598544281,
      "grad_norm": 9.828446388244629,
      "learning_rate": 5.3804096006400844e-06,
      "loss": 0.4102,
      "mean_token_accuracy": 0.8533445596694946,
      "num_tokens": 85780812.0,
      "step": 71320
    },
    {
      "entropy": 1.8348768278956413,
      "epoch": 0.2211167589794778,
      "grad_norm": 8.210378646850586,
      "learning_rate": 5.380032432995013e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8492519795894623,
      "num_tokens": 85794004.0,
      "step": 71330
    },
    {
      "entropy": 1.8405587255954743,
      "epoch": 0.2211477581045275,
      "grad_norm": 8.21066951751709,
      "learning_rate": 5.379655344657373e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8470315858721733,
      "num_tokens": 85807268.0,
      "step": 71340
    },
    {
      "entropy": 1.908400295674801,
      "epoch": 0.2211787572295772,
      "grad_norm": 8.391548156738281,
      "learning_rate": 5.379278335599377e-06,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8420346319675446,
      "num_tokens": 85819177.0,
      "step": 71350
    },
    {
      "entropy": 1.929537844657898,
      "epoch": 0.2212097563546269,
      "grad_norm": 9.014862060546875,
      "learning_rate": 5.378901405793249e-06,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8366276562213898,
      "num_tokens": 85830729.0,
      "step": 71360
    },
    {
      "entropy": 1.9423951536417008,
      "epoch": 0.2212407554796766,
      "grad_norm": 7.291965484619141,
      "learning_rate": 5.378524555211225e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8441259831190109,
      "num_tokens": 85841037.0,
      "step": 71370
    },
    {
      "entropy": 1.88891644179821,
      "epoch": 0.22127175460472628,
      "grad_norm": 8.193086624145508,
      "learning_rate": 5.378147783825558e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8404372721910477,
      "num_tokens": 85852678.0,
      "step": 71380
    },
    {
      "entropy": 1.911217801272869,
      "epoch": 0.22130275372977598,
      "grad_norm": 6.810231685638428,
      "learning_rate": 5.3777710916085125e-06,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8454770565032959,
      "num_tokens": 85864412.0,
      "step": 71390
    },
    {
      "entropy": 1.9492921188473702,
      "epoch": 0.22133375285482568,
      "grad_norm": 9.480756759643555,
      "learning_rate": 5.377394478532367e-06,
      "loss": 0.5085,
      "mean_token_accuracy": 0.843833090364933,
      "num_tokens": 85875825.0,
      "step": 71400
    },
    {
      "entropy": 1.8765543788671493,
      "epoch": 0.22136475197987537,
      "grad_norm": 10.186013221740723,
      "learning_rate": 5.377017944569414e-06,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8491380795836448,
      "num_tokens": 85887745.0,
      "step": 71410
    },
    {
      "entropy": 1.9640293627977372,
      "epoch": 0.22139575110492507,
      "grad_norm": 4.53644323348999,
      "learning_rate": 5.376641489691959e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8418907791376113,
      "num_tokens": 85898854.0,
      "step": 71420
    },
    {
      "entropy": 1.9712566941976548,
      "epoch": 0.22142675022997477,
      "grad_norm": 8.63244915008545,
      "learning_rate": 5.37626511387232e-06,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8289510354399681,
      "num_tokens": 85909616.0,
      "step": 71430
    },
    {
      "entropy": 1.9296884000301362,
      "epoch": 0.22145774935502446,
      "grad_norm": 8.477029800415039,
      "learning_rate": 5.375888817082833e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8430458202958107,
      "num_tokens": 85920633.0,
      "step": 71440
    },
    {
      "entropy": 1.8276248887181281,
      "epoch": 0.22148874848007416,
      "grad_norm": 4.144367694854736,
      "learning_rate": 5.37551259929584e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8453167825937271,
      "num_tokens": 85933483.0,
      "step": 71450
    },
    {
      "entropy": 1.9024877399206161,
      "epoch": 0.22151974760512386,
      "grad_norm": 7.674788475036621,
      "learning_rate": 5.375136460483704e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8434097394347191,
      "num_tokens": 85945186.0,
      "step": 71460
    },
    {
      "entropy": 1.8075610235333444,
      "epoch": 0.22155074673017355,
      "grad_norm": 8.375659942626953,
      "learning_rate": 5.374760400618798e-06,
      "loss": 0.4304,
      "mean_token_accuracy": 0.8462614819407464,
      "num_tokens": 85958094.0,
      "step": 71470
    },
    {
      "entropy": 1.8221330136060714,
      "epoch": 0.22158174585522325,
      "grad_norm": 7.434113025665283,
      "learning_rate": 5.37438441967351e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8467920750379563,
      "num_tokens": 85971956.0,
      "step": 71480
    },
    {
      "entropy": 1.9221985384821891,
      "epoch": 0.22161274498027292,
      "grad_norm": 8.287166595458984,
      "learning_rate": 5.374008517620237e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8432189226150513,
      "num_tokens": 85982856.0,
      "step": 71490
    },
    {
      "entropy": 1.9569891929626464,
      "epoch": 0.22164374410532262,
      "grad_norm": 9.78246021270752,
      "learning_rate": 5.373632694431396e-06,
      "loss": 0.563,
      "mean_token_accuracy": 0.826215885579586,
      "num_tokens": 85994575.0,
      "step": 71500
    },
    {
      "entropy": 1.9260801374912262,
      "epoch": 0.2216747432303723,
      "grad_norm": 7.269822120666504,
      "learning_rate": 5.373256950079414e-06,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8399269118905067,
      "num_tokens": 86006104.0,
      "step": 71510
    },
    {
      "entropy": 1.8692869395017624,
      "epoch": 0.221705742355422,
      "grad_norm": 8.818297386169434,
      "learning_rate": 5.372881284536732e-06,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8455665573477745,
      "num_tokens": 86019217.0,
      "step": 71520
    },
    {
      "entropy": 1.897246205806732,
      "epoch": 0.2217367414804717,
      "grad_norm": 7.545614242553711,
      "learning_rate": 5.372505697775805e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8487001821398735,
      "num_tokens": 86030738.0,
      "step": 71530
    },
    {
      "entropy": 1.922917690873146,
      "epoch": 0.2217677406055214,
      "grad_norm": 9.24394416809082,
      "learning_rate": 5.372130189769099e-06,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8366933539509773,
      "num_tokens": 86042441.0,
      "step": 71540
    },
    {
      "entropy": 1.8682702884078026,
      "epoch": 0.2217987397305711,
      "grad_norm": 8.580945014953613,
      "learning_rate": 5.371754760489097e-06,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8417121097445488,
      "num_tokens": 86054427.0,
      "step": 71550
    },
    {
      "entropy": 1.8668757036328316,
      "epoch": 0.2218297388556208,
      "grad_norm": 10.871688842773438,
      "learning_rate": 5.371379409908294e-06,
      "loss": 0.4733,
      "mean_token_accuracy": 0.8479284450411797,
      "num_tokens": 86066725.0,
      "step": 71560
    },
    {
      "entropy": 1.915563191473484,
      "epoch": 0.2218607379806705,
      "grad_norm": 10.612263679504395,
      "learning_rate": 5.371004137999198e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8447510316967964,
      "num_tokens": 86078349.0,
      "step": 71570
    },
    {
      "entropy": 1.897247090935707,
      "epoch": 0.2218917371057202,
      "grad_norm": 8.9105863571167,
      "learning_rate": 5.370628944734331e-06,
      "loss": 0.5511,
      "mean_token_accuracy": 0.8368135377764702,
      "num_tokens": 86091101.0,
      "step": 71580
    },
    {
      "entropy": 1.6814722761511802,
      "epoch": 0.22192273623076988,
      "grad_norm": 3.993140697479248,
      "learning_rate": 5.370253830086228e-06,
      "loss": 0.33,
      "mean_token_accuracy": 0.8649098068475723,
      "num_tokens": 86105722.0,
      "step": 71590
    },
    {
      "entropy": 1.9403679892420769,
      "epoch": 0.22195373535581958,
      "grad_norm": 8.66745662689209,
      "learning_rate": 5.369878794027438e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8391989499330521,
      "num_tokens": 86116963.0,
      "step": 71600
    },
    {
      "entropy": 1.8736673273146152,
      "epoch": 0.22198473448086928,
      "grad_norm": 9.14969539642334,
      "learning_rate": 5.369503836530523e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8440220966935158,
      "num_tokens": 86129346.0,
      "step": 71610
    },
    {
      "entropy": 1.901363869011402,
      "epoch": 0.22201573360591897,
      "grad_norm": 8.975959777832031,
      "learning_rate": 5.369128957568058e-06,
      "loss": 0.487,
      "mean_token_accuracy": 0.843775762617588,
      "num_tokens": 86141322.0,
      "step": 71620
    },
    {
      "entropy": 1.836611707508564,
      "epoch": 0.22204673273096867,
      "grad_norm": 3.2881882190704346,
      "learning_rate": 5.368754157112632e-06,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8536363363265991,
      "num_tokens": 86154410.0,
      "step": 71630
    },
    {
      "entropy": 1.817706936597824,
      "epoch": 0.22207773185601837,
      "grad_norm": 7.758713722229004,
      "learning_rate": 5.368379435136848e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8535854294896126,
      "num_tokens": 86167987.0,
      "step": 71640
    },
    {
      "entropy": 1.8554053410887719,
      "epoch": 0.22210873098106806,
      "grad_norm": 3.6705055236816406,
      "learning_rate": 5.368004791613321e-06,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8438238322734832,
      "num_tokens": 86180395.0,
      "step": 71650
    },
    {
      "entropy": 1.8957013592123986,
      "epoch": 0.22213973010611776,
      "grad_norm": 4.396717548370361,
      "learning_rate": 5.36763022651468e-06,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8409201994538307,
      "num_tokens": 86192690.0,
      "step": 71660
    },
    {
      "entropy": 1.8508854925632476,
      "epoch": 0.22217072923116746,
      "grad_norm": 4.068726539611816,
      "learning_rate": 5.367255739813568e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8503674864768982,
      "num_tokens": 86205641.0,
      "step": 71670
    },
    {
      "entropy": 1.8594465300440788,
      "epoch": 0.22220172835621715,
      "grad_norm": 8.009425163269043,
      "learning_rate": 5.3668813314826414e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8513711959123611,
      "num_tokens": 86218005.0,
      "step": 71680
    },
    {
      "entropy": 1.9636985063552856,
      "epoch": 0.22223272748126685,
      "grad_norm": 9.654029846191406,
      "learning_rate": 5.366507001494568e-06,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8238719955086709,
      "num_tokens": 86229342.0,
      "step": 71690
    },
    {
      "entropy": 1.8041048809885978,
      "epoch": 0.22226372660631655,
      "grad_norm": 8.672904014587402,
      "learning_rate": 5.3661327498220305e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8550098910927773,
      "num_tokens": 86241797.0,
      "step": 71700
    },
    {
      "entropy": 1.78623516112566,
      "epoch": 0.22229472573136624,
      "grad_norm": 8.161778450012207,
      "learning_rate": 5.365758576437724e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8487135574221611,
      "num_tokens": 86255083.0,
      "step": 71710
    },
    {
      "entropy": 1.904472067952156,
      "epoch": 0.22232572485641594,
      "grad_norm": 9.36921501159668,
      "learning_rate": 5.365384481314359e-06,
      "loss": 0.5776,
      "mean_token_accuracy": 0.8299209102988243,
      "num_tokens": 86266459.0,
      "step": 71720
    },
    {
      "entropy": 1.8936071269214154,
      "epoch": 0.2223567239814656,
      "grad_norm": 8.524149894714355,
      "learning_rate": 5.365010464424658e-06,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8460559591650962,
      "num_tokens": 86278539.0,
      "step": 71730
    },
    {
      "entropy": 1.8948934614658355,
      "epoch": 0.2223877231065153,
      "grad_norm": 8.934950828552246,
      "learning_rate": 5.364636525741356e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.838769344985485,
      "num_tokens": 86290914.0,
      "step": 71740
    },
    {
      "entropy": 1.913631896674633,
      "epoch": 0.222418722231565,
      "grad_norm": 3.6874070167541504,
      "learning_rate": 5.364262665237202e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8442276403307915,
      "num_tokens": 86302745.0,
      "step": 71750
    },
    {
      "entropy": 1.92730543166399,
      "epoch": 0.2224497213566147,
      "grad_norm": 8.071089744567871,
      "learning_rate": 5.363888882884958e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8468718230724335,
      "num_tokens": 86314083.0,
      "step": 71760
    },
    {
      "entropy": 1.9443558216094972,
      "epoch": 0.2224807204816644,
      "grad_norm": 8.098677635192871,
      "learning_rate": 5.363515178657401e-06,
      "loss": 0.5392,
      "mean_token_accuracy": 0.8254296198487282,
      "num_tokens": 86325749.0,
      "step": 71770
    },
    {
      "entropy": 1.8496044874191284,
      "epoch": 0.2225117196067141,
      "grad_norm": 9.303723335266113,
      "learning_rate": 5.36314155252732e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8531778752803802,
      "num_tokens": 86339269.0,
      "step": 71780
    },
    {
      "entropy": 1.8947973191738128,
      "epoch": 0.2225427187317638,
      "grad_norm": 3.389211654663086,
      "learning_rate": 5.362768004467516e-06,
      "loss": 0.4637,
      "mean_token_accuracy": 0.8556076481938362,
      "num_tokens": 86351044.0,
      "step": 71790
    },
    {
      "entropy": 1.9089409783482552,
      "epoch": 0.22257371785681349,
      "grad_norm": 9.990097999572754,
      "learning_rate": 5.362394534450803e-06,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8289933249354362,
      "num_tokens": 86362914.0,
      "step": 71800
    },
    {
      "entropy": 1.8263745561242104,
      "epoch": 0.22260471698186318,
      "grad_norm": 8.100030899047852,
      "learning_rate": 5.362021142450014e-06,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8659976094961166,
      "num_tokens": 86375568.0,
      "step": 71810
    },
    {
      "entropy": 1.9488488882780075,
      "epoch": 0.22263571610691288,
      "grad_norm": 8.042920112609863,
      "learning_rate": 5.361647828437985e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8422571077942849,
      "num_tokens": 86386717.0,
      "step": 71820
    },
    {
      "entropy": 1.9189837276935577,
      "epoch": 0.22266671523196258,
      "grad_norm": 7.569156646728516,
      "learning_rate": 5.361274592387578e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.8239206343889236,
      "num_tokens": 86399499.0,
      "step": 71830
    },
    {
      "entropy": 1.8419446378946305,
      "epoch": 0.22269771435701227,
      "grad_norm": 3.669304609298706,
      "learning_rate": 5.360901434271656e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8487843006849289,
      "num_tokens": 86412252.0,
      "step": 71840
    },
    {
      "entropy": 1.8220521062612534,
      "epoch": 0.22272871348206197,
      "grad_norm": 5.168817043304443,
      "learning_rate": 5.360528354063102e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8524244442582131,
      "num_tokens": 86424935.0,
      "step": 71850
    },
    {
      "entropy": 1.836111642420292,
      "epoch": 0.22275971260711167,
      "grad_norm": 4.996338367462158,
      "learning_rate": 5.360155351734812e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8411974281072616,
      "num_tokens": 86438125.0,
      "step": 71860
    },
    {
      "entropy": 1.985448981821537,
      "epoch": 0.22279071173216136,
      "grad_norm": 8.615861892700195,
      "learning_rate": 5.359782427259694e-06,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8280460953712463,
      "num_tokens": 86448970.0,
      "step": 71870
    },
    {
      "entropy": 1.8285388305783272,
      "epoch": 0.22282171085721106,
      "grad_norm": 4.234899520874023,
      "learning_rate": 5.359409580610668e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8518346786499024,
      "num_tokens": 86461661.0,
      "step": 71880
    },
    {
      "entropy": 2.012462750822306,
      "epoch": 0.22285270998226075,
      "grad_norm": 3.0699379444122314,
      "learning_rate": 5.359036811760669e-06,
      "loss": 0.5702,
      "mean_token_accuracy": 0.8206193700432778,
      "num_tokens": 86473970.0,
      "step": 71890
    },
    {
      "entropy": 1.941874098777771,
      "epoch": 0.22288370910731045,
      "grad_norm": 6.572078227996826,
      "learning_rate": 5.358664120682644e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8426774546504021,
      "num_tokens": 86485521.0,
      "step": 71900
    },
    {
      "entropy": 1.9939334601163865,
      "epoch": 0.22291470823236015,
      "grad_norm": 9.066808700561523,
      "learning_rate": 5.358291507349554e-06,
      "loss": 0.5545,
      "mean_token_accuracy": 0.8358783975243569,
      "num_tokens": 86496329.0,
      "step": 71910
    },
    {
      "entropy": 1.8996268406510353,
      "epoch": 0.22294570735740984,
      "grad_norm": 7.218948841094971,
      "learning_rate": 5.357918971734374e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8539462700486183,
      "num_tokens": 86508655.0,
      "step": 71920
    },
    {
      "entropy": 1.8882664322853089,
      "epoch": 0.22297670648245954,
      "grad_norm": 7.739889621734619,
      "learning_rate": 5.35754651381009e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.8399605572223663,
      "num_tokens": 86521278.0,
      "step": 71930
    },
    {
      "entropy": 1.89051483720541,
      "epoch": 0.22300770560750924,
      "grad_norm": 3.720012903213501,
      "learning_rate": 5.357174133549702e-06,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8391882091760635,
      "num_tokens": 86533583.0,
      "step": 71940
    },
    {
      "entropy": 1.8325127944350244,
      "epoch": 0.22303870473255893,
      "grad_norm": 4.477136611938477,
      "learning_rate": 5.356801830926224e-06,
      "loss": 0.4477,
      "mean_token_accuracy": 0.837264607846737,
      "num_tokens": 86546642.0,
      "step": 71950
    },
    {
      "entropy": 1.916852205991745,
      "epoch": 0.22306970385760863,
      "grad_norm": 8.114299774169922,
      "learning_rate": 5.356429605912681e-06,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8481098964810372,
      "num_tokens": 86558470.0,
      "step": 71960
    },
    {
      "entropy": 1.869056871533394,
      "epoch": 0.22310070298265833,
      "grad_norm": 4.9179182052612305,
      "learning_rate": 5.356057458482115e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8442103147506714,
      "num_tokens": 86571713.0,
      "step": 71970
    },
    {
      "entropy": 1.9630917876958847,
      "epoch": 0.223131702107708,
      "grad_norm": 9.305983543395996,
      "learning_rate": 5.355685388607575e-06,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8500207409262657,
      "num_tokens": 86582632.0,
      "step": 71980
    },
    {
      "entropy": 1.9708934336900712,
      "epoch": 0.2231627012327577,
      "grad_norm": 9.161256790161133,
      "learning_rate": 5.3553133962621305e-06,
      "loss": 0.5558,
      "mean_token_accuracy": 0.8375864192843437,
      "num_tokens": 86593114.0,
      "step": 71990
    },
    {
      "entropy": 1.892447827756405,
      "epoch": 0.2231937003578074,
      "grad_norm": 8.874557495117188,
      "learning_rate": 5.35494148141886e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.8418182253837585,
      "num_tokens": 86605405.0,
      "step": 72000
    },
    {
      "entropy": 1.9689023926854134,
      "epoch": 0.2232246994828571,
      "grad_norm": 8.011334419250488,
      "learning_rate": 5.354569644050853e-06,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8215542823076248,
      "num_tokens": 86616609.0,
      "step": 72010
    },
    {
      "entropy": 1.7857394725084306,
      "epoch": 0.22325569860790678,
      "grad_norm": 8.13255786895752,
      "learning_rate": 5.354197884131216e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8540096297860146,
      "num_tokens": 86630878.0,
      "step": 72020
    },
    {
      "entropy": 1.8428432375192643,
      "epoch": 0.22328669773295648,
      "grad_norm": 10.879768371582031,
      "learning_rate": 5.353826201633068e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8473896354436874,
      "num_tokens": 86643315.0,
      "step": 72030
    },
    {
      "entropy": 1.9004595071077346,
      "epoch": 0.22331769685800618,
      "grad_norm": 10.056571960449219,
      "learning_rate": 5.3534545965295384e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8414341598749161,
      "num_tokens": 86654883.0,
      "step": 72040
    },
    {
      "entropy": 1.8974163249135017,
      "epoch": 0.22334869598305587,
      "grad_norm": 4.734619140625,
      "learning_rate": 5.353083068793772e-06,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8393331229686737,
      "num_tokens": 86666404.0,
      "step": 72050
    },
    {
      "entropy": 1.8827458634972571,
      "epoch": 0.22337969510810557,
      "grad_norm": 7.965644359588623,
      "learning_rate": 5.352711618398927e-06,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8474840223789215,
      "num_tokens": 86678813.0,
      "step": 72060
    },
    {
      "entropy": 1.9129985481500626,
      "epoch": 0.22341069423315527,
      "grad_norm": 3.682452917098999,
      "learning_rate": 5.352340245318172e-06,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8426467835903168,
      "num_tokens": 86691188.0,
      "step": 72070
    },
    {
      "entropy": 1.8033430457115174,
      "epoch": 0.22344169335820496,
      "grad_norm": 8.872450828552246,
      "learning_rate": 5.351968949524691e-06,
      "loss": 0.3794,
      "mean_token_accuracy": 0.8609629422426224,
      "num_tokens": 86704614.0,
      "step": 72080
    },
    {
      "entropy": 1.8621386557817459,
      "epoch": 0.22347269248325466,
      "grad_norm": 3.9304046630859375,
      "learning_rate": 5.351597730991682e-06,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8484826713800431,
      "num_tokens": 86717142.0,
      "step": 72090
    },
    {
      "entropy": 1.978152585029602,
      "epoch": 0.22350369160830436,
      "grad_norm": 8.142024040222168,
      "learning_rate": 5.351226589692352e-06,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8380045786499977,
      "num_tokens": 86727790.0,
      "step": 72100
    },
    {
      "entropy": 1.861580203473568,
      "epoch": 0.22353469073335405,
      "grad_norm": 9.076419830322266,
      "learning_rate": 5.350855525599924e-06,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8400816440582275,
      "num_tokens": 86739827.0,
      "step": 72110
    },
    {
      "entropy": 1.813391050696373,
      "epoch": 0.22356568985840375,
      "grad_norm": 8.701443672180176,
      "learning_rate": 5.350484538687634e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8458169072866439,
      "num_tokens": 86752977.0,
      "step": 72120
    },
    {
      "entropy": 1.9347064226865769,
      "epoch": 0.22359668898345345,
      "grad_norm": 8.15721607208252,
      "learning_rate": 5.350113628928731e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.8386592581868172,
      "num_tokens": 86764531.0,
      "step": 72130
    },
    {
      "entropy": 1.82811646014452,
      "epoch": 0.22362768810850314,
      "grad_norm": 3.8640589714050293,
      "learning_rate": 5.349742796296475e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8612099349498749,
      "num_tokens": 86777148.0,
      "step": 72140
    },
    {
      "entropy": 1.9188909232616425,
      "epoch": 0.22365868723355284,
      "grad_norm": 8.598167419433594,
      "learning_rate": 5.349372040764139e-06,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8361388146877289,
      "num_tokens": 86788330.0,
      "step": 72150
    },
    {
      "entropy": 1.8747940585017204,
      "epoch": 0.22368968635860254,
      "grad_norm": 4.909088134765625,
      "learning_rate": 5.349001362305013e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.836989839375019,
      "num_tokens": 86800353.0,
      "step": 72160
    },
    {
      "entropy": 1.9335097655653954,
      "epoch": 0.22372068548365223,
      "grad_norm": 8.049686431884766,
      "learning_rate": 5.348630760892396e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8366736158728599,
      "num_tokens": 86811632.0,
      "step": 72170
    },
    {
      "entropy": 1.8739148452877998,
      "epoch": 0.22375168460870193,
      "grad_norm": 8.556473731994629,
      "learning_rate": 5.3482602364996015e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8425387993454934,
      "num_tokens": 86824293.0,
      "step": 72180
    },
    {
      "entropy": 1.9024907439947127,
      "epoch": 0.22378268373375163,
      "grad_norm": 8.516105651855469,
      "learning_rate": 5.347889789099956e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.825544822216034,
      "num_tokens": 86836349.0,
      "step": 72190
    },
    {
      "entropy": 1.897923794388771,
      "epoch": 0.22381368285880132,
      "grad_norm": 7.900509357452393,
      "learning_rate": 5.347519418666795e-06,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8481002271175384,
      "num_tokens": 86847118.0,
      "step": 72200
    },
    {
      "entropy": 1.8177719444036484,
      "epoch": 0.22384468198385102,
      "grad_norm": 7.703773498535156,
      "learning_rate": 5.347149125173477e-06,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8518023356795311,
      "num_tokens": 86860044.0,
      "step": 72210
    },
    {
      "entropy": 1.8685875788331032,
      "epoch": 0.22387568110890071,
      "grad_norm": 6.817202091217041,
      "learning_rate": 5.3467789085933605e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8450110226869583,
      "num_tokens": 86872544.0,
      "step": 72220
    },
    {
      "entropy": 1.9101796388626098,
      "epoch": 0.22390668023395038,
      "grad_norm": 8.793787002563477,
      "learning_rate": 5.346408768899827e-06,
      "loss": 0.538,
      "mean_token_accuracy": 0.8359600931406022,
      "num_tokens": 86884268.0,
      "step": 72230
    },
    {
      "entropy": 1.8767978221178054,
      "epoch": 0.22393767935900008,
      "grad_norm": 8.353222846984863,
      "learning_rate": 5.3460387060662665e-06,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8356214210391044,
      "num_tokens": 86895860.0,
      "step": 72240
    },
    {
      "entropy": 1.8598055988550186,
      "epoch": 0.22396867848404978,
      "grad_norm": 8.898566246032715,
      "learning_rate": 5.345668720066082e-06,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8349154770374299,
      "num_tokens": 86907668.0,
      "step": 72250
    },
    {
      "entropy": 1.9466781616210938,
      "epoch": 0.22399967760909947,
      "grad_norm": 8.537947654724121,
      "learning_rate": 5.34529881087269e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8424281865358353,
      "num_tokens": 86919261.0,
      "step": 72260
    },
    {
      "entropy": 1.89094198346138,
      "epoch": 0.22403067673414917,
      "grad_norm": 8.919629096984863,
      "learning_rate": 5.344928978459521e-06,
      "loss": 0.469,
      "mean_token_accuracy": 0.8445121660828591,
      "num_tokens": 86930676.0,
      "step": 72270
    },
    {
      "entropy": 1.896059663593769,
      "epoch": 0.22406167585919887,
      "grad_norm": 8.955682754516602,
      "learning_rate": 5.344559222800014e-06,
      "loss": 0.5143,
      "mean_token_accuracy": 0.8374183923006058,
      "num_tokens": 86941738.0,
      "step": 72280
    },
    {
      "entropy": 1.883140115439892,
      "epoch": 0.22409267498424856,
      "grad_norm": 4.690348148345947,
      "learning_rate": 5.344189543867627e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.844383516907692,
      "num_tokens": 86953896.0,
      "step": 72290
    },
    {
      "entropy": 1.95089001506567,
      "epoch": 0.22412367410929826,
      "grad_norm": 10.894330024719238,
      "learning_rate": 5.3438199416358285e-06,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8346414759755134,
      "num_tokens": 86965358.0,
      "step": 72300
    },
    {
      "entropy": 1.991261574625969,
      "epoch": 0.22415467323434796,
      "grad_norm": 8.711053848266602,
      "learning_rate": 5.343450416078097e-06,
      "loss": 0.5846,
      "mean_token_accuracy": 0.8269345715641976,
      "num_tokens": 86976161.0,
      "step": 72310
    },
    {
      "entropy": 1.9148213535547256,
      "epoch": 0.22418567235939765,
      "grad_norm": 8.75413703918457,
      "learning_rate": 5.343080967167927e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8315785259008408,
      "num_tokens": 86988247.0,
      "step": 72320
    },
    {
      "entropy": 1.967397329211235,
      "epoch": 0.22421667148444735,
      "grad_norm": 8.436080932617188,
      "learning_rate": 5.342711594878823e-06,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8347294554114342,
      "num_tokens": 86999086.0,
      "step": 72330
    },
    {
      "entropy": 2.000737062096596,
      "epoch": 0.22424767060949705,
      "grad_norm": 9.230005264282227,
      "learning_rate": 5.342342299184309e-06,
      "loss": 0.6598,
      "mean_token_accuracy": 0.8300457715988159,
      "num_tokens": 87010382.0,
      "step": 72340
    },
    {
      "entropy": 1.9067621529102325,
      "epoch": 0.22427866973454674,
      "grad_norm": 8.503578186035156,
      "learning_rate": 5.341973080057913e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.831950668990612,
      "num_tokens": 87022828.0,
      "step": 72350
    },
    {
      "entropy": 1.9217410042881966,
      "epoch": 0.22430966885959644,
      "grad_norm": 9.517120361328125,
      "learning_rate": 5.34160393747318e-06,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8437418282032013,
      "num_tokens": 87034285.0,
      "step": 72360
    },
    {
      "entropy": 1.8522989198565483,
      "epoch": 0.22434066798464614,
      "grad_norm": 9.239526748657227,
      "learning_rate": 5.34123487140367e-06,
      "loss": 0.4158,
      "mean_token_accuracy": 0.8460891872644425,
      "num_tokens": 87046754.0,
      "step": 72370
    },
    {
      "entropy": 1.9420243352651596,
      "epoch": 0.22437166710969583,
      "grad_norm": 8.482245445251465,
      "learning_rate": 5.340865881822951e-06,
      "loss": 0.5542,
      "mean_token_accuracy": 0.8409739390015603,
      "num_tokens": 87058047.0,
      "step": 72380
    },
    {
      "entropy": 1.9989713937044145,
      "epoch": 0.22440266623474553,
      "grad_norm": 10.160581588745117,
      "learning_rate": 5.340496968704607e-06,
      "loss": 0.6208,
      "mean_token_accuracy": 0.82189432233572,
      "num_tokens": 87068796.0,
      "step": 72390
    },
    {
      "entropy": 1.8937209725379944,
      "epoch": 0.22443366535979523,
      "grad_norm": 8.602519035339355,
      "learning_rate": 5.340128132022235e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8464359551668167,
      "num_tokens": 87081062.0,
      "step": 72400
    },
    {
      "entropy": 1.923159296810627,
      "epoch": 0.22446466448484492,
      "grad_norm": 3.398913860321045,
      "learning_rate": 5.339759371749443e-06,
      "loss": 0.501,
      "mean_token_accuracy": 0.8402744174003601,
      "num_tokens": 87093053.0,
      "step": 72410
    },
    {
      "entropy": 1.8120801776647568,
      "epoch": 0.22449566360989462,
      "grad_norm": 4.736013412475586,
      "learning_rate": 5.339390687859851e-06,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8469538927078247,
      "num_tokens": 87106652.0,
      "step": 72420
    },
    {
      "entropy": 1.976495975255966,
      "epoch": 0.22452666273494432,
      "grad_norm": 9.108383178710938,
      "learning_rate": 5.339022080327097e-06,
      "loss": 0.5906,
      "mean_token_accuracy": 0.8242412880063057,
      "num_tokens": 87117466.0,
      "step": 72430
    },
    {
      "entropy": 1.8691235825419426,
      "epoch": 0.224557661859994,
      "grad_norm": 3.6879398822784424,
      "learning_rate": 5.338653549124824e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8466471612453461,
      "num_tokens": 87130079.0,
      "step": 72440
    },
    {
      "entropy": 1.872739316523075,
      "epoch": 0.2245886609850437,
      "grad_norm": 8.234794616699219,
      "learning_rate": 5.338285094226693e-06,
      "loss": 0.5574,
      "mean_token_accuracy": 0.8330441653728485,
      "num_tokens": 87143430.0,
      "step": 72450
    },
    {
      "entropy": 1.911289119720459,
      "epoch": 0.2246196601100934,
      "grad_norm": 11.46174430847168,
      "learning_rate": 5.337916715606378e-06,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8353003263473511,
      "num_tokens": 87155144.0,
      "step": 72460
    },
    {
      "entropy": 1.9616554155945778,
      "epoch": 0.22465065923514307,
      "grad_norm": 7.765041351318359,
      "learning_rate": 5.337548413237561e-06,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8335460603237153,
      "num_tokens": 87166518.0,
      "step": 72470
    },
    {
      "entropy": 1.9011431649327277,
      "epoch": 0.22468165836019277,
      "grad_norm": 8.76104736328125,
      "learning_rate": 5.337180187093943e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8393405005335808,
      "num_tokens": 87178008.0,
      "step": 72480
    },
    {
      "entropy": 1.8503070279955864,
      "epoch": 0.22471265748524247,
      "grad_norm": 10.410445213317871,
      "learning_rate": 5.336812037149233e-06,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8424379363656044,
      "num_tokens": 87191011.0,
      "step": 72490
    },
    {
      "entropy": 1.9341685771942139,
      "epoch": 0.22474365661029216,
      "grad_norm": 3.9696474075317383,
      "learning_rate": 5.336443963377155e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8476704210042953,
      "num_tokens": 87202897.0,
      "step": 72500
    },
    {
      "entropy": 1.9511468440294266,
      "epoch": 0.22477465573534186,
      "grad_norm": 9.15610122680664,
      "learning_rate": 5.336075965751444e-06,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8327615946531296,
      "num_tokens": 87213746.0,
      "step": 72510
    },
    {
      "entropy": 1.9269428536295892,
      "epoch": 0.22480565486039156,
      "grad_norm": 9.162718772888184,
      "learning_rate": 5.335708044245848e-06,
      "loss": 0.5052,
      "mean_token_accuracy": 0.840380209684372,
      "num_tokens": 87225670.0,
      "step": 72520
    },
    {
      "entropy": 1.940378698706627,
      "epoch": 0.22483665398544125,
      "grad_norm": 6.751579284667969,
      "learning_rate": 5.335340198834132e-06,
      "loss": 0.532,
      "mean_token_accuracy": 0.8332238659262657,
      "num_tokens": 87236688.0,
      "step": 72530
    },
    {
      "entropy": 1.958859845995903,
      "epoch": 0.22486765311049095,
      "grad_norm": 10.634821891784668,
      "learning_rate": 5.334972429490065e-06,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8343855604529381,
      "num_tokens": 87247914.0,
      "step": 72540
    },
    {
      "entropy": 1.9445181697607041,
      "epoch": 0.22489865223554065,
      "grad_norm": 9.388453483581543,
      "learning_rate": 5.334604736187437e-06,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8429577887058258,
      "num_tokens": 87258915.0,
      "step": 72550
    },
    {
      "entropy": 1.920157741010189,
      "epoch": 0.22492965136059034,
      "grad_norm": 8.916130065917969,
      "learning_rate": 5.334237118900046e-06,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8405413702130318,
      "num_tokens": 87271134.0,
      "step": 72560
    },
    {
      "entropy": 1.9094747498631477,
      "epoch": 0.22496065048564004,
      "grad_norm": 8.803078651428223,
      "learning_rate": 5.333869577601703e-06,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8441318541765213,
      "num_tokens": 87282774.0,
      "step": 72570
    },
    {
      "entropy": 1.8855116859078407,
      "epoch": 0.22499164961068974,
      "grad_norm": 9.416301727294922,
      "learning_rate": 5.333502112266234e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8383870780467987,
      "num_tokens": 87295016.0,
      "step": 72580
    },
    {
      "entropy": 1.8912652000784873,
      "epoch": 0.22502264873573943,
      "grad_norm": 7.569722652435303,
      "learning_rate": 5.333134722867477e-06,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8289127513766289,
      "num_tokens": 87307357.0,
      "step": 72590
    },
    {
      "entropy": 1.8563939124345779,
      "epoch": 0.22505364786078913,
      "grad_norm": 3.901181221008301,
      "learning_rate": 5.332767409379278e-06,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8382441207766533,
      "num_tokens": 87321430.0,
      "step": 72600
    },
    {
      "entropy": 1.7983256116509438,
      "epoch": 0.22508464698583883,
      "grad_norm": 8.528167724609375,
      "learning_rate": 5.332400171775503e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8504438042640686,
      "num_tokens": 87334567.0,
      "step": 72610
    },
    {
      "entropy": 1.8781714573502541,
      "epoch": 0.22511564611088852,
      "grad_norm": 8.67835521697998,
      "learning_rate": 5.332033010030026e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8410458341240883,
      "num_tokens": 87347091.0,
      "step": 72620
    },
    {
      "entropy": 1.869327275454998,
      "epoch": 0.22514664523593822,
      "grad_norm": 7.16377592086792,
      "learning_rate": 5.331665924116734e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8444514736533165,
      "num_tokens": 87359313.0,
      "step": 72630
    },
    {
      "entropy": 1.9245384186506271,
      "epoch": 0.22517764436098792,
      "grad_norm": 8.11259651184082,
      "learning_rate": 5.331298914009525e-06,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8420319616794586,
      "num_tokens": 87370592.0,
      "step": 72640
    },
    {
      "entropy": 1.9539393037557602,
      "epoch": 0.2252086434860376,
      "grad_norm": 8.318745613098145,
      "learning_rate": 5.3309319796823165e-06,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8356306239962578,
      "num_tokens": 87381867.0,
      "step": 72650
    },
    {
      "entropy": 1.860501480102539,
      "epoch": 0.2252396426110873,
      "grad_norm": 11.79776668548584,
      "learning_rate": 5.33056512110903e-06,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8360062450170517,
      "num_tokens": 87394488.0,
      "step": 72660
    },
    {
      "entropy": 1.9126732975244523,
      "epoch": 0.225270641736137,
      "grad_norm": 7.536765098571777,
      "learning_rate": 5.330198338263605e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8378516137599945,
      "num_tokens": 87406482.0,
      "step": 72670
    },
    {
      "entropy": 1.950472255051136,
      "epoch": 0.2253016408611867,
      "grad_norm": 6.861098766326904,
      "learning_rate": 5.329831631119992e-06,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8359086707234382,
      "num_tokens": 87417682.0,
      "step": 72680
    },
    {
      "entropy": 1.8644071131944657,
      "epoch": 0.2253326399862364,
      "grad_norm": 3.9185712337493896,
      "learning_rate": 5.329464999652151e-06,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8424749970436096,
      "num_tokens": 87430239.0,
      "step": 72690
    },
    {
      "entropy": 1.8736916035413742,
      "epoch": 0.2253636391112861,
      "grad_norm": 8.611660957336426,
      "learning_rate": 5.329098443834062e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.857531276345253,
      "num_tokens": 87442834.0,
      "step": 72700
    },
    {
      "entropy": 1.903275479376316,
      "epoch": 0.2253946382363358,
      "grad_norm": 2.624007225036621,
      "learning_rate": 5.328731963639709e-06,
      "loss": 0.492,
      "mean_token_accuracy": 0.8541991651058197,
      "num_tokens": 87453877.0,
      "step": 72710
    },
    {
      "entropy": 1.8286138609051705,
      "epoch": 0.22542563736138546,
      "grad_norm": 2.311256170272827,
      "learning_rate": 5.328365559043095e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8497246339917183,
      "num_tokens": 87466560.0,
      "step": 72720
    },
    {
      "entropy": 1.7663344264030456,
      "epoch": 0.22545663648643516,
      "grad_norm": 9.070472717285156,
      "learning_rate": 5.327999230018231e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8562262952327728,
      "num_tokens": 87480091.0,
      "step": 72730
    },
    {
      "entropy": 1.8812779873609542,
      "epoch": 0.22548763561148485,
      "grad_norm": 10.331746101379395,
      "learning_rate": 5.327632976539146e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8345619887113571,
      "num_tokens": 87492404.0,
      "step": 72740
    },
    {
      "entropy": 1.8396708235144614,
      "epoch": 0.22551863473653455,
      "grad_norm": 4.0618462562561035,
      "learning_rate": 5.327266798579874e-06,
      "loss": 0.414,
      "mean_token_accuracy": 0.8638743400573731,
      "num_tokens": 87505236.0,
      "step": 72750
    },
    {
      "entropy": 1.8074708625674247,
      "epoch": 0.22554963386158425,
      "grad_norm": 4.086684703826904,
      "learning_rate": 5.326900696114468e-06,
      "loss": 0.3903,
      "mean_token_accuracy": 0.8573868215084076,
      "num_tokens": 87518657.0,
      "step": 72760
    },
    {
      "entropy": 1.912179996073246,
      "epoch": 0.22558063298663394,
      "grad_norm": 6.899926662445068,
      "learning_rate": 5.326534669116988e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8395515128970146,
      "num_tokens": 87530348.0,
      "step": 72770
    },
    {
      "entropy": 1.95667557567358,
      "epoch": 0.22561163211168364,
      "grad_norm": 7.5052361488342285,
      "learning_rate": 5.326168717561514e-06,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8295163378119469,
      "num_tokens": 87541743.0,
      "step": 72780
    },
    {
      "entropy": 1.7888675391674043,
      "epoch": 0.22564263123673334,
      "grad_norm": 6.81760311126709,
      "learning_rate": 5.325802841422131e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8517789766192436,
      "num_tokens": 87554783.0,
      "step": 72790
    },
    {
      "entropy": 1.9340469419956208,
      "epoch": 0.22567363036178303,
      "grad_norm": 8.59512996673584,
      "learning_rate": 5.325437040672939e-06,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8468787103891373,
      "num_tokens": 87565222.0,
      "step": 72800
    },
    {
      "entropy": 1.8804991841316223,
      "epoch": 0.22570462948683273,
      "grad_norm": 4.417396545410156,
      "learning_rate": 5.3250713152880525e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8418024495244026,
      "num_tokens": 87577513.0,
      "step": 72810
    },
    {
      "entropy": 1.889168956875801,
      "epoch": 0.22573562861188243,
      "grad_norm": 6.085160732269287,
      "learning_rate": 5.3247056652415955e-06,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8587741032242775,
      "num_tokens": 87589761.0,
      "step": 72820
    },
    {
      "entropy": 1.8290639758110045,
      "epoch": 0.22576662773693212,
      "grad_norm": 9.560019493103027,
      "learning_rate": 5.324340090507707e-06,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8536442771553994,
      "num_tokens": 87602453.0,
      "step": 72830
    },
    {
      "entropy": 1.9466052517294883,
      "epoch": 0.22579762686198182,
      "grad_norm": 10.381073951721191,
      "learning_rate": 5.323974591060536e-06,
      "loss": 0.5507,
      "mean_token_accuracy": 0.8318102985620499,
      "num_tokens": 87614036.0,
      "step": 72840
    },
    {
      "entropy": 1.8601842939853668,
      "epoch": 0.22582862598703152,
      "grad_norm": 7.986728668212891,
      "learning_rate": 5.323609166874244e-06,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8537519767880439,
      "num_tokens": 87625934.0,
      "step": 72850
    },
    {
      "entropy": 1.9298621758818626,
      "epoch": 0.2258596251120812,
      "grad_norm": 7.938553810119629,
      "learning_rate": 5.3232438179230086e-06,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8388584434986115,
      "num_tokens": 87637522.0,
      "step": 72860
    },
    {
      "entropy": 1.8683239459991454,
      "epoch": 0.2258906242371309,
      "grad_norm": 7.229837894439697,
      "learning_rate": 5.322878544181015e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8578355640172959,
      "num_tokens": 87649965.0,
      "step": 72870
    },
    {
      "entropy": 1.9087668746709823,
      "epoch": 0.2259216233621806,
      "grad_norm": 8.52033519744873,
      "learning_rate": 5.322513345622464e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8454820603132248,
      "num_tokens": 87662212.0,
      "step": 72880
    },
    {
      "entropy": 1.8324894472956657,
      "epoch": 0.2259526224872303,
      "grad_norm": 9.297342300415039,
      "learning_rate": 5.322148222221568e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8492673173546791,
      "num_tokens": 87675147.0,
      "step": 72890
    },
    {
      "entropy": 1.8852784946560859,
      "epoch": 0.22598362161228,
      "grad_norm": 3.968653440475464,
      "learning_rate": 5.3217831739525515e-06,
      "loss": 0.53,
      "mean_token_accuracy": 0.8285475313663483,
      "num_tokens": 87686898.0,
      "step": 72900
    },
    {
      "entropy": 1.9112021766602993,
      "epoch": 0.2260146207373297,
      "grad_norm": 4.359899044036865,
      "learning_rate": 5.321418200789649e-06,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8378346741199494,
      "num_tokens": 87699547.0,
      "step": 72910
    },
    {
      "entropy": 1.9921528965234756,
      "epoch": 0.2260456198623794,
      "grad_norm": 8.43260383605957,
      "learning_rate": 5.321053302707114e-06,
      "loss": 0.5883,
      "mean_token_accuracy": 0.8277798056602478,
      "num_tokens": 87710308.0,
      "step": 72920
    },
    {
      "entropy": 1.847052039206028,
      "epoch": 0.2260766189874291,
      "grad_norm": 8.61585521697998,
      "learning_rate": 5.320688479679204e-06,
      "loss": 0.4228,
      "mean_token_accuracy": 0.855183681845665,
      "num_tokens": 87723526.0,
      "step": 72930
    },
    {
      "entropy": 1.9104703694581986,
      "epoch": 0.2261076181124788,
      "grad_norm": 9.710681915283203,
      "learning_rate": 5.320323731680197e-06,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8491173967719078,
      "num_tokens": 87735916.0,
      "step": 72940
    },
    {
      "entropy": 1.9582516461610795,
      "epoch": 0.22613861723752848,
      "grad_norm": 7.941116809844971,
      "learning_rate": 5.319959058684375e-06,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8455424636602402,
      "num_tokens": 87747201.0,
      "step": 72950
    },
    {
      "entropy": 1.8938917353749276,
      "epoch": 0.22616961636257818,
      "grad_norm": 8.022253036499023,
      "learning_rate": 5.319594460666041e-06,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8349814653396607,
      "num_tokens": 87758501.0,
      "step": 72960
    },
    {
      "entropy": 1.9582582622766496,
      "epoch": 0.22620061548762785,
      "grad_norm": 8.384016036987305,
      "learning_rate": 5.319229937599502e-06,
      "loss": 0.5318,
      "mean_token_accuracy": 0.8396169364452362,
      "num_tokens": 87769542.0,
      "step": 72970
    },
    {
      "entropy": 1.9805827289819717,
      "epoch": 0.22623161461267755,
      "grad_norm": 7.869809150695801,
      "learning_rate": 5.318865489459086e-06,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8493814244866371,
      "num_tokens": 87779962.0,
      "step": 72980
    },
    {
      "entropy": 1.9055181667208672,
      "epoch": 0.22626261373772724,
      "grad_norm": 8.316291809082031,
      "learning_rate": 5.3185011162191226e-06,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8454388037323952,
      "num_tokens": 87791102.0,
      "step": 72990
    },
    {
      "entropy": 1.9191015899181365,
      "epoch": 0.22629361286277694,
      "grad_norm": 7.545352458953857,
      "learning_rate": 5.318136817853964e-06,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8402454987168312,
      "num_tokens": 87803046.0,
      "step": 73000
    },
    {
      "entropy": 1.9037675946950912,
      "epoch": 0.22632461198782663,
      "grad_norm": 9.174158096313477,
      "learning_rate": 5.317772594337969e-06,
      "loss": 0.5138,
      "mean_token_accuracy": 0.8370854437351227,
      "num_tokens": 87814781.0,
      "step": 73010
    },
    {
      "entropy": 1.941840186715126,
      "epoch": 0.22635561111287633,
      "grad_norm": 11.267866134643555,
      "learning_rate": 5.317408445645512e-06,
      "loss": 0.5302,
      "mean_token_accuracy": 0.834077525138855,
      "num_tokens": 87826000.0,
      "step": 73020
    },
    {
      "entropy": 1.9527758836746216,
      "epoch": 0.22638661023792603,
      "grad_norm": 7.3283514976501465,
      "learning_rate": 5.3170443717509745e-06,
      "loss": 0.5012,
      "mean_token_accuracy": 0.8412612110376358,
      "num_tokens": 87836886.0,
      "step": 73030
    },
    {
      "entropy": 1.8808117777109146,
      "epoch": 0.22641760936297572,
      "grad_norm": 10.775453567504883,
      "learning_rate": 5.316680372628757e-06,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8417520016431809,
      "num_tokens": 87848489.0,
      "step": 73040
    },
    {
      "entropy": 1.8674452692270278,
      "epoch": 0.22644860848802542,
      "grad_norm": 4.4659223556518555,
      "learning_rate": 5.316316448253266e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8506308153271676,
      "num_tokens": 87860795.0,
      "step": 73050
    },
    {
      "entropy": 1.8244705572724342,
      "epoch": 0.22647960761307512,
      "grad_norm": 8.522405624389648,
      "learning_rate": 5.315952598598925e-06,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8474853068590165,
      "num_tokens": 87873677.0,
      "step": 73060
    },
    {
      "entropy": 1.923226311802864,
      "epoch": 0.22651060673812481,
      "grad_norm": 7.727634906768799,
      "learning_rate": 5.315588823640166e-06,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8321527540683746,
      "num_tokens": 87885403.0,
      "step": 73070
    },
    {
      "entropy": 1.8770409598946571,
      "epoch": 0.2265416058631745,
      "grad_norm": 7.156225681304932,
      "learning_rate": 5.315225123351437e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8440602973103524,
      "num_tokens": 87897387.0,
      "step": 73080
    },
    {
      "entropy": 1.8762115344405175,
      "epoch": 0.2265726049882242,
      "grad_norm": 7.807554721832275,
      "learning_rate": 5.3148614977071956e-06,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8331007972359658,
      "num_tokens": 87910085.0,
      "step": 73090
    },
    {
      "entropy": 1.9102078214287759,
      "epoch": 0.2266036041132739,
      "grad_norm": 7.591177940368652,
      "learning_rate": 5.314497946681913e-06,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8310081690549851,
      "num_tokens": 87922858.0,
      "step": 73100
    },
    {
      "entropy": 1.9448012560606003,
      "epoch": 0.2266346032383236,
      "grad_norm": 8.734160423278809,
      "learning_rate": 5.31413447025007e-06,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8439796343445778,
      "num_tokens": 87933821.0,
      "step": 73110
    },
    {
      "entropy": 1.8482655853033065,
      "epoch": 0.2266656023633733,
      "grad_norm": 6.68172550201416,
      "learning_rate": 5.3137710683861635e-06,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8570908337831498,
      "num_tokens": 87946480.0,
      "step": 73120
    },
    {
      "entropy": 1.9056684881448747,
      "epoch": 0.226696601488423,
      "grad_norm": 7.89918327331543,
      "learning_rate": 5.3134077410647004e-06,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8478031545877457,
      "num_tokens": 87958529.0,
      "step": 73130
    },
    {
      "entropy": 1.8404195442795754,
      "epoch": 0.2267276006134727,
      "grad_norm": 8.014739990234375,
      "learning_rate": 5.3130444882601994e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8541824370622635,
      "num_tokens": 87971303.0,
      "step": 73140
    },
    {
      "entropy": 1.8912012234330178,
      "epoch": 0.2267585997385224,
      "grad_norm": 9.917512893676758,
      "learning_rate": 5.312681309947193e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8446354269981384,
      "num_tokens": 87983955.0,
      "step": 73150
    },
    {
      "entropy": 1.9031813889741898,
      "epoch": 0.22678959886357208,
      "grad_norm": 8.86811637878418,
      "learning_rate": 5.3123182061002245e-06,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8356417074799538,
      "num_tokens": 87995437.0,
      "step": 73160
    },
    {
      "entropy": 1.9914841502904892,
      "epoch": 0.22682059798862178,
      "grad_norm": 8.847204208374023,
      "learning_rate": 5.311955176693849e-06,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8395547032356262,
      "num_tokens": 88006625.0,
      "step": 73170
    },
    {
      "entropy": 1.9578437075018882,
      "epoch": 0.22685159711367148,
      "grad_norm": 8.657424926757812,
      "learning_rate": 5.311592221702637e-06,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8370061025023461,
      "num_tokens": 88017862.0,
      "step": 73180
    },
    {
      "entropy": 1.843288530409336,
      "epoch": 0.22688259623872117,
      "grad_norm": 7.968525409698486,
      "learning_rate": 5.311229341101166e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8541227072477341,
      "num_tokens": 88030463.0,
      "step": 73190
    },
    {
      "entropy": 1.949458932876587,
      "epoch": 0.22691359536377087,
      "grad_norm": 8.945792198181152,
      "learning_rate": 5.3108665348640306e-06,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8297375872731209,
      "num_tokens": 88041754.0,
      "step": 73200
    },
    {
      "entropy": 1.858209890127182,
      "epoch": 0.22694459448882057,
      "grad_norm": 4.584981441497803,
      "learning_rate": 5.3105038029658355e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.849499624967575,
      "num_tokens": 88054627.0,
      "step": 73210
    },
    {
      "entropy": 1.9475027039647101,
      "epoch": 0.22697559361387024,
      "grad_norm": 3.7373814582824707,
      "learning_rate": 5.310141145381194e-06,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8340747594833374,
      "num_tokens": 88066580.0,
      "step": 73220
    },
    {
      "entropy": 1.8866298824548722,
      "epoch": 0.22700659273891993,
      "grad_norm": 8.264451026916504,
      "learning_rate": 5.30977856208474e-06,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8434574559330941,
      "num_tokens": 88078483.0,
      "step": 73230
    },
    {
      "entropy": 1.8932233542203902,
      "epoch": 0.22703759186396963,
      "grad_norm": 12.378700256347656,
      "learning_rate": 5.309416053051112e-06,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8487199395895004,
      "num_tokens": 88090079.0,
      "step": 73240
    },
    {
      "entropy": 1.8719527080655098,
      "epoch": 0.22706859098901933,
      "grad_norm": 9.49503231048584,
      "learning_rate": 5.309053618254963e-06,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8392317876219749,
      "num_tokens": 88103225.0,
      "step": 73250
    },
    {
      "entropy": 1.8768180578947067,
      "epoch": 0.22709959011406902,
      "grad_norm": 4.083436965942383,
      "learning_rate": 5.308691257670956e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8495214492082596,
      "num_tokens": 88115707.0,
      "step": 73260
    },
    {
      "entropy": 1.9521862715482712,
      "epoch": 0.22713058923911872,
      "grad_norm": 4.048558712005615,
      "learning_rate": 5.308328971273773e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8412155851721763,
      "num_tokens": 88126836.0,
      "step": 73270
    },
    {
      "entropy": 1.9520546540617942,
      "epoch": 0.22716158836416842,
      "grad_norm": 4.938279151916504,
      "learning_rate": 5.3079667590381004e-06,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8300433591008186,
      "num_tokens": 88138872.0,
      "step": 73280
    },
    {
      "entropy": 1.8862208157777787,
      "epoch": 0.2271925874892181,
      "grad_norm": 7.787830829620361,
      "learning_rate": 5.3076046209386405e-06,
      "loss": 0.475,
      "mean_token_accuracy": 0.8360537528991699,
      "num_tokens": 88151734.0,
      "step": 73290
    },
    {
      "entropy": 1.9074003919959068,
      "epoch": 0.2272235866142678,
      "grad_norm": 7.785778045654297,
      "learning_rate": 5.307242556950106e-06,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8495361775159835,
      "num_tokens": 88163810.0,
      "step": 73300
    },
    {
      "entropy": 1.9791472971439361,
      "epoch": 0.2272545857393175,
      "grad_norm": 8.446614265441895,
      "learning_rate": 5.306880567047223e-06,
      "loss": 0.6625,
      "mean_token_accuracy": 0.8312054976820946,
      "num_tokens": 88176079.0,
      "step": 73310
    },
    {
      "entropy": 1.8590899035334587,
      "epoch": 0.2272855848643672,
      "grad_norm": 4.120956897735596,
      "learning_rate": 5.306518651204732e-06,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8539949178695678,
      "num_tokens": 88188688.0,
      "step": 73320
    },
    {
      "entropy": 1.92610152810812,
      "epoch": 0.2273165839894169,
      "grad_norm": 8.005793571472168,
      "learning_rate": 5.306156809397379e-06,
      "loss": 0.4976,
      "mean_token_accuracy": 0.8419401630759239,
      "num_tokens": 88200847.0,
      "step": 73330
    },
    {
      "entropy": 1.9598787605762482,
      "epoch": 0.2273475831144666,
      "grad_norm": 8.136838912963867,
      "learning_rate": 5.305795041599927e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8463526502251625,
      "num_tokens": 88212189.0,
      "step": 73340
    },
    {
      "entropy": 1.8861487179994583,
      "epoch": 0.2273785822395163,
      "grad_norm": 8.918937683105469,
      "learning_rate": 5.30543334778715e-06,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8471100255846977,
      "num_tokens": 88223697.0,
      "step": 73350
    },
    {
      "entropy": 2.0112386524677275,
      "epoch": 0.227409581364566,
      "grad_norm": 8.22529411315918,
      "learning_rate": 5.305071727933835e-06,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8350095435976982,
      "num_tokens": 88233981.0,
      "step": 73360
    },
    {
      "entropy": 1.9829117342829705,
      "epoch": 0.22744058048961568,
      "grad_norm": 8.34417724609375,
      "learning_rate": 5.304710182014778e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.8332961440086365,
      "num_tokens": 88245931.0,
      "step": 73370
    },
    {
      "entropy": 1.8644984647631646,
      "epoch": 0.22747157961466538,
      "grad_norm": 9.018733024597168,
      "learning_rate": 5.304348710004791e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8427291125059128,
      "num_tokens": 88258399.0,
      "step": 73380
    },
    {
      "entropy": 1.8523142382502555,
      "epoch": 0.22750257873971508,
      "grad_norm": 7.3071370124816895,
      "learning_rate": 5.303987311878693e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.856279893219471,
      "num_tokens": 88270512.0,
      "step": 73390
    },
    {
      "entropy": 1.948920576274395,
      "epoch": 0.22753357786476477,
      "grad_norm": 8.815826416015625,
      "learning_rate": 5.303625987611321e-06,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8310361623764038,
      "num_tokens": 88282293.0,
      "step": 73400
    },
    {
      "entropy": 1.8547866210341453,
      "epoch": 0.22756457698981447,
      "grad_norm": 4.00172233581543,
      "learning_rate": 5.30326473717752e-06,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8407047167420387,
      "num_tokens": 88295869.0,
      "step": 73410
    },
    {
      "entropy": 1.9882050842046737,
      "epoch": 0.22759557611486417,
      "grad_norm": 4.15764045715332,
      "learning_rate": 5.3029035605521485e-06,
      "loss": 0.5662,
      "mean_token_accuracy": 0.824670821428299,
      "num_tokens": 88307629.0,
      "step": 73420
    },
    {
      "entropy": 1.896325621008873,
      "epoch": 0.22762657523991386,
      "grad_norm": 7.764687538146973,
      "learning_rate": 5.302542457710075e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8539949029684066,
      "num_tokens": 88319325.0,
      "step": 73430
    },
    {
      "entropy": 1.9740970045328141,
      "epoch": 0.22765757436496356,
      "grad_norm": 8.697678565979004,
      "learning_rate": 5.302181428626182e-06,
      "loss": 0.5632,
      "mean_token_accuracy": 0.8374375849962234,
      "num_tokens": 88330494.0,
      "step": 73440
    },
    {
      "entropy": 1.9539963483810425,
      "epoch": 0.22768857349001326,
      "grad_norm": 9.064693450927734,
      "learning_rate": 5.301820473275364e-06,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8300345674157142,
      "num_tokens": 88341702.0,
      "step": 73450
    },
    {
      "entropy": 1.8306120559573174,
      "epoch": 0.22771957261506293,
      "grad_norm": 9.045393943786621,
      "learning_rate": 5.301459591632527e-06,
      "loss": 0.4311,
      "mean_token_accuracy": 0.847631786763668,
      "num_tokens": 88354694.0,
      "step": 73460
    },
    {
      "entropy": 1.9496394276618958,
      "epoch": 0.22775057174011262,
      "grad_norm": 9.405258178710938,
      "learning_rate": 5.301098783672588e-06,
      "loss": 0.5447,
      "mean_token_accuracy": 0.837113332748413,
      "num_tokens": 88365566.0,
      "step": 73470
    },
    {
      "entropy": 1.832155992090702,
      "epoch": 0.22778157086516232,
      "grad_norm": 8.784136772155762,
      "learning_rate": 5.300738049370477e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8398911356925964,
      "num_tokens": 88378055.0,
      "step": 73480
    },
    {
      "entropy": 1.8397590324282647,
      "epoch": 0.22781256999021202,
      "grad_norm": 10.566763877868652,
      "learning_rate": 5.3003773887011364e-06,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8380313396453858,
      "num_tokens": 88391505.0,
      "step": 73490
    },
    {
      "entropy": 1.9307547613978386,
      "epoch": 0.2278435691152617,
      "grad_norm": 7.682570934295654,
      "learning_rate": 5.3000168016395195e-06,
      "loss": 0.529,
      "mean_token_accuracy": 0.8304173111915588,
      "num_tokens": 88402732.0,
      "step": 73500
    },
    {
      "entropy": 1.9600440084934234,
      "epoch": 0.2278745682403114,
      "grad_norm": 8.14580249786377,
      "learning_rate": 5.299656288160591e-06,
      "loss": 0.5391,
      "mean_token_accuracy": 0.8324763268232346,
      "num_tokens": 88414140.0,
      "step": 73510
    },
    {
      "entropy": 1.9035349115729332,
      "epoch": 0.2279055673653611,
      "grad_norm": 8.5955171585083,
      "learning_rate": 5.299295848239329e-06,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8473742425441741,
      "num_tokens": 88425821.0,
      "step": 73520
    },
    {
      "entropy": 1.9537647753953933,
      "epoch": 0.2279365664904108,
      "grad_norm": 8.174680709838867,
      "learning_rate": 5.298935481850723e-06,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8415812566876412,
      "num_tokens": 88436784.0,
      "step": 73530
    },
    {
      "entropy": 1.9054030612111093,
      "epoch": 0.2279675656154605,
      "grad_norm": 9.483674049377441,
      "learning_rate": 5.2985751889697746e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8276939764618874,
      "num_tokens": 88448671.0,
      "step": 73540
    },
    {
      "entropy": 1.7816397354006768,
      "epoch": 0.2279985647405102,
      "grad_norm": 9.179033279418945,
      "learning_rate": 5.2982149695714964e-06,
      "loss": 0.3919,
      "mean_token_accuracy": 0.8552224412560463,
      "num_tokens": 88463062.0,
      "step": 73550
    },
    {
      "entropy": 1.926460900902748,
      "epoch": 0.2280295638655599,
      "grad_norm": 8.515421867370605,
      "learning_rate": 5.297854823630913e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8481031745672226,
      "num_tokens": 88474333.0,
      "step": 73560
    },
    {
      "entropy": 1.944996650516987,
      "epoch": 0.2280605629906096,
      "grad_norm": 7.795518398284912,
      "learning_rate": 5.2974947511230635e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.847733362019062,
      "num_tokens": 88486183.0,
      "step": 73570
    },
    {
      "entropy": 1.924484086036682,
      "epoch": 0.22809156211565929,
      "grad_norm": 9.433268547058105,
      "learning_rate": 5.297134752022996e-06,
      "loss": 0.5195,
      "mean_token_accuracy": 0.83953125923872,
      "num_tokens": 88498763.0,
      "step": 73580
    },
    {
      "entropy": 1.8909065306186676,
      "epoch": 0.22812256124070898,
      "grad_norm": 4.520410537719727,
      "learning_rate": 5.2967748263057685e-06,
      "loss": 0.4292,
      "mean_token_accuracy": 0.8501924559473991,
      "num_tokens": 88511080.0,
      "step": 73590
    },
    {
      "entropy": 1.855413518846035,
      "epoch": 0.22815356036575868,
      "grad_norm": 7.426053047180176,
      "learning_rate": 5.296414973946457e-06,
      "loss": 0.4606,
      "mean_token_accuracy": 0.848104490339756,
      "num_tokens": 88523301.0,
      "step": 73600
    },
    {
      "entropy": 1.9100253120064736,
      "epoch": 0.22818455949080838,
      "grad_norm": 3.6768534183502197,
      "learning_rate": 5.2960551949201445e-06,
      "loss": 0.5232,
      "mean_token_accuracy": 0.831768749654293,
      "num_tokens": 88535393.0,
      "step": 73610
    },
    {
      "entropy": 1.9303764522075653,
      "epoch": 0.22821555861585807,
      "grad_norm": 9.903493881225586,
      "learning_rate": 5.295695489201927e-06,
      "loss": 0.522,
      "mean_token_accuracy": 0.8348423153162002,
      "num_tokens": 88547298.0,
      "step": 73620
    },
    {
      "entropy": 1.8810430273413659,
      "epoch": 0.22824655774090777,
      "grad_norm": 8.662088394165039,
      "learning_rate": 5.295335856766913e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.836891371011734,
      "num_tokens": 88559818.0,
      "step": 73630
    },
    {
      "entropy": 1.9283401042222976,
      "epoch": 0.22827755686595746,
      "grad_norm": 8.903078079223633,
      "learning_rate": 5.294976297590223e-06,
      "loss": 0.5273,
      "mean_token_accuracy": 0.8383133113384247,
      "num_tokens": 88570970.0,
      "step": 73640
    },
    {
      "entropy": 1.9302942425012588,
      "epoch": 0.22830855599100716,
      "grad_norm": 8.803302764892578,
      "learning_rate": 5.294616811646988e-06,
      "loss": 0.5433,
      "mean_token_accuracy": 0.833921717107296,
      "num_tokens": 88582398.0,
      "step": 73650
    },
    {
      "entropy": 1.8926509320735931,
      "epoch": 0.22833955511605686,
      "grad_norm": 7.454058647155762,
      "learning_rate": 5.294257398912351e-06,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8442865967750549,
      "num_tokens": 88594112.0,
      "step": 73660
    },
    {
      "entropy": 1.811450758576393,
      "epoch": 0.22837055424110655,
      "grad_norm": 3.648226499557495,
      "learning_rate": 5.29389805936147e-06,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8563667595386505,
      "num_tokens": 88606926.0,
      "step": 73670
    },
    {
      "entropy": 1.927273753285408,
      "epoch": 0.22840155336615625,
      "grad_norm": 7.629867076873779,
      "learning_rate": 5.293538792969509e-06,
      "loss": 0.55,
      "mean_token_accuracy": 0.8345969125628472,
      "num_tokens": 88617995.0,
      "step": 73680
    },
    {
      "entropy": 1.866581754386425,
      "epoch": 0.22843255249120595,
      "grad_norm": 10.446778297424316,
      "learning_rate": 5.293179599711649e-06,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8569563254714012,
      "num_tokens": 88630088.0,
      "step": 73690
    },
    {
      "entropy": 1.9038088709115981,
      "epoch": 0.22846355161625564,
      "grad_norm": 8.493614196777344,
      "learning_rate": 5.29282047956308e-06,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8368158757686615,
      "num_tokens": 88642108.0,
      "step": 73700
    },
    {
      "entropy": 1.8956671848893165,
      "epoch": 0.2284945507413053,
      "grad_norm": 8.689042091369629,
      "learning_rate": 5.2924614324990045e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8425026118755341,
      "num_tokens": 88653698.0,
      "step": 73710
    },
    {
      "entropy": 1.8952711433172227,
      "epoch": 0.228525549866355,
      "grad_norm": 8.7789306640625,
      "learning_rate": 5.292102458494637e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8493198037147522,
      "num_tokens": 88665017.0,
      "step": 73720
    },
    {
      "entropy": 1.925877857208252,
      "epoch": 0.2285565489914047,
      "grad_norm": 9.960886001586914,
      "learning_rate": 5.2917435575252045e-06,
      "loss": 0.5461,
      "mean_token_accuracy": 0.833611187338829,
      "num_tokens": 88676741.0,
      "step": 73730
    },
    {
      "entropy": 1.9205459102988243,
      "epoch": 0.2285875481164544,
      "grad_norm": 4.133283615112305,
      "learning_rate": 5.291384729565944e-06,
      "loss": 0.525,
      "mean_token_accuracy": 0.8389394223690033,
      "num_tokens": 88688321.0,
      "step": 73740
    },
    {
      "entropy": 1.974983049929142,
      "epoch": 0.2286185472415041,
      "grad_norm": 8.154329299926758,
      "learning_rate": 5.291025974592104e-06,
      "loss": 0.582,
      "mean_token_accuracy": 0.8283642366528511,
      "num_tokens": 88700262.0,
      "step": 73750
    },
    {
      "entropy": 1.8185023814439774,
      "epoch": 0.2286495463665538,
      "grad_norm": 8.861298561096191,
      "learning_rate": 5.290667292578948e-06,
      "loss": 0.469,
      "mean_token_accuracy": 0.8509408429265022,
      "num_tokens": 88713127.0,
      "step": 73760
    },
    {
      "entropy": 1.8385506071150304,
      "epoch": 0.2286805454916035,
      "grad_norm": 3.4774937629699707,
      "learning_rate": 5.290308683501748e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8416382968425751,
      "num_tokens": 88726534.0,
      "step": 73770
    },
    {
      "entropy": 2.0139817029237745,
      "epoch": 0.2287115446166532,
      "grad_norm": 9.335980415344238,
      "learning_rate": 5.289950147335788e-06,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8362293437123298,
      "num_tokens": 88737523.0,
      "step": 73780
    },
    {
      "entropy": 1.7716790959239006,
      "epoch": 0.2287425437417029,
      "grad_norm": 2.6843504905700684,
      "learning_rate": 5.2895916840563675e-06,
      "loss": 0.3855,
      "mean_token_accuracy": 0.8626002490520477,
      "num_tokens": 88751293.0,
      "step": 73790
    },
    {
      "entropy": 1.9235130712389945,
      "epoch": 0.22877354286675258,
      "grad_norm": 9.159330368041992,
      "learning_rate": 5.289233293638791e-06,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8392347306013107,
      "num_tokens": 88763247.0,
      "step": 73800
    },
    {
      "entropy": 1.8758223891258239,
      "epoch": 0.22880454199180228,
      "grad_norm": 7.979211330413818,
      "learning_rate": 5.288874976058381e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8446521013975143,
      "num_tokens": 88775511.0,
      "step": 73810
    },
    {
      "entropy": 1.9276006370782852,
      "epoch": 0.22883554111685198,
      "grad_norm": 7.672868251800537,
      "learning_rate": 5.288516731290468e-06,
      "loss": 0.514,
      "mean_token_accuracy": 0.8356775805354119,
      "num_tokens": 88786826.0,
      "step": 73820
    },
    {
      "entropy": 1.9099656268954277,
      "epoch": 0.22886654024190167,
      "grad_norm": 9.19827651977539,
      "learning_rate": 5.288158559310397e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8499478042125702,
      "num_tokens": 88799014.0,
      "step": 73830
    },
    {
      "entropy": 1.8794591814279555,
      "epoch": 0.22889753936695137,
      "grad_norm": 9.211488723754883,
      "learning_rate": 5.287800460093521e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8426800593733788,
      "num_tokens": 88811087.0,
      "step": 73840
    },
    {
      "entropy": 1.8956602096557618,
      "epoch": 0.22892853849200107,
      "grad_norm": 8.783346176147461,
      "learning_rate": 5.287442433615207e-06,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8364124745130539,
      "num_tokens": 88823246.0,
      "step": 73850
    },
    {
      "entropy": 1.8410479247570037,
      "epoch": 0.22895953761705076,
      "grad_norm": 7.995346546173096,
      "learning_rate": 5.287084479850834e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8437046140432358,
      "num_tokens": 88836150.0,
      "step": 73860
    },
    {
      "entropy": 1.9268400833010673,
      "epoch": 0.22899053674210046,
      "grad_norm": 6.432771682739258,
      "learning_rate": 5.286726598775794e-06,
      "loss": 0.559,
      "mean_token_accuracy": 0.8304910391569138,
      "num_tokens": 88846935.0,
      "step": 73870
    },
    {
      "entropy": 1.967933678627014,
      "epoch": 0.22902153586715016,
      "grad_norm": 8.738357543945312,
      "learning_rate": 5.286368790365485e-06,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8432664573192596,
      "num_tokens": 88857965.0,
      "step": 73880
    },
    {
      "entropy": 1.8915558218955995,
      "epoch": 0.22905253499219985,
      "grad_norm": 8.925724983215332,
      "learning_rate": 5.286011054595324e-06,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8440207138657569,
      "num_tokens": 88869852.0,
      "step": 73890
    },
    {
      "entropy": 1.9460553407669068,
      "epoch": 0.22908353411724955,
      "grad_norm": 10.097593307495117,
      "learning_rate": 5.285653391440732e-06,
      "loss": 0.5563,
      "mean_token_accuracy": 0.8398781731724739,
      "num_tokens": 88880507.0,
      "step": 73900
    },
    {
      "entropy": 1.9188208684325219,
      "epoch": 0.22911453324229925,
      "grad_norm": 7.534275531768799,
      "learning_rate": 5.285295800877149e-06,
      "loss": 0.4985,
      "mean_token_accuracy": 0.845045380294323,
      "num_tokens": 88891773.0,
      "step": 73910
    },
    {
      "entropy": 1.9317274510860443,
      "epoch": 0.22914553236734894,
      "grad_norm": 7.515310764312744,
      "learning_rate": 5.284938282880022e-06,
      "loss": 0.5103,
      "mean_token_accuracy": 0.830031855404377,
      "num_tokens": 88903628.0,
      "step": 73920
    },
    {
      "entropy": 1.8418959081172943,
      "epoch": 0.22917653149239864,
      "grad_norm": 10.1459379196167,
      "learning_rate": 5.284580837424812e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8359077215194702,
      "num_tokens": 88915956.0,
      "step": 73930
    },
    {
      "entropy": 1.9436316937208176,
      "epoch": 0.22920753061744834,
      "grad_norm": 9.291312217712402,
      "learning_rate": 5.2842234644869895e-06,
      "loss": 0.56,
      "mean_token_accuracy": 0.8290729984641075,
      "num_tokens": 88926709.0,
      "step": 73940
    },
    {
      "entropy": 1.997511848807335,
      "epoch": 0.22923852974249803,
      "grad_norm": 8.855123519897461,
      "learning_rate": 5.283866164042037e-06,
      "loss": 0.5755,
      "mean_token_accuracy": 0.8371048003435135,
      "num_tokens": 88937889.0,
      "step": 73950
    },
    {
      "entropy": 1.9249091997742653,
      "epoch": 0.2292695288675477,
      "grad_norm": 10.297185897827148,
      "learning_rate": 5.283508936065452e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8376726984977723,
      "num_tokens": 88950020.0,
      "step": 73960
    },
    {
      "entropy": 1.974366408586502,
      "epoch": 0.2293005279925974,
      "grad_norm": 9.853002548217773,
      "learning_rate": 5.283151780532737e-06,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8404949456453323,
      "num_tokens": 88960953.0,
      "step": 73970
    },
    {
      "entropy": 1.9446517005562782,
      "epoch": 0.2293315271176471,
      "grad_norm": 7.498370170593262,
      "learning_rate": 5.282794697419412e-06,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8361097246408462,
      "num_tokens": 88972293.0,
      "step": 73980
    },
    {
      "entropy": 1.9057669252157212,
      "epoch": 0.2293625262426968,
      "grad_norm": 9.31442928314209,
      "learning_rate": 5.282437686701009e-06,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8370606362819671,
      "num_tokens": 88983890.0,
      "step": 73990
    },
    {
      "entropy": 1.9226253479719162,
      "epoch": 0.2293935253677465,
      "grad_norm": 8.424675941467285,
      "learning_rate": 5.2820807483530635e-06,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8275117412209511,
      "num_tokens": 88995595.0,
      "step": 74000
    },
    {
      "entropy": 1.9122430935502053,
      "epoch": 0.22942452449279618,
      "grad_norm": 7.728245258331299,
      "learning_rate": 5.281723882351132e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8327098101377487,
      "num_tokens": 89008305.0,
      "step": 74010
    },
    {
      "entropy": 1.8688533097505569,
      "epoch": 0.22945552361784588,
      "grad_norm": 4.136612415313721,
      "learning_rate": 5.281367088670779e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.852810050547123,
      "num_tokens": 89020921.0,
      "step": 74020
    },
    {
      "entropy": 1.8244116827845573,
      "epoch": 0.22948652274289558,
      "grad_norm": 9.678080558776855,
      "learning_rate": 5.281010367287579e-06,
      "loss": 0.3859,
      "mean_token_accuracy": 0.8560571730136871,
      "num_tokens": 89034197.0,
      "step": 74030
    },
    {
      "entropy": 1.9837775856256485,
      "epoch": 0.22951752186794527,
      "grad_norm": 8.24919319152832,
      "learning_rate": 5.280653718177119e-06,
      "loss": 0.519,
      "mean_token_accuracy": 0.83905139118433,
      "num_tokens": 89044845.0,
      "step": 74040
    },
    {
      "entropy": 1.9067578569054604,
      "epoch": 0.22954852099299497,
      "grad_norm": 4.405426502227783,
      "learning_rate": 5.2802971413149995e-06,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8524753168225289,
      "num_tokens": 89056288.0,
      "step": 74050
    },
    {
      "entropy": 1.8309132128953933,
      "epoch": 0.22957952011804467,
      "grad_norm": 7.820222854614258,
      "learning_rate": 5.279940636676828e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8572394847869873,
      "num_tokens": 89069166.0,
      "step": 74060
    },
    {
      "entropy": 1.9012271910905838,
      "epoch": 0.22961051924309436,
      "grad_norm": 9.459297180175781,
      "learning_rate": 5.27958420423823e-06,
      "loss": 0.4733,
      "mean_token_accuracy": 0.8430428206920624,
      "num_tokens": 89081030.0,
      "step": 74070
    },
    {
      "entropy": 1.937754437327385,
      "epoch": 0.22964151836814406,
      "grad_norm": 7.659812927246094,
      "learning_rate": 5.279227843974837e-06,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8409286737442017,
      "num_tokens": 89093165.0,
      "step": 74080
    },
    {
      "entropy": 1.8637029841542243,
      "epoch": 0.22967251749319376,
      "grad_norm": 8.62972354888916,
      "learning_rate": 5.278871555862294e-06,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8332642212510109,
      "num_tokens": 89105778.0,
      "step": 74090
    },
    {
      "entropy": 1.9526731699705124,
      "epoch": 0.22970351661824345,
      "grad_norm": 8.800896644592285,
      "learning_rate": 5.278515339876257e-06,
      "loss": 0.5603,
      "mean_token_accuracy": 0.8348133593797684,
      "num_tokens": 89116443.0,
      "step": 74100
    },
    {
      "entropy": 1.906109546124935,
      "epoch": 0.22973451574329315,
      "grad_norm": 8.518561363220215,
      "learning_rate": 5.278159195992395e-06,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8357308685779572,
      "num_tokens": 89128471.0,
      "step": 74110
    },
    {
      "entropy": 1.9231334283947945,
      "epoch": 0.22976551486834285,
      "grad_norm": 5.9242401123046875,
      "learning_rate": 5.277803124186387e-06,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8310579225420952,
      "num_tokens": 89140663.0,
      "step": 74120
    },
    {
      "entropy": 1.8632835254073143,
      "epoch": 0.22979651399339254,
      "grad_norm": 7.401041507720947,
      "learning_rate": 5.277447124433924e-06,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8411552801728248,
      "num_tokens": 89153122.0,
      "step": 74130
    },
    {
      "entropy": 1.8588541388511657,
      "epoch": 0.22982751311844224,
      "grad_norm": 8.74480152130127,
      "learning_rate": 5.277091196710709e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8416764706373214,
      "num_tokens": 89166007.0,
      "step": 74140
    },
    {
      "entropy": 1.953472825884819,
      "epoch": 0.22985851224349194,
      "grad_norm": 10.276031494140625,
      "learning_rate": 5.276735340992454e-06,
      "loss": 0.5669,
      "mean_token_accuracy": 0.835921137034893,
      "num_tokens": 89177892.0,
      "step": 74150
    },
    {
      "entropy": 1.9886727631092072,
      "epoch": 0.22988951136854163,
      "grad_norm": 8.81214714050293,
      "learning_rate": 5.276379557254886e-06,
      "loss": 0.5502,
      "mean_token_accuracy": 0.8341899603605271,
      "num_tokens": 89188418.0,
      "step": 74160
    },
    {
      "entropy": 1.8839535772800446,
      "epoch": 0.22992051049359133,
      "grad_norm": 7.365606307983398,
      "learning_rate": 5.276023845473741e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8486846655607223,
      "num_tokens": 89201029.0,
      "step": 74170
    },
    {
      "entropy": 1.8606535702943803,
      "epoch": 0.22995150961864103,
      "grad_norm": 6.745984077453613,
      "learning_rate": 5.275668205624769e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8506320536136627,
      "num_tokens": 89213185.0,
      "step": 74180
    },
    {
      "entropy": 1.9073518499732018,
      "epoch": 0.22998250874369072,
      "grad_norm": 8.412069320678711,
      "learning_rate": 5.275312637683727e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.839865280687809,
      "num_tokens": 89225044.0,
      "step": 74190
    },
    {
      "entropy": 1.9035982072353363,
      "epoch": 0.2300135078687404,
      "grad_norm": 9.212803840637207,
      "learning_rate": 5.274957141626388e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8478123605251312,
      "num_tokens": 89236364.0,
      "step": 74200
    },
    {
      "entropy": 1.847705954313278,
      "epoch": 0.2300445069937901,
      "grad_norm": 8.281103134155273,
      "learning_rate": 5.274601717428534e-06,
      "loss": 0.4457,
      "mean_token_accuracy": 0.8482792422175407,
      "num_tokens": 89249228.0,
      "step": 74210
    },
    {
      "entropy": 1.8957882165908813,
      "epoch": 0.23007550611883978,
      "grad_norm": 9.440603256225586,
      "learning_rate": 5.274246365065958e-06,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8471298664808273,
      "num_tokens": 89260514.0,
      "step": 74220
    },
    {
      "entropy": 1.8138180442154408,
      "epoch": 0.23010650524388948,
      "grad_norm": 3.8155839443206787,
      "learning_rate": 5.273891084514467e-06,
      "loss": 0.3897,
      "mean_token_accuracy": 0.8571711733937264,
      "num_tokens": 89274014.0,
      "step": 74230
    },
    {
      "entropy": 1.913990643620491,
      "epoch": 0.23013750436893918,
      "grad_norm": 9.489317893981934,
      "learning_rate": 5.273535875749878e-06,
      "loss": 0.5619,
      "mean_token_accuracy": 0.832446351647377,
      "num_tokens": 89286179.0,
      "step": 74240
    },
    {
      "entropy": 1.921476237475872,
      "epoch": 0.23016850349398887,
      "grad_norm": 4.419705867767334,
      "learning_rate": 5.273180738748017e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8520811811089516,
      "num_tokens": 89298896.0,
      "step": 74250
    },
    {
      "entropy": 1.9407318532466888,
      "epoch": 0.23019950261903857,
      "grad_norm": 9.360230445861816,
      "learning_rate": 5.2728256734847265e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8388498574495316,
      "num_tokens": 89310196.0,
      "step": 74260
    },
    {
      "entropy": 1.822094811499119,
      "epoch": 0.23023050174408827,
      "grad_norm": 9.136000633239746,
      "learning_rate": 5.272470679935853e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8359728991985321,
      "num_tokens": 89323522.0,
      "step": 74270
    },
    {
      "entropy": 1.8379074111580849,
      "epoch": 0.23026150086913796,
      "grad_norm": 8.908660888671875,
      "learning_rate": 5.2721157580772635e-06,
      "loss": 0.4626,
      "mean_token_accuracy": 0.8487562105059624,
      "num_tokens": 89336641.0,
      "step": 74280
    },
    {
      "entropy": 1.9069863885641098,
      "epoch": 0.23029249999418766,
      "grad_norm": 9.512563705444336,
      "learning_rate": 5.27176090788483e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8323235869407654,
      "num_tokens": 89348756.0,
      "step": 74290
    },
    {
      "entropy": 1.9753981336951256,
      "epoch": 0.23032349911923736,
      "grad_norm": 8.980364799499512,
      "learning_rate": 5.271406129334436e-06,
      "loss": 0.5229,
      "mean_token_accuracy": 0.8389256626367569,
      "num_tokens": 89360026.0,
      "step": 74300
    },
    {
      "entropy": 1.9259448662400245,
      "epoch": 0.23035449824428705,
      "grad_norm": 8.680480003356934,
      "learning_rate": 5.271051422401982e-06,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8409480020403862,
      "num_tokens": 89372328.0,
      "step": 74310
    },
    {
      "entropy": 1.9059909775853157,
      "epoch": 0.23038549736933675,
      "grad_norm": 3.84014892578125,
      "learning_rate": 5.2706967870633704e-06,
      "loss": 0.507,
      "mean_token_accuracy": 0.8374297887086868,
      "num_tokens": 89385781.0,
      "step": 74320
    },
    {
      "entropy": 1.9672224968671799,
      "epoch": 0.23041649649438645,
      "grad_norm": 8.767590522766113,
      "learning_rate": 5.270342223294524e-06,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8284627929329872,
      "num_tokens": 89396717.0,
      "step": 74330
    },
    {
      "entropy": 1.9303541094064713,
      "epoch": 0.23044749561943614,
      "grad_norm": 7.687719345092773,
      "learning_rate": 5.2699877310713735e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.846529133617878,
      "num_tokens": 89408365.0,
      "step": 74340
    },
    {
      "entropy": 1.8513358294963838,
      "epoch": 0.23047849474448584,
      "grad_norm": 4.211158275604248,
      "learning_rate": 5.26963331036986e-06,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8417376056313515,
      "num_tokens": 89420728.0,
      "step": 74350
    },
    {
      "entropy": 1.9158269882202148,
      "epoch": 0.23050949386953554,
      "grad_norm": 8.824563026428223,
      "learning_rate": 5.2692789611659345e-06,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8564553380012512,
      "num_tokens": 89432042.0,
      "step": 74360
    },
    {
      "entropy": 1.8483793511986732,
      "epoch": 0.23054049299458523,
      "grad_norm": 8.350893020629883,
      "learning_rate": 5.268924683435563e-06,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8427222266793251,
      "num_tokens": 89445106.0,
      "step": 74370
    },
    {
      "entropy": 1.9533838108181953,
      "epoch": 0.23057149211963493,
      "grad_norm": 7.688497066497803,
      "learning_rate": 5.268570477154723e-06,
      "loss": 0.487,
      "mean_token_accuracy": 0.8480934947729111,
      "num_tokens": 89456517.0,
      "step": 74380
    },
    {
      "entropy": 1.7940612569451333,
      "epoch": 0.23060249124468463,
      "grad_norm": 10.271604537963867,
      "learning_rate": 5.268216342299399e-06,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8567822203040123,
      "num_tokens": 89470177.0,
      "step": 74390
    },
    {
      "entropy": 1.905691820383072,
      "epoch": 0.23063349036973432,
      "grad_norm": 9.242095947265625,
      "learning_rate": 5.267862278845591e-06,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8492756888270379,
      "num_tokens": 89481980.0,
      "step": 74400
    },
    {
      "entropy": 1.9306585028767587,
      "epoch": 0.23066448949478402,
      "grad_norm": 8.073739051818848,
      "learning_rate": 5.267508286769307e-06,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8462639227509499,
      "num_tokens": 89493310.0,
      "step": 74410
    },
    {
      "entropy": 1.9487823605537415,
      "epoch": 0.23069548861983372,
      "grad_norm": 7.810174942016602,
      "learning_rate": 5.267154366046571e-06,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8337769046425819,
      "num_tokens": 89504297.0,
      "step": 74420
    },
    {
      "entropy": 2.020465725660324,
      "epoch": 0.2307264877448834,
      "grad_norm": 8.833086967468262,
      "learning_rate": 5.266800516653412e-06,
      "loss": 0.5802,
      "mean_token_accuracy": 0.8244875445961952,
      "num_tokens": 89515014.0,
      "step": 74430
    },
    {
      "entropy": 1.8529538474977016,
      "epoch": 0.2307574868699331,
      "grad_norm": 8.668907165527344,
      "learning_rate": 5.266446738565875e-06,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8453396156430244,
      "num_tokens": 89527739.0,
      "step": 74440
    },
    {
      "entropy": 1.938523431122303,
      "epoch": 0.23078848599498278,
      "grad_norm": 3.760244846343994,
      "learning_rate": 5.266093031760013e-06,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8392288982868195,
      "num_tokens": 89539449.0,
      "step": 74450
    },
    {
      "entropy": 1.8795533359050751,
      "epoch": 0.23081948512003247,
      "grad_norm": 8.304437637329102,
      "learning_rate": 5.265739396211895e-06,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8489022374153137,
      "num_tokens": 89551459.0,
      "step": 74460
    },
    {
      "entropy": 1.8862609788775444,
      "epoch": 0.23085048424508217,
      "grad_norm": 8.59811782836914,
      "learning_rate": 5.265385831897596e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8412604182958603,
      "num_tokens": 89563656.0,
      "step": 74470
    },
    {
      "entropy": 1.906290727853775,
      "epoch": 0.23088148337013187,
      "grad_norm": 8.959325790405273,
      "learning_rate": 5.2650323387932055e-06,
      "loss": 0.5424,
      "mean_token_accuracy": 0.8369390457868576,
      "num_tokens": 89575960.0,
      "step": 74480
    },
    {
      "entropy": 1.8360799536108972,
      "epoch": 0.23091248249518156,
      "grad_norm": 8.049399375915527,
      "learning_rate": 5.264678916874822e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8528993338346481,
      "num_tokens": 89588989.0,
      "step": 74490
    },
    {
      "entropy": 1.9210985012352466,
      "epoch": 0.23094348162023126,
      "grad_norm": 9.01938533782959,
      "learning_rate": 5.264325566118559e-06,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8337433129549027,
      "num_tokens": 89601183.0,
      "step": 74500
    },
    {
      "entropy": 1.925881953537464,
      "epoch": 0.23097448074528096,
      "grad_norm": 9.546213150024414,
      "learning_rate": 5.263972286500535e-06,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8354784622788429,
      "num_tokens": 89612244.0,
      "step": 74510
    },
    {
      "entropy": 1.85896704941988,
      "epoch": 0.23100547987033065,
      "grad_norm": 9.076498031616211,
      "learning_rate": 5.263619077996888e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8546704173088073,
      "num_tokens": 89623851.0,
      "step": 74520
    },
    {
      "entropy": 1.9326560199260712,
      "epoch": 0.23103647899538035,
      "grad_norm": 3.5278940200805664,
      "learning_rate": 5.263265940583757e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8429828837513924,
      "num_tokens": 89635763.0,
      "step": 74530
    },
    {
      "entropy": 1.8581456407904624,
      "epoch": 0.23106747812043005,
      "grad_norm": 3.4591639041900635,
      "learning_rate": 5.262912874237302e-06,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8445979550480842,
      "num_tokens": 89648396.0,
      "step": 74540
    },
    {
      "entropy": 1.983565354347229,
      "epoch": 0.23109847724547974,
      "grad_norm": 9.090567588806152,
      "learning_rate": 5.262559878933689e-06,
      "loss": 0.54,
      "mean_token_accuracy": 0.8390719383955002,
      "num_tokens": 89659197.0,
      "step": 74550
    },
    {
      "entropy": 1.8371982112526895,
      "epoch": 0.23112947637052944,
      "grad_norm": 8.092364311218262,
      "learning_rate": 5.262206954649097e-06,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8549615368247032,
      "num_tokens": 89671458.0,
      "step": 74560
    },
    {
      "entropy": 1.8131123587489129,
      "epoch": 0.23116047549557914,
      "grad_norm": 4.235384941101074,
      "learning_rate": 5.2618541013597135e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8519744485616684,
      "num_tokens": 89684238.0,
      "step": 74570
    },
    {
      "entropy": 1.8373139381408692,
      "epoch": 0.23119147462062883,
      "grad_norm": 7.887211322784424,
      "learning_rate": 5.26150131904174e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8530314326286316,
      "num_tokens": 89697119.0,
      "step": 74580
    },
    {
      "entropy": 1.9169560462236404,
      "epoch": 0.23122247374567853,
      "grad_norm": 7.602973937988281,
      "learning_rate": 5.261148607671387e-06,
      "loss": 0.493,
      "mean_token_accuracy": 0.8439723640680313,
      "num_tokens": 89708694.0,
      "step": 74590
    },
    {
      "entropy": 1.9264726474881173,
      "epoch": 0.23125347287072823,
      "grad_norm": 8.548664093017578,
      "learning_rate": 5.26079596722488e-06,
      "loss": 0.4944,
      "mean_token_accuracy": 0.844902828335762,
      "num_tokens": 89720007.0,
      "step": 74600
    },
    {
      "entropy": 1.9451578676700592,
      "epoch": 0.23128447199577792,
      "grad_norm": 8.304434776306152,
      "learning_rate": 5.260443397678451e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8561795771121978,
      "num_tokens": 89731316.0,
      "step": 74610
    },
    {
      "entropy": 1.9096089273691177,
      "epoch": 0.23131547112082762,
      "grad_norm": 5.2652459144592285,
      "learning_rate": 5.260090899008346e-06,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8433238387107849,
      "num_tokens": 89742801.0,
      "step": 74620
    },
    {
      "entropy": 1.9577663838863373,
      "epoch": 0.23134647024587732,
      "grad_norm": 9.235634803771973,
      "learning_rate": 5.25973847119082e-06,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8379147708415985,
      "num_tokens": 89754662.0,
      "step": 74630
    },
    {
      "entropy": 1.8390830487012864,
      "epoch": 0.231377469370927,
      "grad_norm": 10.85385799407959,
      "learning_rate": 5.259386114202142e-06,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8555236831307411,
      "num_tokens": 89767357.0,
      "step": 74640
    },
    {
      "entropy": 1.8444815009832383,
      "epoch": 0.2314084684959767,
      "grad_norm": 7.597925186157227,
      "learning_rate": 5.25903382801859e-06,
      "loss": 0.466,
      "mean_token_accuracy": 0.8467239618301392,
      "num_tokens": 89779687.0,
      "step": 74650
    },
    {
      "entropy": 1.8531509324908257,
      "epoch": 0.2314394676210264,
      "grad_norm": 9.575505256652832,
      "learning_rate": 5.2586816126164544e-06,
      "loss": 0.4716,
      "mean_token_accuracy": 0.846321189403534,
      "num_tokens": 89792615.0,
      "step": 74660
    },
    {
      "entropy": 1.8784852564334868,
      "epoch": 0.2314704667460761,
      "grad_norm": 6.865838527679443,
      "learning_rate": 5.258329467972034e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8459220543503762,
      "num_tokens": 89804495.0,
      "step": 74670
    },
    {
      "entropy": 1.8415795341134071,
      "epoch": 0.2315014658711258,
      "grad_norm": 4.12829065322876,
      "learning_rate": 5.257977394061643e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8459180906414986,
      "num_tokens": 89816717.0,
      "step": 74680
    },
    {
      "entropy": 1.8776524886488914,
      "epoch": 0.2315324649961755,
      "grad_norm": 8.387575149536133,
      "learning_rate": 5.257625390861604e-06,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8276556923985481,
      "num_tokens": 89828650.0,
      "step": 74690
    },
    {
      "entropy": 1.8143211975693703,
      "epoch": 0.23156346412122517,
      "grad_norm": 9.098214149475098,
      "learning_rate": 5.25727345834825e-06,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8600962340831757,
      "num_tokens": 89841929.0,
      "step": 74700
    },
    {
      "entropy": 1.9715292781591416,
      "epoch": 0.23159446324627486,
      "grad_norm": 8.993027687072754,
      "learning_rate": 5.256921596497926e-06,
      "loss": 0.5908,
      "mean_token_accuracy": 0.8187263280153274,
      "num_tokens": 89853031.0,
      "step": 74710
    },
    {
      "entropy": 1.8571769908070563,
      "epoch": 0.23162546237132456,
      "grad_norm": 9.328154563903809,
      "learning_rate": 5.256569805286989e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8409965321421623,
      "num_tokens": 89866198.0,
      "step": 74720
    },
    {
      "entropy": 1.930174747109413,
      "epoch": 0.23165646149637426,
      "grad_norm": 8.642340660095215,
      "learning_rate": 5.256218084691808e-06,
      "loss": 0.532,
      "mean_token_accuracy": 0.8321129947900772,
      "num_tokens": 89877322.0,
      "step": 74730
    },
    {
      "entropy": 1.8410791546106338,
      "epoch": 0.23168746062142395,
      "grad_norm": 7.149337291717529,
      "learning_rate": 5.255866434688759e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8385511308908462,
      "num_tokens": 89890110.0,
      "step": 74740
    },
    {
      "entropy": 1.8869677096605302,
      "epoch": 0.23171845974647365,
      "grad_norm": 4.138514518737793,
      "learning_rate": 5.255514855254232e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8496659383177757,
      "num_tokens": 89901498.0,
      "step": 74750
    },
    {
      "entropy": 1.9108739644289017,
      "epoch": 0.23174945887152335,
      "grad_norm": 3.6030426025390625,
      "learning_rate": 5.255163346364628e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8481365412473678,
      "num_tokens": 89913794.0,
      "step": 74760
    },
    {
      "entropy": 1.8742002308368684,
      "epoch": 0.23178045799657304,
      "grad_norm": 4.183920383453369,
      "learning_rate": 5.25481190799636e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.854676042497158,
      "num_tokens": 89926016.0,
      "step": 74770
    },
    {
      "entropy": 1.8252388328313827,
      "epoch": 0.23181145712162274,
      "grad_norm": 8.657533645629883,
      "learning_rate": 5.254460540125848e-06,
      "loss": 0.4536,
      "mean_token_accuracy": 0.855050428211689,
      "num_tokens": 89939164.0,
      "step": 74780
    },
    {
      "entropy": 1.891065989434719,
      "epoch": 0.23184245624667243,
      "grad_norm": 7.831802845001221,
      "learning_rate": 5.254109242729526e-06,
      "loss": 0.5505,
      "mean_token_accuracy": 0.837081053853035,
      "num_tokens": 89950621.0,
      "step": 74790
    },
    {
      "entropy": 1.8497989937663077,
      "epoch": 0.23187345537172213,
      "grad_norm": 9.593158721923828,
      "learning_rate": 5.2537580157838395e-06,
      "loss": 0.487,
      "mean_token_accuracy": 0.8481601446866989,
      "num_tokens": 89962644.0,
      "step": 74800
    },
    {
      "entropy": 1.9117231592535973,
      "epoch": 0.23190445449677183,
      "grad_norm": 8.208306312561035,
      "learning_rate": 5.253406859265246e-06,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8326343983411789,
      "num_tokens": 89973841.0,
      "step": 74810
    },
    {
      "entropy": 1.8912553757429122,
      "epoch": 0.23193545362182152,
      "grad_norm": 6.744791507720947,
      "learning_rate": 5.253055773150209e-06,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8465361073613167,
      "num_tokens": 89985153.0,
      "step": 74820
    },
    {
      "entropy": 1.9291711524128914,
      "epoch": 0.23196645274687122,
      "grad_norm": 8.88097858428955,
      "learning_rate": 5.252704757415207e-06,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8393183097243309,
      "num_tokens": 89996749.0,
      "step": 74830
    },
    {
      "entropy": 1.9174196228384972,
      "epoch": 0.23199745187192092,
      "grad_norm": 8.680536270141602,
      "learning_rate": 5.25235381203673e-06,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8294728398323059,
      "num_tokens": 90008694.0,
      "step": 74840
    },
    {
      "entropy": 1.8946995601058005,
      "epoch": 0.23202845099697061,
      "grad_norm": 7.604076385498047,
      "learning_rate": 5.252002936991277e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.843429696559906,
      "num_tokens": 90021433.0,
      "step": 74850
    },
    {
      "entropy": 1.8540204659104347,
      "epoch": 0.2320594501220203,
      "grad_norm": 4.1174635887146,
      "learning_rate": 5.251652132255359e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.851498831808567,
      "num_tokens": 90032932.0,
      "step": 74860
    },
    {
      "entropy": 1.823809015750885,
      "epoch": 0.23209044924707,
      "grad_norm": 8.341029167175293,
      "learning_rate": 5.251301397805497e-06,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8412299945950508,
      "num_tokens": 90045600.0,
      "step": 74870
    },
    {
      "entropy": 1.804415312409401,
      "epoch": 0.2321214483721197,
      "grad_norm": 9.450553894042969,
      "learning_rate": 5.250950733618225e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8472607269883156,
      "num_tokens": 90059093.0,
      "step": 74880
    },
    {
      "entropy": 1.8313181832432748,
      "epoch": 0.2321524474971694,
      "grad_norm": 8.513847351074219,
      "learning_rate": 5.250600139670086e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8442484304308892,
      "num_tokens": 90071931.0,
      "step": 74890
    },
    {
      "entropy": 1.805875188112259,
      "epoch": 0.2321834466222191,
      "grad_norm": 8.688248634338379,
      "learning_rate": 5.2502496159376335e-06,
      "loss": 0.4062,
      "mean_token_accuracy": 0.8494523644447327,
      "num_tokens": 90085027.0,
      "step": 74900
    },
    {
      "entropy": 1.8620700597763062,
      "epoch": 0.2322144457472688,
      "grad_norm": 7.525083065032959,
      "learning_rate": 5.249899162397435e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8424142166972161,
      "num_tokens": 90097799.0,
      "step": 74910
    },
    {
      "entropy": 1.8805269077420235,
      "epoch": 0.2322454448723185,
      "grad_norm": 9.527615547180176,
      "learning_rate": 5.249548779026064e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8365740343928337,
      "num_tokens": 90110343.0,
      "step": 74920
    },
    {
      "entropy": 1.88951036632061,
      "epoch": 0.2322764439973682,
      "grad_norm": 8.741109848022461,
      "learning_rate": 5.249198465800112e-06,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8468944728374481,
      "num_tokens": 90122279.0,
      "step": 74930
    },
    {
      "entropy": 1.921808835864067,
      "epoch": 0.23230744312241786,
      "grad_norm": 7.858125686645508,
      "learning_rate": 5.248848222696175e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8521002262830735,
      "num_tokens": 90133878.0,
      "step": 74940
    },
    {
      "entropy": 1.9586533635854722,
      "epoch": 0.23233844224746755,
      "grad_norm": 7.217672348022461,
      "learning_rate": 5.248498049690861e-06,
      "loss": 0.5898,
      "mean_token_accuracy": 0.8337603956460953,
      "num_tokens": 90144689.0,
      "step": 74950
    },
    {
      "entropy": 1.942648607492447,
      "epoch": 0.23236944137251725,
      "grad_norm": 8.679248809814453,
      "learning_rate": 5.248147946760793e-06,
      "loss": 0.5142,
      "mean_token_accuracy": 0.8394230246543884,
      "num_tokens": 90155720.0,
      "step": 74960
    },
    {
      "entropy": 1.791871838271618,
      "epoch": 0.23240044049756695,
      "grad_norm": 4.113439559936523,
      "learning_rate": 5.247797913882602e-06,
      "loss": 0.3892,
      "mean_token_accuracy": 0.8623890161514283,
      "num_tokens": 90169661.0,
      "step": 74970
    },
    {
      "entropy": 1.8388600483536721,
      "epoch": 0.23243143962261664,
      "grad_norm": 3.5801477432250977,
      "learning_rate": 5.247447951032928e-06,
      "loss": 0.4552,
      "mean_token_accuracy": 0.8412294924259186,
      "num_tokens": 90182346.0,
      "step": 74980
    },
    {
      "entropy": 1.814172099530697,
      "epoch": 0.23246243874766634,
      "grad_norm": 3.8935399055480957,
      "learning_rate": 5.247098058188425e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8537375554442406,
      "num_tokens": 90195133.0,
      "step": 74990
    },
    {
      "entropy": 1.9371330052614213,
      "epoch": 0.23249343787271604,
      "grad_norm": 7.545095443725586,
      "learning_rate": 5.246748235325756e-06,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8489918291568757,
      "num_tokens": 90205935.0,
      "step": 75000
    },
    {
      "entropy": 1.9675442904233933,
      "epoch": 0.23252443699776573,
      "grad_norm": 7.987549304962158,
      "learning_rate": 5.246398482421598e-06,
      "loss": 0.555,
      "mean_token_accuracy": 0.822739377617836,
      "num_tokens": 90216793.0,
      "step": 75010
    },
    {
      "entropy": 1.9193257197737694,
      "epoch": 0.23255543612281543,
      "grad_norm": 8.14727783203125,
      "learning_rate": 5.246048799452634e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8483793914318085,
      "num_tokens": 90227782.0,
      "step": 75020
    },
    {
      "entropy": 1.8773325115442276,
      "epoch": 0.23258643524786513,
      "grad_norm": 8.173250198364258,
      "learning_rate": 5.245699186395562e-06,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8404614195227623,
      "num_tokens": 90239587.0,
      "step": 75030
    },
    {
      "entropy": 1.78584865629673,
      "epoch": 0.23261743437291482,
      "grad_norm": 8.468997955322266,
      "learning_rate": 5.24534964322709e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8459458023309707,
      "num_tokens": 90253213.0,
      "step": 75040
    },
    {
      "entropy": 1.860848817229271,
      "epoch": 0.23264843349796452,
      "grad_norm": 8.477463722229004,
      "learning_rate": 5.245000169923935e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8364793375134468,
      "num_tokens": 90265280.0,
      "step": 75050
    },
    {
      "entropy": 1.9089404866099358,
      "epoch": 0.23267943262301422,
      "grad_norm": 8.954151153564453,
      "learning_rate": 5.244650766462827e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8399009183049202,
      "num_tokens": 90276657.0,
      "step": 75060
    },
    {
      "entropy": 1.9542981505393981,
      "epoch": 0.2327104317480639,
      "grad_norm": 8.272551536560059,
      "learning_rate": 5.244301432820503e-06,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8442383721470833,
      "num_tokens": 90287635.0,
      "step": 75070
    },
    {
      "entropy": 1.9224041730165482,
      "epoch": 0.2327414308731136,
      "grad_norm": 8.821447372436523,
      "learning_rate": 5.243952168973718e-06,
      "loss": 0.6275,
      "mean_token_accuracy": 0.8298459380865097,
      "num_tokens": 90300036.0,
      "step": 75080
    },
    {
      "entropy": 1.9240923762321471,
      "epoch": 0.2327724299981633,
      "grad_norm": 3.3205227851867676,
      "learning_rate": 5.243602974899231e-06,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8405191898345947,
      "num_tokens": 90310850.0,
      "step": 75090
    },
    {
      "entropy": 1.7889164254069327,
      "epoch": 0.232803429123213,
      "grad_norm": 8.739767074584961,
      "learning_rate": 5.243253850573816e-06,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8499681517481804,
      "num_tokens": 90323572.0,
      "step": 75100
    },
    {
      "entropy": 1.928827765583992,
      "epoch": 0.2328344282482627,
      "grad_norm": 6.958417892456055,
      "learning_rate": 5.2429047959742555e-06,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8277982458472252,
      "num_tokens": 90334888.0,
      "step": 75110
    },
    {
      "entropy": 1.8841903924942016,
      "epoch": 0.2328654273733124,
      "grad_norm": 4.459338665008545,
      "learning_rate": 5.242555811077344e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8400102391839027,
      "num_tokens": 90346732.0,
      "step": 75120
    },
    {
      "entropy": 1.8876192346215248,
      "epoch": 0.2328964264983621,
      "grad_norm": 8.372060775756836,
      "learning_rate": 5.242206895859884e-06,
      "loss": 0.536,
      "mean_token_accuracy": 0.8402991250157357,
      "num_tokens": 90358781.0,
      "step": 75130
    },
    {
      "entropy": 1.9489302426576613,
      "epoch": 0.2329274256234118,
      "grad_norm": 7.9244384765625,
      "learning_rate": 5.241858050298695e-06,
      "loss": 0.5783,
      "mean_token_accuracy": 0.8289978623390197,
      "num_tokens": 90369723.0,
      "step": 75140
    },
    {
      "entropy": 1.9526596829295157,
      "epoch": 0.23295842474846148,
      "grad_norm": 9.39733600616455,
      "learning_rate": 5.241509274370601e-06,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8326995015144348,
      "num_tokens": 90381573.0,
      "step": 75150
    },
    {
      "entropy": 1.855046857893467,
      "epoch": 0.23298942387351118,
      "grad_norm": 8.04930305480957,
      "learning_rate": 5.24116056805244e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8553052991628647,
      "num_tokens": 90394574.0,
      "step": 75160
    },
    {
      "entropy": 1.9039323568344115,
      "epoch": 0.23302042299856088,
      "grad_norm": 8.174781799316406,
      "learning_rate": 5.24081193132106e-06,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8432982727885246,
      "num_tokens": 90406242.0,
      "step": 75170
    },
    {
      "entropy": 1.8142014488577842,
      "epoch": 0.23305142212361057,
      "grad_norm": 10.09156608581543,
      "learning_rate": 5.240463364153321e-06,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8498436227440834,
      "num_tokens": 90419691.0,
      "step": 75180
    },
    {
      "entropy": 1.8953629210591316,
      "epoch": 0.23308242124866024,
      "grad_norm": 7.493961334228516,
      "learning_rate": 5.24011486652609e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.8336662292480469,
      "num_tokens": 90432358.0,
      "step": 75190
    },
    {
      "entropy": 1.8964366748929025,
      "epoch": 0.23311342037370994,
      "grad_norm": 8.871479034423828,
      "learning_rate": 5.2397664384162504e-06,
      "loss": 0.496,
      "mean_token_accuracy": 0.8406847149133683,
      "num_tokens": 90443925.0,
      "step": 75200
    },
    {
      "entropy": 1.8410288974642754,
      "epoch": 0.23314441949875964,
      "grad_norm": 9.23766803741455,
      "learning_rate": 5.239418079800691e-06,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8419799253344535,
      "num_tokens": 90456444.0,
      "step": 75210
    },
    {
      "entropy": 1.9168146222829818,
      "epoch": 0.23317541862380933,
      "grad_norm": 8.451478004455566,
      "learning_rate": 5.239069790656316e-06,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8360653147101402,
      "num_tokens": 90467399.0,
      "step": 75220
    },
    {
      "entropy": 1.7803557753562926,
      "epoch": 0.23320641774885903,
      "grad_norm": 4.116580963134766,
      "learning_rate": 5.238721570960036e-06,
      "loss": 0.4092,
      "mean_token_accuracy": 0.853205819427967,
      "num_tokens": 90479796.0,
      "step": 75230
    },
    {
      "entropy": 1.8654976204037665,
      "epoch": 0.23323741687390873,
      "grad_norm": 9.181920051574707,
      "learning_rate": 5.238373420688775e-06,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8490437477827072,
      "num_tokens": 90491819.0,
      "step": 75240
    },
    {
      "entropy": 1.9017037138342858,
      "epoch": 0.23326841599895842,
      "grad_norm": 7.338005542755127,
      "learning_rate": 5.238025339819467e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.838523656129837,
      "num_tokens": 90503606.0,
      "step": 75250
    },
    {
      "entropy": 1.8169363185763359,
      "epoch": 0.23329941512400812,
      "grad_norm": 11.015266418457031,
      "learning_rate": 5.237677328329057e-06,
      "loss": 0.423,
      "mean_token_accuracy": 0.8545296162366867,
      "num_tokens": 90516076.0,
      "step": 75260
    },
    {
      "entropy": 1.8559274673461914,
      "epoch": 0.23333041424905782,
      "grad_norm": 7.690032482147217,
      "learning_rate": 5.237329386194502e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8428187996149064,
      "num_tokens": 90528110.0,
      "step": 75270
    },
    {
      "entropy": 1.8206583276391028,
      "epoch": 0.2333614133741075,
      "grad_norm": 4.115688323974609,
      "learning_rate": 5.236981513392766e-06,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8493341699242591,
      "num_tokens": 90540476.0,
      "step": 75280
    },
    {
      "entropy": 1.8399672955274582,
      "epoch": 0.2333924124991572,
      "grad_norm": 8.14948558807373,
      "learning_rate": 5.2366337099008265e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.848198390007019,
      "num_tokens": 90552496.0,
      "step": 75290
    },
    {
      "entropy": 1.7906280860304833,
      "epoch": 0.2334234116242069,
      "grad_norm": 6.901556968688965,
      "learning_rate": 5.236285975695673e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8607045665383339,
      "num_tokens": 90565388.0,
      "step": 75300
    },
    {
      "entropy": 1.8817636847496033,
      "epoch": 0.2334544107492566,
      "grad_norm": 10.196159362792969,
      "learning_rate": 5.235938310754303e-06,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8407386288046836,
      "num_tokens": 90577351.0,
      "step": 75310
    },
    {
      "entropy": 1.9196124032139779,
      "epoch": 0.2334854098743063,
      "grad_norm": 7.629261016845703,
      "learning_rate": 5.2355907150537245e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8403740376234055,
      "num_tokens": 90588716.0,
      "step": 75320
    },
    {
      "entropy": 1.9231160968542098,
      "epoch": 0.233516408999356,
      "grad_norm": 8.21264362335205,
      "learning_rate": 5.2352431885709585e-06,
      "loss": 0.5624,
      "mean_token_accuracy": 0.8309063494205475,
      "num_tokens": 90599515.0,
      "step": 75330
    },
    {
      "entropy": 1.8250703021883965,
      "epoch": 0.2335474081244057,
      "grad_norm": 8.739277839660645,
      "learning_rate": 5.234895731283034e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8584131434559822,
      "num_tokens": 90612420.0,
      "step": 75340
    },
    {
      "entropy": 1.832010643184185,
      "epoch": 0.2335784072494554,
      "grad_norm": 8.311301231384277,
      "learning_rate": 5.234548343166994e-06,
      "loss": 0.494,
      "mean_token_accuracy": 0.837817907333374,
      "num_tokens": 90624549.0,
      "step": 75350
    },
    {
      "entropy": 1.9003486275672912,
      "epoch": 0.23360940637450509,
      "grad_norm": 8.429434776306152,
      "learning_rate": 5.234201024199889e-06,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8493073508143425,
      "num_tokens": 90635571.0,
      "step": 75360
    },
    {
      "entropy": 1.8496991708874702,
      "epoch": 0.23364040549955478,
      "grad_norm": 8.533575057983398,
      "learning_rate": 5.233853774358781e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8495924681425094,
      "num_tokens": 90647578.0,
      "step": 75370
    },
    {
      "entropy": 1.9252197608351707,
      "epoch": 0.23367140462460448,
      "grad_norm": 7.957611083984375,
      "learning_rate": 5.233506593620745e-06,
      "loss": 0.5422,
      "mean_token_accuracy": 0.834588311612606,
      "num_tokens": 90658684.0,
      "step": 75380
    },
    {
      "entropy": 1.914403536915779,
      "epoch": 0.23370240374965418,
      "grad_norm": 8.381974220275879,
      "learning_rate": 5.233159481962864e-06,
      "loss": 0.544,
      "mean_token_accuracy": 0.836005237698555,
      "num_tokens": 90670080.0,
      "step": 75390
    },
    {
      "entropy": 1.8845387056469918,
      "epoch": 0.23373340287470387,
      "grad_norm": 7.998114109039307,
      "learning_rate": 5.232812439362231e-06,
      "loss": 0.4658,
      "mean_token_accuracy": 0.8451636284589767,
      "num_tokens": 90682811.0,
      "step": 75400
    },
    {
      "entropy": 1.8718926057219505,
      "epoch": 0.23376440199975357,
      "grad_norm": 3.7296555042266846,
      "learning_rate": 5.232465465795954e-06,
      "loss": 0.569,
      "mean_token_accuracy": 0.8414446488022804,
      "num_tokens": 90694457.0,
      "step": 75410
    },
    {
      "entropy": 1.9601527959108354,
      "epoch": 0.23379540112480326,
      "grad_norm": 8.428439140319824,
      "learning_rate": 5.232118561241144e-06,
      "loss": 0.5494,
      "mean_token_accuracy": 0.8311134546995163,
      "num_tokens": 90705460.0,
      "step": 75420
    },
    {
      "entropy": 1.803976234793663,
      "epoch": 0.23382640024985296,
      "grad_norm": 6.161757946014404,
      "learning_rate": 5.231771725674932e-06,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8395594924688339,
      "num_tokens": 90718488.0,
      "step": 75430
    },
    {
      "entropy": 1.8432627946138382,
      "epoch": 0.23385739937490263,
      "grad_norm": 8.392098426818848,
      "learning_rate": 5.231424959074452e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8638657048344612,
      "num_tokens": 90730557.0,
      "step": 75440
    },
    {
      "entropy": 1.9326068967580796,
      "epoch": 0.23388839849995233,
      "grad_norm": 9.61757755279541,
      "learning_rate": 5.2310782614168524e-06,
      "loss": 0.5731,
      "mean_token_accuracy": 0.828363224864006,
      "num_tokens": 90741224.0,
      "step": 75450
    },
    {
      "entropy": 1.9034574344754218,
      "epoch": 0.23391939762500202,
      "grad_norm": 9.683328628540039,
      "learning_rate": 5.230731632679291e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8382599085569382,
      "num_tokens": 90752495.0,
      "step": 75460
    },
    {
      "entropy": 1.889817087352276,
      "epoch": 0.23395039675005172,
      "grad_norm": 7.614642143249512,
      "learning_rate": 5.230385072838936e-06,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8519553110003472,
      "num_tokens": 90764504.0,
      "step": 75470
    },
    {
      "entropy": 1.846573531627655,
      "epoch": 0.23398139587510142,
      "grad_norm": 8.647238731384277,
      "learning_rate": 5.230038581872968e-06,
      "loss": 0.4794,
      "mean_token_accuracy": 0.845521879196167,
      "num_tokens": 90777223.0,
      "step": 75480
    },
    {
      "entropy": 1.887799009680748,
      "epoch": 0.2340123950001511,
      "grad_norm": 9.934295654296875,
      "learning_rate": 5.2296921597585774e-06,
      "loss": 0.513,
      "mean_token_accuracy": 0.8402909606695175,
      "num_tokens": 90788632.0,
      "step": 75490
    },
    {
      "entropy": 1.8547486677765845,
      "epoch": 0.2340433941252008,
      "grad_norm": 7.495747089385986,
      "learning_rate": 5.229345806472961e-06,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8464630424976349,
      "num_tokens": 90800492.0,
      "step": 75500
    },
    {
      "entropy": 1.8723068907856941,
      "epoch": 0.2340743932502505,
      "grad_norm": 8.363115310668945,
      "learning_rate": 5.228999521993333e-06,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8466327175498009,
      "num_tokens": 90812390.0,
      "step": 75510
    },
    {
      "entropy": 1.8291691452264787,
      "epoch": 0.2341053923753002,
      "grad_norm": 4.456087112426758,
      "learning_rate": 5.228653306296913e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8438918322324753,
      "num_tokens": 90825014.0,
      "step": 75520
    },
    {
      "entropy": 1.9418904185295105,
      "epoch": 0.2341363915003499,
      "grad_norm": 8.06656265258789,
      "learning_rate": 5.228307159360937e-06,
      "loss": 0.5371,
      "mean_token_accuracy": 0.8240366965532303,
      "num_tokens": 90836051.0,
      "step": 75530
    },
    {
      "entropy": 1.8666341871023178,
      "epoch": 0.2341673906253996,
      "grad_norm": 9.48065185546875,
      "learning_rate": 5.2279610811626425e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8425935983657837,
      "num_tokens": 90847835.0,
      "step": 75540
    },
    {
      "entropy": 1.8927869558334351,
      "epoch": 0.2341983897504493,
      "grad_norm": 9.044870376586914,
      "learning_rate": 5.227615071679285e-06,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8357618257403374,
      "num_tokens": 90859514.0,
      "step": 75550
    },
    {
      "entropy": 1.7879324212670327,
      "epoch": 0.234229388875499,
      "grad_norm": 4.232240200042725,
      "learning_rate": 5.227269130888129e-06,
      "loss": 0.368,
      "mean_token_accuracy": 0.8541668400168418,
      "num_tokens": 90873762.0,
      "step": 75560
    },
    {
      "entropy": 1.86094261854887,
      "epoch": 0.23426038800054869,
      "grad_norm": 2.110261917114258,
      "learning_rate": 5.226923258766447e-06,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8436441987752914,
      "num_tokens": 90886426.0,
      "step": 75570
    },
    {
      "entropy": 1.7484543666243553,
      "epoch": 0.23429138712559838,
      "grad_norm": 4.710477352142334,
      "learning_rate": 5.226577455291525e-06,
      "loss": 0.3791,
      "mean_token_accuracy": 0.8541481211781502,
      "num_tokens": 90900094.0,
      "step": 75580
    },
    {
      "entropy": 1.945685862004757,
      "epoch": 0.23432238625064808,
      "grad_norm": 8.194548606872559,
      "learning_rate": 5.226231720440659e-06,
      "loss": 0.5229,
      "mean_token_accuracy": 0.8371816173195838,
      "num_tokens": 90911642.0,
      "step": 75590
    },
    {
      "entropy": 1.8348077073693276,
      "epoch": 0.23435338537569778,
      "grad_norm": 9.100287437438965,
      "learning_rate": 5.2258860541911514e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8539791435003281,
      "num_tokens": 90924678.0,
      "step": 75600
    },
    {
      "entropy": 1.8494689092040062,
      "epoch": 0.23438438450074747,
      "grad_norm": 7.481054782867432,
      "learning_rate": 5.2255404565203234e-06,
      "loss": 0.469,
      "mean_token_accuracy": 0.8396744161844254,
      "num_tokens": 90937460.0,
      "step": 75610
    },
    {
      "entropy": 1.9320793822407722,
      "epoch": 0.23441538362579717,
      "grad_norm": 9.441415786743164,
      "learning_rate": 5.225194927405498e-06,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8343402355909347,
      "num_tokens": 90949563.0,
      "step": 75620
    },
    {
      "entropy": 1.8647829070687294,
      "epoch": 0.23444638275084687,
      "grad_norm": 8.169805526733398,
      "learning_rate": 5.224849466824015e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.839415080845356,
      "num_tokens": 90961677.0,
      "step": 75630
    },
    {
      "entropy": 1.8755572125315667,
      "epoch": 0.23447738187589656,
      "grad_norm": 7.456679344177246,
      "learning_rate": 5.2245040747532205e-06,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8273000553250313,
      "num_tokens": 90974918.0,
      "step": 75640
    },
    {
      "entropy": 1.9732760652899741,
      "epoch": 0.23450838100094626,
      "grad_norm": 9.377166748046875,
      "learning_rate": 5.224158751170473e-06,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8365205094218254,
      "num_tokens": 90986087.0,
      "step": 75650
    },
    {
      "entropy": 1.9110179245471954,
      "epoch": 0.23453938012599596,
      "grad_norm": 8.639516830444336,
      "learning_rate": 5.223813496053142e-06,
      "loss": 0.4806,
      "mean_token_accuracy": 0.8477229610085487,
      "num_tokens": 90997997.0,
      "step": 75660
    },
    {
      "entropy": 1.933686862885952,
      "epoch": 0.23457037925104565,
      "grad_norm": 8.902299880981445,
      "learning_rate": 5.223468309378605e-06,
      "loss": 0.5609,
      "mean_token_accuracy": 0.8341797336935997,
      "num_tokens": 91009272.0,
      "step": 75670
    },
    {
      "entropy": 1.968608994781971,
      "epoch": 0.23460137837609532,
      "grad_norm": 4.5855712890625,
      "learning_rate": 5.2231231911242555e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8370487987995148,
      "num_tokens": 91020661.0,
      "step": 75680
    },
    {
      "entropy": 1.8787442803382874,
      "epoch": 0.23463237750114502,
      "grad_norm": 7.962042331695557,
      "learning_rate": 5.222778141267488e-06,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8470729902386666,
      "num_tokens": 91033230.0,
      "step": 75690
    },
    {
      "entropy": 1.8811139300465585,
      "epoch": 0.23466337662619471,
      "grad_norm": 7.824191093444824,
      "learning_rate": 5.222433159785718e-06,
      "loss": 0.481,
      "mean_token_accuracy": 0.8375716552138328,
      "num_tokens": 91045252.0,
      "step": 75700
    },
    {
      "entropy": 1.9257870644330979,
      "epoch": 0.2346943757512444,
      "grad_norm": 7.708620548248291,
      "learning_rate": 5.222088246656365e-06,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8385186135768891,
      "num_tokens": 91056958.0,
      "step": 75710
    },
    {
      "entropy": 1.9052812889218331,
      "epoch": 0.2347253748762941,
      "grad_norm": 7.920125484466553,
      "learning_rate": 5.22174340185686e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8514411136507988,
      "num_tokens": 91068010.0,
      "step": 75720
    },
    {
      "entropy": 1.8879928544163704,
      "epoch": 0.2347563740013438,
      "grad_norm": 8.99319839477539,
      "learning_rate": 5.221398625364644e-06,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8493816792964936,
      "num_tokens": 91079694.0,
      "step": 75730
    },
    {
      "entropy": 1.860645368695259,
      "epoch": 0.2347873731263935,
      "grad_norm": 8.87778377532959,
      "learning_rate": 5.2210539171571715e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8425444066524506,
      "num_tokens": 91091618.0,
      "step": 75740
    },
    {
      "entropy": 1.8990991547703744,
      "epoch": 0.2348183722514432,
      "grad_norm": 4.461852550506592,
      "learning_rate": 5.220709277211903e-06,
      "loss": 0.519,
      "mean_token_accuracy": 0.83267183303833,
      "num_tokens": 91104008.0,
      "step": 75750
    },
    {
      "entropy": 1.9108994349837303,
      "epoch": 0.2348493713764929,
      "grad_norm": 7.736074924468994,
      "learning_rate": 5.220364705506313e-06,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8414252296090126,
      "num_tokens": 91115759.0,
      "step": 75760
    },
    {
      "entropy": 1.9764820635318756,
      "epoch": 0.2348803705015426,
      "grad_norm": 9.64189624786377,
      "learning_rate": 5.220020202017887e-06,
      "loss": 0.5797,
      "mean_token_accuracy": 0.8334431812167168,
      "num_tokens": 91125911.0,
      "step": 75770
    },
    {
      "entropy": 1.8024020060896873,
      "epoch": 0.2349113696265923,
      "grad_norm": 6.673051357269287,
      "learning_rate": 5.219675766724114e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8567843988537789,
      "num_tokens": 91139395.0,
      "step": 75780
    },
    {
      "entropy": 1.9167088627815247,
      "epoch": 0.23494236875164198,
      "grad_norm": 6.79614782333374,
      "learning_rate": 5.219331399602503e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8450982421636581,
      "num_tokens": 91151124.0,
      "step": 75790
    },
    {
      "entropy": 1.8575886994600297,
      "epoch": 0.23497336787669168,
      "grad_norm": 9.397269248962402,
      "learning_rate": 5.218987100630566e-06,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8560428529977798,
      "num_tokens": 91163168.0,
      "step": 75800
    },
    {
      "entropy": 1.9414154648780824,
      "epoch": 0.23500436700174138,
      "grad_norm": 6.633089065551758,
      "learning_rate": 5.21864286978583e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8463586211204529,
      "num_tokens": 91174212.0,
      "step": 75810
    },
    {
      "entropy": 1.7952500343322755,
      "epoch": 0.23503536612679107,
      "grad_norm": 3.056459426879883,
      "learning_rate": 5.218298707045828e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8485618963837623,
      "num_tokens": 91188049.0,
      "step": 75820
    },
    {
      "entropy": 1.9243362337350844,
      "epoch": 0.23506636525184077,
      "grad_norm": 8.155814170837402,
      "learning_rate": 5.217954612388109e-06,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8437091827392578,
      "num_tokens": 91199119.0,
      "step": 75830
    },
    {
      "entropy": 1.9802058964967728,
      "epoch": 0.23509736437689047,
      "grad_norm": 8.803181648254395,
      "learning_rate": 5.217610585790226e-06,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8368562594056129,
      "num_tokens": 91209740.0,
      "step": 75840
    },
    {
      "entropy": 1.8397732719779014,
      "epoch": 0.23512836350194016,
      "grad_norm": 7.574740886688232,
      "learning_rate": 5.217266627229748e-06,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8465480893850327,
      "num_tokens": 91223580.0,
      "step": 75850
    },
    {
      "entropy": 1.9436632886528968,
      "epoch": 0.23515936262698986,
      "grad_norm": 7.771301746368408,
      "learning_rate": 5.216922736684251e-06,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8337218299508095,
      "num_tokens": 91235380.0,
      "step": 75860
    },
    {
      "entropy": 1.8707226827740668,
      "epoch": 0.23519036175203956,
      "grad_norm": 7.848099231719971,
      "learning_rate": 5.216578914131323e-06,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8528139501810074,
      "num_tokens": 91247966.0,
      "step": 75870
    },
    {
      "entropy": 2.011591988801956,
      "epoch": 0.23522136087708925,
      "grad_norm": 9.857985496520996,
      "learning_rate": 5.216235159548561e-06,
      "loss": 0.6008,
      "mean_token_accuracy": 0.8229044884443283,
      "num_tokens": 91259077.0,
      "step": 75880
    },
    {
      "entropy": 1.9336156010627747,
      "epoch": 0.23525236000213895,
      "grad_norm": 3.4189701080322266,
      "learning_rate": 5.215891472913572e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8479844376444816,
      "num_tokens": 91271077.0,
      "step": 75890
    },
    {
      "entropy": 1.9687816753983498,
      "epoch": 0.23528335912718865,
      "grad_norm": 10.6126070022583,
      "learning_rate": 5.215547854203976e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.839890718460083,
      "num_tokens": 91282469.0,
      "step": 75900
    },
    {
      "entropy": 1.8437663704156875,
      "epoch": 0.23531435825223834,
      "grad_norm": 4.00623083114624,
      "learning_rate": 5.2152043033974e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8508341655135154,
      "num_tokens": 91295928.0,
      "step": 75910
    },
    {
      "entropy": 1.9553633004426956,
      "epoch": 0.23534535737728804,
      "grad_norm": 8.508515357971191,
      "learning_rate": 5.214860820471484e-06,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8452081516385078,
      "num_tokens": 91307197.0,
      "step": 75920
    },
    {
      "entropy": 1.880798263847828,
      "epoch": 0.2353763565023377,
      "grad_norm": 6.8838791847229,
      "learning_rate": 5.214517405403878e-06,
      "loss": 0.415,
      "mean_token_accuracy": 0.8529353275895118,
      "num_tokens": 91319587.0,
      "step": 75930
    },
    {
      "entropy": 1.8892224609851838,
      "epoch": 0.2354073556273874,
      "grad_norm": 8.371787071228027,
      "learning_rate": 5.214174058172241e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8420174464583396,
      "num_tokens": 91332095.0,
      "step": 75940
    },
    {
      "entropy": 1.9798435151576996,
      "epoch": 0.2354383547524371,
      "grad_norm": 10.347806930541992,
      "learning_rate": 5.213830778754241e-06,
      "loss": 0.6038,
      "mean_token_accuracy": 0.8171585813164711,
      "num_tokens": 91342922.0,
      "step": 75950
    },
    {
      "entropy": 1.9197808623313903,
      "epoch": 0.2354693538774868,
      "grad_norm": 8.591217041015625,
      "learning_rate": 5.213487567127559e-06,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8502022624015808,
      "num_tokens": 91354825.0,
      "step": 75960
    },
    {
      "entropy": 1.9522427856922149,
      "epoch": 0.2355003530025365,
      "grad_norm": 9.180562973022461,
      "learning_rate": 5.213144423269887e-06,
      "loss": 0.538,
      "mean_token_accuracy": 0.8390685006976127,
      "num_tokens": 91365900.0,
      "step": 75970
    },
    {
      "entropy": 1.9248853296041488,
      "epoch": 0.2355313521275862,
      "grad_norm": 7.8382248878479,
      "learning_rate": 5.212801347158925e-06,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8395644649863243,
      "num_tokens": 91377884.0,
      "step": 75980
    },
    {
      "entropy": 1.869673204421997,
      "epoch": 0.2355623512526359,
      "grad_norm": 10.19648551940918,
      "learning_rate": 5.212458338772383e-06,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8442730501294136,
      "num_tokens": 91389637.0,
      "step": 75990
    },
    {
      "entropy": 1.8467582926154136,
      "epoch": 0.23559335037768558,
      "grad_norm": 4.066962718963623,
      "learning_rate": 5.212115398087981e-06,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8452418327331543,
      "num_tokens": 91402099.0,
      "step": 76000
    },
    {
      "entropy": 1.8998699069023133,
      "epoch": 0.23562434950273528,
      "grad_norm": 7.766880035400391,
      "learning_rate": 5.211772525083454e-06,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8390342697501183,
      "num_tokens": 91415037.0,
      "step": 76010
    },
    {
      "entropy": 1.8747850939631463,
      "epoch": 0.23565534862778498,
      "grad_norm": 9.178610801696777,
      "learning_rate": 5.2114297197365406e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8318077206611634,
      "num_tokens": 91427375.0,
      "step": 76020
    },
    {
      "entropy": 1.8805608436465264,
      "epoch": 0.23568634775283467,
      "grad_norm": 8.223982810974121,
      "learning_rate": 5.211086982024995e-06,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8412626773118973,
      "num_tokens": 91439742.0,
      "step": 76030
    },
    {
      "entropy": 1.8259792655706406,
      "epoch": 0.23571734687788437,
      "grad_norm": 4.048516273498535,
      "learning_rate": 5.210744311926578e-06,
      "loss": 0.4072,
      "mean_token_accuracy": 0.8551635116338729,
      "num_tokens": 91451962.0,
      "step": 76040
    },
    {
      "entropy": 1.9941584020853043,
      "epoch": 0.23574834600293407,
      "grad_norm": 8.59123420715332,
      "learning_rate": 5.210401709419061e-06,
      "loss": 0.6048,
      "mean_token_accuracy": 0.8241194665431977,
      "num_tokens": 91462894.0,
      "step": 76050
    },
    {
      "entropy": 1.9185423642396926,
      "epoch": 0.23577934512798376,
      "grad_norm": 11.106711387634277,
      "learning_rate": 5.210059174480229e-06,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8400084570050239,
      "num_tokens": 91474440.0,
      "step": 76060
    },
    {
      "entropy": 1.9143582820892333,
      "epoch": 0.23581034425303346,
      "grad_norm": 7.75733757019043,
      "learning_rate": 5.2097167070878754e-06,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8454731091856956,
      "num_tokens": 91485553.0,
      "step": 76070
    },
    {
      "entropy": 1.9490876853466035,
      "epoch": 0.23584134337808316,
      "grad_norm": 11.029200553894043,
      "learning_rate": 5.209374307219801e-06,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8368668779730797,
      "num_tokens": 91495910.0,
      "step": 76080
    },
    {
      "entropy": 1.9713003784418106,
      "epoch": 0.23587234250313285,
      "grad_norm": 8.084342956542969,
      "learning_rate": 5.20903197485382e-06,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8417750880122185,
      "num_tokens": 91506865.0,
      "step": 76090
    },
    {
      "entropy": 1.8941795021295547,
      "epoch": 0.23590334162818255,
      "grad_norm": 7.648937702178955,
      "learning_rate": 5.208689709967756e-06,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8375332474708557,
      "num_tokens": 91518800.0,
      "step": 76100
    },
    {
      "entropy": 1.8584771752357483,
      "epoch": 0.23593434075323225,
      "grad_norm": 8.80497932434082,
      "learning_rate": 5.208347512539442e-06,
      "loss": 0.5252,
      "mean_token_accuracy": 0.8386746376752854,
      "num_tokens": 91530792.0,
      "step": 76110
    },
    {
      "entropy": 1.8739138320088387,
      "epoch": 0.23596533987828194,
      "grad_norm": 8.877378463745117,
      "learning_rate": 5.2080053825467235e-06,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8397540152072906,
      "num_tokens": 91542992.0,
      "step": 76120
    },
    {
      "entropy": 1.840997065603733,
      "epoch": 0.23599633900333164,
      "grad_norm": 7.287625789642334,
      "learning_rate": 5.207663319967453e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8495836406946182,
      "num_tokens": 91555954.0,
      "step": 76130
    },
    {
      "entropy": 1.826866079866886,
      "epoch": 0.23602733812838134,
      "grad_norm": 8.53991985321045,
      "learning_rate": 5.207321324779495e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8605758026242256,
      "num_tokens": 91568711.0,
      "step": 76140
    },
    {
      "entropy": 1.8429628267884255,
      "epoch": 0.23605833725343103,
      "grad_norm": 9.361772537231445,
      "learning_rate": 5.2069793969607265e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8468737691640854,
      "num_tokens": 91581173.0,
      "step": 76150
    },
    {
      "entropy": 1.8500198744237424,
      "epoch": 0.23608933637848073,
      "grad_norm": 8.414274215698242,
      "learning_rate": 5.206637536489028e-06,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8451015651226044,
      "num_tokens": 91594341.0,
      "step": 76160
    },
    {
      "entropy": 1.93178521245718,
      "epoch": 0.23612033550353043,
      "grad_norm": 7.025452136993408,
      "learning_rate": 5.206295743342297e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8408794820308685,
      "num_tokens": 91605890.0,
      "step": 76170
    },
    {
      "entropy": 1.8896970108151436,
      "epoch": 0.2361513346285801,
      "grad_norm": 10.62279224395752,
      "learning_rate": 5.205954017498437e-06,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8361875951290131,
      "num_tokens": 91617830.0,
      "step": 76180
    },
    {
      "entropy": 1.8840731859207154,
      "epoch": 0.2361823337536298,
      "grad_norm": 8.596649169921875,
      "learning_rate": 5.205612358935365e-06,
      "loss": 0.55,
      "mean_token_accuracy": 0.8344793394207954,
      "num_tokens": 91629222.0,
      "step": 76190
    },
    {
      "entropy": 1.8855000153183936,
      "epoch": 0.2362133328786795,
      "grad_norm": 7.801729202270508,
      "learning_rate": 5.205270767631004e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.849211573600769,
      "num_tokens": 91641322.0,
      "step": 76200
    },
    {
      "entropy": 1.9867444038391113,
      "epoch": 0.23624433200372918,
      "grad_norm": 10.91826057434082,
      "learning_rate": 5.2049292435632915e-06,
      "loss": 0.5572,
      "mean_token_accuracy": 0.8317925870418549,
      "num_tokens": 91651690.0,
      "step": 76210
    },
    {
      "entropy": 1.9123092323541642,
      "epoch": 0.23627533112877888,
      "grad_norm": 8.301390647888184,
      "learning_rate": 5.2045877867101715e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8369743451476097,
      "num_tokens": 91662933.0,
      "step": 76220
    },
    {
      "entropy": 1.9738686069846154,
      "epoch": 0.23630633025382858,
      "grad_norm": 9.968904495239258,
      "learning_rate": 5.2042463970496e-06,
      "loss": 0.5797,
      "mean_token_accuracy": 0.8200091898441315,
      "num_tokens": 91674479.0,
      "step": 76230
    },
    {
      "entropy": 1.8900879070162773,
      "epoch": 0.23633732937887827,
      "grad_norm": 10.34799575805664,
      "learning_rate": 5.203905074559543e-06,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8390724554657936,
      "num_tokens": 91686217.0,
      "step": 76240
    },
    {
      "entropy": 1.8866849929094314,
      "epoch": 0.23636832850392797,
      "grad_norm": 8.001737594604492,
      "learning_rate": 5.203563819217977e-06,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8539379611611366,
      "num_tokens": 91697971.0,
      "step": 76250
    },
    {
      "entropy": 1.9403339833021165,
      "epoch": 0.23639932762897767,
      "grad_norm": 8.635519027709961,
      "learning_rate": 5.203222631002886e-06,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8294428750872612,
      "num_tokens": 91709307.0,
      "step": 76260
    },
    {
      "entropy": 1.9339663892984391,
      "epoch": 0.23643032675402736,
      "grad_norm": 8.026914596557617,
      "learning_rate": 5.202881509892268e-06,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8449966043233872,
      "num_tokens": 91720715.0,
      "step": 76270
    },
    {
      "entropy": 1.769917319715023,
      "epoch": 0.23646132587907706,
      "grad_norm": 3.137450933456421,
      "learning_rate": 5.202540455864128e-06,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8480799734592438,
      "num_tokens": 91734185.0,
      "step": 76280
    },
    {
      "entropy": 1.9272101998329163,
      "epoch": 0.23649232500412676,
      "grad_norm": 9.050545692443848,
      "learning_rate": 5.202199468896483e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8370890215039253,
      "num_tokens": 91746051.0,
      "step": 76290
    },
    {
      "entropy": 1.890957424044609,
      "epoch": 0.23652332412917645,
      "grad_norm": 7.916098117828369,
      "learning_rate": 5.201858548967359e-06,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8495743572711945,
      "num_tokens": 91757115.0,
      "step": 76300
    },
    {
      "entropy": 1.904859295487404,
      "epoch": 0.23655432325422615,
      "grad_norm": 7.8038105964660645,
      "learning_rate": 5.201517696054792e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8543992474675178,
      "num_tokens": 91768690.0,
      "step": 76310
    },
    {
      "entropy": 1.8198333993554114,
      "epoch": 0.23658532237927585,
      "grad_norm": 8.247322082519531,
      "learning_rate": 5.2011769101368294e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8562105298042297,
      "num_tokens": 91781845.0,
      "step": 76320
    },
    {
      "entropy": 1.879896378517151,
      "epoch": 0.23661632150432554,
      "grad_norm": 3.9357030391693115,
      "learning_rate": 5.200836191191528e-06,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8404986575245857,
      "num_tokens": 91794225.0,
      "step": 76330
    },
    {
      "entropy": 1.8717378214001656,
      "epoch": 0.23664732062937524,
      "grad_norm": 3.682194948196411,
      "learning_rate": 5.200495539196953e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8454118952155113,
      "num_tokens": 91807385.0,
      "step": 76340
    },
    {
      "entropy": 1.9737626880407333,
      "epoch": 0.23667831975442494,
      "grad_norm": 7.0051069259643555,
      "learning_rate": 5.200154954131182e-06,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8395712092518807,
      "num_tokens": 91818225.0,
      "step": 76350
    },
    {
      "entropy": 1.9173112154006957,
      "epoch": 0.23670931887947463,
      "grad_norm": 7.16079568862915,
      "learning_rate": 5.199814435972302e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8442193806171417,
      "num_tokens": 91830394.0,
      "step": 76360
    },
    {
      "entropy": 1.8628339901566506,
      "epoch": 0.23674031800452433,
      "grad_norm": 4.580597400665283,
      "learning_rate": 5.19947398469841e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8353284701704979,
      "num_tokens": 91843266.0,
      "step": 76370
    },
    {
      "entropy": 1.9519380658864975,
      "epoch": 0.23677131712957403,
      "grad_norm": 7.013167381286621,
      "learning_rate": 5.1991336002876116e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8405268996953964,
      "num_tokens": 91853694.0,
      "step": 76380
    },
    {
      "entropy": 1.8750559836626053,
      "epoch": 0.23680231625462372,
      "grad_norm": 3.834982395172119,
      "learning_rate": 5.198793282718023e-06,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8327495470643044,
      "num_tokens": 91866658.0,
      "step": 76390
    },
    {
      "entropy": 1.778690068423748,
      "epoch": 0.23683331537967342,
      "grad_norm": 8.452054977416992,
      "learning_rate": 5.198453031967774e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8489452451467514,
      "num_tokens": 91880171.0,
      "step": 76400
    },
    {
      "entropy": 1.8419625982642174,
      "epoch": 0.23686431450472312,
      "grad_norm": 5.224298000335693,
      "learning_rate": 5.198112848015e-06,
      "loss": 0.428,
      "mean_token_accuracy": 0.8543521463871002,
      "num_tokens": 91892362.0,
      "step": 76410
    },
    {
      "entropy": 1.8903858274221421,
      "epoch": 0.2368953136297728,
      "grad_norm": 8.055941581726074,
      "learning_rate": 5.197772730837848e-06,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8432309225201606,
      "num_tokens": 91904734.0,
      "step": 76420
    },
    {
      "entropy": 1.8139331057667731,
      "epoch": 0.23692631275482248,
      "grad_norm": 4.4151506423950195,
      "learning_rate": 5.197432680414474e-06,
      "loss": 0.3915,
      "mean_token_accuracy": 0.855991254746914,
      "num_tokens": 91917912.0,
      "step": 76430
    },
    {
      "entropy": 1.7869535863399506,
      "epoch": 0.23695731187987218,
      "grad_norm": 8.555081367492676,
      "learning_rate": 5.1970926967230455e-06,
      "loss": 0.3974,
      "mean_token_accuracy": 0.8637645095586777,
      "num_tokens": 91931141.0,
      "step": 76440
    },
    {
      "entropy": 1.8419126272201538,
      "epoch": 0.23698831100492188,
      "grad_norm": 4.1938910484313965,
      "learning_rate": 5.196752779741738e-06,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8464212834835052,
      "num_tokens": 91943834.0,
      "step": 76450
    },
    {
      "entropy": 1.8724951505661012,
      "epoch": 0.23701931012997157,
      "grad_norm": 9.071846008300781,
      "learning_rate": 5.196412929448742e-06,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8366881161928177,
      "num_tokens": 91956524.0,
      "step": 76460
    },
    {
      "entropy": 1.9315653994679451,
      "epoch": 0.23705030925502127,
      "grad_norm": 4.576693534851074,
      "learning_rate": 5.1960731458222526e-06,
      "loss": 0.5247,
      "mean_token_accuracy": 0.832519143819809,
      "num_tokens": 91968089.0,
      "step": 76470
    },
    {
      "entropy": 1.8646988362073897,
      "epoch": 0.23708130838007097,
      "grad_norm": 9.52219009399414,
      "learning_rate": 5.195733428840475e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8468518033623695,
      "num_tokens": 91979897.0,
      "step": 76480
    },
    {
      "entropy": 1.8760708898305893,
      "epoch": 0.23711230750512066,
      "grad_norm": 8.841971397399902,
      "learning_rate": 5.1953937784816275e-06,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8391115218400955,
      "num_tokens": 91992057.0,
      "step": 76490
    },
    {
      "entropy": 1.9153295263648034,
      "epoch": 0.23714330663017036,
      "grad_norm": 9.128143310546875,
      "learning_rate": 5.195054194723937e-06,
      "loss": 0.5021,
      "mean_token_accuracy": 0.840194196999073,
      "num_tokens": 92004385.0,
      "step": 76500
    },
    {
      "entropy": 1.871627089381218,
      "epoch": 0.23717430575522006,
      "grad_norm": 5.556238651275635,
      "learning_rate": 5.19471467754564e-06,
      "loss": 0.4935,
      "mean_token_accuracy": 0.8399588122963906,
      "num_tokens": 92016826.0,
      "step": 76510
    },
    {
      "entropy": 1.9132946953177452,
      "epoch": 0.23720530488026975,
      "grad_norm": 10.617101669311523,
      "learning_rate": 5.194375226924984e-06,
      "loss": 0.549,
      "mean_token_accuracy": 0.8375571563839912,
      "num_tokens": 92028616.0,
      "step": 76520
    },
    {
      "entropy": 1.8806725934147834,
      "epoch": 0.23723630400531945,
      "grad_norm": 7.826884746551514,
      "learning_rate": 5.194035842840225e-06,
      "loss": 0.501,
      "mean_token_accuracy": 0.8441179230809212,
      "num_tokens": 92040482.0,
      "step": 76530
    },
    {
      "entropy": 1.9325586080551147,
      "epoch": 0.23726730313036914,
      "grad_norm": 10.202268600463867,
      "learning_rate": 5.193696525269629e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8491972535848618,
      "num_tokens": 92052267.0,
      "step": 76540
    },
    {
      "entropy": 1.9318912595510482,
      "epoch": 0.23729830225541884,
      "grad_norm": 8.407309532165527,
      "learning_rate": 5.1933572741914726e-06,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8412642747163772,
      "num_tokens": 92063760.0,
      "step": 76550
    },
    {
      "entropy": 1.9290201410651207,
      "epoch": 0.23732930138046854,
      "grad_norm": 8.155739784240723,
      "learning_rate": 5.193018089584044e-06,
      "loss": 0.517,
      "mean_token_accuracy": 0.8428064584732056,
      "num_tokens": 92075347.0,
      "step": 76560
    },
    {
      "entropy": 1.8454096555709838,
      "epoch": 0.23736030050551823,
      "grad_norm": 10.54609203338623,
      "learning_rate": 5.192678971425639e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8462561905384064,
      "num_tokens": 92087679.0,
      "step": 76570
    },
    {
      "entropy": 1.8874484971165657,
      "epoch": 0.23739129963056793,
      "grad_norm": 9.038010597229004,
      "learning_rate": 5.192339919694561e-06,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8541138485074043,
      "num_tokens": 92099382.0,
      "step": 76580
    },
    {
      "entropy": 1.9192816317081451,
      "epoch": 0.23742229875561763,
      "grad_norm": 9.369587898254395,
      "learning_rate": 5.192000934369129e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8446416437625885,
      "num_tokens": 92111204.0,
      "step": 76590
    },
    {
      "entropy": 1.9248986691236496,
      "epoch": 0.23745329788066732,
      "grad_norm": 8.209547996520996,
      "learning_rate": 5.19166201542767e-06,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8277073204517365,
      "num_tokens": 92122562.0,
      "step": 76600
    },
    {
      "entropy": 1.8582263216376305,
      "epoch": 0.23748429700571702,
      "grad_norm": 8.712994575500488,
      "learning_rate": 5.191323162848518e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8368935108184814,
      "num_tokens": 92135073.0,
      "step": 76610
    },
    {
      "entropy": 1.9173593461513518,
      "epoch": 0.23751529613076672,
      "grad_norm": 7.943591594696045,
      "learning_rate": 5.190984376610021e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8503361091017723,
      "num_tokens": 92146436.0,
      "step": 76620
    },
    {
      "entropy": 1.8543553605675698,
      "epoch": 0.23754629525581641,
      "grad_norm": 7.117832183837891,
      "learning_rate": 5.190645656690533e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.851584343612194,
      "num_tokens": 92158280.0,
      "step": 76630
    },
    {
      "entropy": 1.868790790438652,
      "epoch": 0.2375772943808661,
      "grad_norm": 7.763202667236328,
      "learning_rate": 5.19030700306842e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8473842933773994,
      "num_tokens": 92170413.0,
      "step": 76640
    },
    {
      "entropy": 1.9384655416011811,
      "epoch": 0.2376082935059158,
      "grad_norm": 8.59603214263916,
      "learning_rate": 5.189968415722057e-06,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8368967100977898,
      "num_tokens": 92181528.0,
      "step": 76650
    },
    {
      "entropy": 1.7697623759508132,
      "epoch": 0.2376392926309655,
      "grad_norm": 9.119267463684082,
      "learning_rate": 5.189629894629832e-06,
      "loss": 0.414,
      "mean_token_accuracy": 0.8625062674283981,
      "num_tokens": 92195483.0,
      "step": 76660
    },
    {
      "entropy": 1.8749955371022224,
      "epoch": 0.23767029175601517,
      "grad_norm": 7.122738361358643,
      "learning_rate": 5.189291439770136e-06,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8516977295279503,
      "num_tokens": 92208548.0,
      "step": 76670
    },
    {
      "entropy": 1.9658852204680444,
      "epoch": 0.23770129088106487,
      "grad_norm": 9.981183052062988,
      "learning_rate": 5.18895305112138e-06,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8255149856209755,
      "num_tokens": 92219807.0,
      "step": 76680
    },
    {
      "entropy": 1.9098512694239616,
      "epoch": 0.23773229000611457,
      "grad_norm": 8.188959121704102,
      "learning_rate": 5.188614728661975e-06,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8477085337042809,
      "num_tokens": 92231598.0,
      "step": 76690
    },
    {
      "entropy": 1.9508659318089485,
      "epoch": 0.23776328913116426,
      "grad_norm": 9.962569236755371,
      "learning_rate": 5.188276472370346e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8311663627624511,
      "num_tokens": 92243313.0,
      "step": 76700
    },
    {
      "entropy": 1.880191344022751,
      "epoch": 0.23779428825621396,
      "grad_norm": 8.909346580505371,
      "learning_rate": 5.187938282224929e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8435371667146683,
      "num_tokens": 92255339.0,
      "step": 76710
    },
    {
      "entropy": 1.8568480342626572,
      "epoch": 0.23782528738126366,
      "grad_norm": 3.7800729274749756,
      "learning_rate": 5.187600158204169e-06,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8489156365394592,
      "num_tokens": 92268366.0,
      "step": 76720
    },
    {
      "entropy": 1.9763164937496185,
      "epoch": 0.23785628650631335,
      "grad_norm": 7.895925998687744,
      "learning_rate": 5.187262100286519e-06,
      "loss": 0.6004,
      "mean_token_accuracy": 0.8200380340218544,
      "num_tokens": 92278966.0,
      "step": 76730
    },
    {
      "entropy": 1.9572820693254471,
      "epoch": 0.23788728563136305,
      "grad_norm": 8.173888206481934,
      "learning_rate": 5.186924108450444e-06,
      "loss": 0.5823,
      "mean_token_accuracy": 0.831984531879425,
      "num_tokens": 92290141.0,
      "step": 76740
    },
    {
      "entropy": 1.9257027983665467,
      "epoch": 0.23791828475641275,
      "grad_norm": 9.356558799743652,
      "learning_rate": 5.186586182674418e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8511083468794822,
      "num_tokens": 92301596.0,
      "step": 76750
    },
    {
      "entropy": 1.9547690361738206,
      "epoch": 0.23794928388146244,
      "grad_norm": 9.034090995788574,
      "learning_rate": 5.186248322936925e-06,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8326706364750862,
      "num_tokens": 92313300.0,
      "step": 76760
    },
    {
      "entropy": 1.933908286690712,
      "epoch": 0.23798028300651214,
      "grad_norm": 7.58587646484375,
      "learning_rate": 5.1859105292164594e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8463876083493233,
      "num_tokens": 92324745.0,
      "step": 76770
    },
    {
      "entropy": 1.9384202346205712,
      "epoch": 0.23801128213156184,
      "grad_norm": 7.760578632354736,
      "learning_rate": 5.185572801491523e-06,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8321873590350151,
      "num_tokens": 92335893.0,
      "step": 76780
    },
    {
      "entropy": 1.7943998739123344,
      "epoch": 0.23804228125661153,
      "grad_norm": 3.816833734512329,
      "learning_rate": 5.18523513974063e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8502035543322564,
      "num_tokens": 92348555.0,
      "step": 76790
    },
    {
      "entropy": 1.9405929937958717,
      "epoch": 0.23807328038166123,
      "grad_norm": 8.975422859191895,
      "learning_rate": 5.184897543942303e-06,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8349242195487022,
      "num_tokens": 92359797.0,
      "step": 76800
    },
    {
      "entropy": 1.9422178596258164,
      "epoch": 0.23810427950671093,
      "grad_norm": 11.456317901611328,
      "learning_rate": 5.184560014075075e-06,
      "loss": 0.5249,
      "mean_token_accuracy": 0.8406445398926735,
      "num_tokens": 92370595.0,
      "step": 76810
    },
    {
      "entropy": 1.7581544771790505,
      "epoch": 0.23813527863176062,
      "grad_norm": 4.420853614807129,
      "learning_rate": 5.184222550117491e-06,
      "loss": 0.3966,
      "mean_token_accuracy": 0.8657904848456383,
      "num_tokens": 92383924.0,
      "step": 76820
    },
    {
      "entropy": 1.8917575940489768,
      "epoch": 0.23816627775681032,
      "grad_norm": 7.945245742797852,
      "learning_rate": 5.1838851520481e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8527548462152481,
      "num_tokens": 92395855.0,
      "step": 76830
    },
    {
      "entropy": 1.9285854771733284,
      "epoch": 0.23819727688186002,
      "grad_norm": 8.082216262817383,
      "learning_rate": 5.1835478198454654e-06,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8462440460920334,
      "num_tokens": 92406773.0,
      "step": 76840
    },
    {
      "entropy": 1.887254835665226,
      "epoch": 0.2382282760069097,
      "grad_norm": 8.663360595703125,
      "learning_rate": 5.1832105534881614e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8472326830029487,
      "num_tokens": 92418785.0,
      "step": 76850
    },
    {
      "entropy": 1.8737476989626884,
      "epoch": 0.2382592751319594,
      "grad_norm": 8.888916969299316,
      "learning_rate": 5.182873352954766e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8380342468619346,
      "num_tokens": 92431715.0,
      "step": 76860
    },
    {
      "entropy": 1.8650729537010193,
      "epoch": 0.2382902742570091,
      "grad_norm": 4.145284175872803,
      "learning_rate": 5.182536218223874e-06,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8427535608410835,
      "num_tokens": 92443505.0,
      "step": 76870
    },
    {
      "entropy": 1.9061666548252105,
      "epoch": 0.2383212733820588,
      "grad_norm": 9.312609672546387,
      "learning_rate": 5.182199149274083e-06,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8470250204205513,
      "num_tokens": 92454897.0,
      "step": 76880
    },
    {
      "entropy": 1.9185165598988534,
      "epoch": 0.2383522725071085,
      "grad_norm": 7.761290550231934,
      "learning_rate": 5.181862146084008e-06,
      "loss": 0.5389,
      "mean_token_accuracy": 0.8365787997841835,
      "num_tokens": 92467145.0,
      "step": 76890
    },
    {
      "entropy": 1.893602092564106,
      "epoch": 0.2383832716321582,
      "grad_norm": 9.334866523742676,
      "learning_rate": 5.181525208632266e-06,
      "loss": 0.5021,
      "mean_token_accuracy": 0.834423853456974,
      "num_tokens": 92478937.0,
      "step": 76900
    },
    {
      "entropy": 1.9183743864297866,
      "epoch": 0.2384142707572079,
      "grad_norm": 8.78905200958252,
      "learning_rate": 5.18118833689749e-06,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8335022673010826,
      "num_tokens": 92490133.0,
      "step": 76910
    },
    {
      "entropy": 1.9075113162398338,
      "epoch": 0.23844526988225756,
      "grad_norm": 8.945623397827148,
      "learning_rate": 5.18085153085832e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8388106390833855,
      "num_tokens": 92502046.0,
      "step": 76920
    },
    {
      "entropy": 1.904564779996872,
      "epoch": 0.23847626900730726,
      "grad_norm": 7.865208625793457,
      "learning_rate": 5.180514790493405e-06,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8397575378417969,
      "num_tokens": 92513471.0,
      "step": 76930
    },
    {
      "entropy": 1.9215335533022881,
      "epoch": 0.23850726813235695,
      "grad_norm": 4.170154571533203,
      "learning_rate": 5.180178115781404e-06,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8430800303816796,
      "num_tokens": 92525752.0,
      "step": 76940
    },
    {
      "entropy": 1.8328320786356926,
      "epoch": 0.23853826725740665,
      "grad_norm": 10.626273155212402,
      "learning_rate": 5.179841506700989e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8432081758975982,
      "num_tokens": 92539459.0,
      "step": 76950
    },
    {
      "entropy": 1.8328202441334724,
      "epoch": 0.23856926638245635,
      "grad_norm": 9.927977561950684,
      "learning_rate": 5.179504963230835e-06,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8494842126965523,
      "num_tokens": 92551165.0,
      "step": 76960
    },
    {
      "entropy": 1.9268303319811821,
      "epoch": 0.23860026550750604,
      "grad_norm": 8.204046249389648,
      "learning_rate": 5.179168485349633e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8421676337718964,
      "num_tokens": 92562511.0,
      "step": 76970
    },
    {
      "entropy": 1.799267826974392,
      "epoch": 0.23863126463255574,
      "grad_norm": 3.754833221435547,
      "learning_rate": 5.178832073036083e-06,
      "loss": 0.4261,
      "mean_token_accuracy": 0.860039333999157,
      "num_tokens": 92575004.0,
      "step": 76980
    },
    {
      "entropy": 1.9153159961104393,
      "epoch": 0.23866226375760544,
      "grad_norm": 8.974563598632812,
      "learning_rate": 5.178495726268889e-06,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8295607075095177,
      "num_tokens": 92586146.0,
      "step": 76990
    },
    {
      "entropy": 1.8078723505139351,
      "epoch": 0.23869326288265513,
      "grad_norm": 7.464177131652832,
      "learning_rate": 5.178159445026772e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.863148321211338,
      "num_tokens": 92599499.0,
      "step": 77000
    },
    {
      "entropy": 1.9081050038337708,
      "epoch": 0.23872426200770483,
      "grad_norm": 10.655149459838867,
      "learning_rate": 5.17782322928846e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8429564580321312,
      "num_tokens": 92610592.0,
      "step": 77010
    },
    {
      "entropy": 1.9061660870909691,
      "epoch": 0.23875526113275453,
      "grad_norm": 8.927163124084473,
      "learning_rate": 5.177487079032687e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8398579880595207,
      "num_tokens": 92622264.0,
      "step": 77020
    },
    {
      "entropy": 1.8448384143412113,
      "epoch": 0.23878626025780422,
      "grad_norm": 9.433097839355469,
      "learning_rate": 5.177150994238202e-06,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8562486276030541,
      "num_tokens": 92634783.0,
      "step": 77030
    },
    {
      "entropy": 1.8873264119029045,
      "epoch": 0.23881725938285392,
      "grad_norm": 3.7170372009277344,
      "learning_rate": 5.176814974883761e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8441817805171012,
      "num_tokens": 92646458.0,
      "step": 77040
    },
    {
      "entropy": 1.9377914816141129,
      "epoch": 0.23884825850790362,
      "grad_norm": 7.460235118865967,
      "learning_rate": 5.176479020948127e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8441906869411469,
      "num_tokens": 92657504.0,
      "step": 77050
    },
    {
      "entropy": 1.9209259897470474,
      "epoch": 0.2388792576329533,
      "grad_norm": 8.706901550292969,
      "learning_rate": 5.1761431324100805e-06,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8378788083791733,
      "num_tokens": 92668549.0,
      "step": 77060
    },
    {
      "entropy": 1.939936462044716,
      "epoch": 0.238910256758003,
      "grad_norm": 7.477108001708984,
      "learning_rate": 5.175807309248405e-06,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8331178665161133,
      "num_tokens": 92679326.0,
      "step": 77070
    },
    {
      "entropy": 1.9476049482822417,
      "epoch": 0.2389412558830527,
      "grad_norm": 10.547106742858887,
      "learning_rate": 5.175471551441896e-06,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8249987348914146,
      "num_tokens": 92690605.0,
      "step": 77080
    },
    {
      "entropy": 1.8310050159692763,
      "epoch": 0.2389722550081024,
      "grad_norm": 8.774078369140625,
      "learning_rate": 5.175135858969356e-06,
      "loss": 0.466,
      "mean_token_accuracy": 0.8467973753809929,
      "num_tokens": 92702798.0,
      "step": 77090
    },
    {
      "entropy": 1.8465019643306733,
      "epoch": 0.2390032541331521,
      "grad_norm": 4.085997104644775,
      "learning_rate": 5.174800231809601e-06,
      "loss": 0.455,
      "mean_token_accuracy": 0.8533832848072052,
      "num_tokens": 92714580.0,
      "step": 77100
    },
    {
      "entropy": 1.85148034542799,
      "epoch": 0.2390342532582018,
      "grad_norm": 9.144734382629395,
      "learning_rate": 5.174464669941455e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8366668865084648,
      "num_tokens": 92726728.0,
      "step": 77110
    },
    {
      "entropy": 1.8749561220407487,
      "epoch": 0.2390652523832515,
      "grad_norm": 4.692014217376709,
      "learning_rate": 5.17412917334375e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8414742290973664,
      "num_tokens": 92739555.0,
      "step": 77120
    },
    {
      "entropy": 1.850008523464203,
      "epoch": 0.2390962515083012,
      "grad_norm": 4.577549934387207,
      "learning_rate": 5.17379374199533e-06,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8429133415222168,
      "num_tokens": 92751167.0,
      "step": 77130
    },
    {
      "entropy": 1.7252720057964326,
      "epoch": 0.23912725063335089,
      "grad_norm": 2.2258358001708984,
      "learning_rate": 5.173458375875047e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8607361957430839,
      "num_tokens": 92764800.0,
      "step": 77140
    },
    {
      "entropy": 1.7218406319618225,
      "epoch": 0.23915824975840058,
      "grad_norm": 4.97647762298584,
      "learning_rate": 5.1731230749617645e-06,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8532850712537765,
      "num_tokens": 92779776.0,
      "step": 77150
    },
    {
      "entropy": 1.80644121915102,
      "epoch": 0.23918924888345028,
      "grad_norm": 9.306987762451172,
      "learning_rate": 5.172787839234355e-06,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8367184489965439,
      "num_tokens": 92793137.0,
      "step": 77160
    },
    {
      "entropy": 1.7925639390945434,
      "epoch": 0.23922024800849995,
      "grad_norm": 10.766707420349121,
      "learning_rate": 5.172452668671697e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8472514748573303,
      "num_tokens": 92806007.0,
      "step": 77170
    },
    {
      "entropy": 1.9033477440476418,
      "epoch": 0.23925124713354964,
      "grad_norm": 7.035861492156982,
      "learning_rate": 5.172117563252683e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8353909865021706,
      "num_tokens": 92817579.0,
      "step": 77180
    },
    {
      "entropy": 1.7921052902936936,
      "epoch": 0.23928224625859934,
      "grad_norm": 7.975375175476074,
      "learning_rate": 5.171782522956215e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8562724024057389,
      "num_tokens": 92830855.0,
      "step": 77190
    },
    {
      "entropy": 1.8471863463521003,
      "epoch": 0.23931324538364904,
      "grad_norm": 7.514707088470459,
      "learning_rate": 5.1714475477612005e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8399458363652229,
      "num_tokens": 92843224.0,
      "step": 77200
    },
    {
      "entropy": 1.9269508570432663,
      "epoch": 0.23934424450869873,
      "grad_norm": 7.373228073120117,
      "learning_rate": 5.17111263764656e-06,
      "loss": 0.535,
      "mean_token_accuracy": 0.8384528383612633,
      "num_tokens": 92854284.0,
      "step": 77210
    },
    {
      "entropy": 1.9275295376777648,
      "epoch": 0.23937524363374843,
      "grad_norm": 7.924658298492432,
      "learning_rate": 5.170777792591225e-06,
      "loss": 0.521,
      "mean_token_accuracy": 0.8433918222784996,
      "num_tokens": 92865684.0,
      "step": 77220
    },
    {
      "entropy": 1.902247828245163,
      "epoch": 0.23940624275879813,
      "grad_norm": 12.593827247619629,
      "learning_rate": 5.170443012574131e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8516241997480393,
      "num_tokens": 92876734.0,
      "step": 77230
    },
    {
      "entropy": 1.8969348236918449,
      "epoch": 0.23943724188384782,
      "grad_norm": 8.934786796569824,
      "learning_rate": 5.170108297574229e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8509429812431335,
      "num_tokens": 92888454.0,
      "step": 77240
    },
    {
      "entropy": 1.9256169840693473,
      "epoch": 0.23946824100889752,
      "grad_norm": 8.596548080444336,
      "learning_rate": 5.169773647570475e-06,
      "loss": 0.5858,
      "mean_token_accuracy": 0.8136735886335373,
      "num_tokens": 92900087.0,
      "step": 77250
    },
    {
      "entropy": 1.9090922564268111,
      "epoch": 0.23949924013394722,
      "grad_norm": 8.0187406539917,
      "learning_rate": 5.169439062541838e-06,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8492131948471069,
      "num_tokens": 92910900.0,
      "step": 77260
    },
    {
      "entropy": 1.9346697509288788,
      "epoch": 0.2395302392589969,
      "grad_norm": 8.380687713623047,
      "learning_rate": 5.1691045424672945e-06,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8416820123791695,
      "num_tokens": 92922693.0,
      "step": 77270
    },
    {
      "entropy": 1.9072400107979774,
      "epoch": 0.2395612383840466,
      "grad_norm": 6.178520679473877,
      "learning_rate": 5.16877008732583e-06,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8457990005612374,
      "num_tokens": 92934239.0,
      "step": 77280
    },
    {
      "entropy": 1.8505976639688015,
      "epoch": 0.2395922375090963,
      "grad_norm": 8.494392395019531,
      "learning_rate": 5.16843569709644e-06,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8447763130068779,
      "num_tokens": 92948046.0,
      "step": 77290
    },
    {
      "entropy": 1.8789535805583,
      "epoch": 0.239623236634146,
      "grad_norm": 8.061935424804688,
      "learning_rate": 5.168101371758133e-06,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8349943101406098,
      "num_tokens": 92959929.0,
      "step": 77300
    },
    {
      "entropy": 1.827258250117302,
      "epoch": 0.2396542357591957,
      "grad_norm": 3.748342275619507,
      "learning_rate": 5.1677671112899204e-06,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8416868060827255,
      "num_tokens": 92972031.0,
      "step": 77310
    },
    {
      "entropy": 1.892835232615471,
      "epoch": 0.2396852348842454,
      "grad_norm": 8.331387519836426,
      "learning_rate": 5.1674329156708305e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8418567150831222,
      "num_tokens": 92983623.0,
      "step": 77320
    },
    {
      "entropy": 1.8563826560974122,
      "epoch": 0.2397162340092951,
      "grad_norm": 7.926171779632568,
      "learning_rate": 5.1670987848798935e-06,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8450048446655274,
      "num_tokens": 92996094.0,
      "step": 77330
    },
    {
      "entropy": 1.93873221129179,
      "epoch": 0.2397472331343448,
      "grad_norm": 3.8182835578918457,
      "learning_rate": 5.1667647188961544e-06,
      "loss": 0.5358,
      "mean_token_accuracy": 0.8399959117174148,
      "num_tokens": 93007758.0,
      "step": 77340
    },
    {
      "entropy": 1.8997079834342003,
      "epoch": 0.23977823225939449,
      "grad_norm": 3.8245768547058105,
      "learning_rate": 5.166430717698667e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8303521856665611,
      "num_tokens": 93019975.0,
      "step": 77350
    },
    {
      "entropy": 1.9382802799344063,
      "epoch": 0.23980923138444418,
      "grad_norm": 9.995282173156738,
      "learning_rate": 5.166096781266493e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.846211564540863,
      "num_tokens": 93031399.0,
      "step": 77360
    },
    {
      "entropy": 1.872287529706955,
      "epoch": 0.23984023050949388,
      "grad_norm": 10.423619270324707,
      "learning_rate": 5.1657629095787045e-06,
      "loss": 0.5554,
      "mean_token_accuracy": 0.8343263044953346,
      "num_tokens": 93044307.0,
      "step": 77370
    },
    {
      "entropy": 1.9333751276135445,
      "epoch": 0.23987122963454358,
      "grad_norm": 9.734416007995605,
      "learning_rate": 5.165429102614382e-06,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8426417291164399,
      "num_tokens": 93055636.0,
      "step": 77380
    },
    {
      "entropy": 1.858459161221981,
      "epoch": 0.23990222875959327,
      "grad_norm": 8.448528289794922,
      "learning_rate": 5.165095360352618e-06,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8568536981940269,
      "num_tokens": 93068414.0,
      "step": 77390
    },
    {
      "entropy": 1.884053786098957,
      "epoch": 0.23993322788464297,
      "grad_norm": 9.751893043518066,
      "learning_rate": 5.164761682772511e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8394820794463158,
      "num_tokens": 93080506.0,
      "step": 77400
    },
    {
      "entropy": 1.9235117584466934,
      "epoch": 0.23996422700969264,
      "grad_norm": 8.76278305053711,
      "learning_rate": 5.164428069853172e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.8517300844192505,
      "num_tokens": 93091210.0,
      "step": 77410
    },
    {
      "entropy": 1.8631570398807527,
      "epoch": 0.23999522613474233,
      "grad_norm": 8.752660751342773,
      "learning_rate": 5.16409452157372e-06,
      "loss": 0.4473,
      "mean_token_accuracy": 0.845121631026268,
      "num_tokens": 93103714.0,
      "step": 77420
    },
    {
      "entropy": 1.831365491449833,
      "epoch": 0.24002622525979203,
      "grad_norm": 3.933938503265381,
      "learning_rate": 5.163761037913284e-06,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8512499988079071,
      "num_tokens": 93116096.0,
      "step": 77430
    },
    {
      "entropy": 1.8161715433001517,
      "epoch": 0.24005722438484173,
      "grad_norm": 7.886687755584717,
      "learning_rate": 5.163427618851002e-06,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8520355001091957,
      "num_tokens": 93128310.0,
      "step": 77440
    },
    {
      "entropy": 1.9185721650719643,
      "epoch": 0.24008822350989142,
      "grad_norm": 8.354714393615723,
      "learning_rate": 5.163094264366018e-06,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8505765274167061,
      "num_tokens": 93139885.0,
      "step": 77450
    },
    {
      "entropy": 1.881207676231861,
      "epoch": 0.24011922263494112,
      "grad_norm": 3.9551658630371094,
      "learning_rate": 5.162760974437495e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8480890035629273,
      "num_tokens": 93152362.0,
      "step": 77460
    },
    {
      "entropy": 1.8166447281837463,
      "epoch": 0.24015022175999082,
      "grad_norm": 9.925832748413086,
      "learning_rate": 5.162427749044595e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8519127413630485,
      "num_tokens": 93165382.0,
      "step": 77470
    },
    {
      "entropy": 1.9012980431318283,
      "epoch": 0.24018122088504051,
      "grad_norm": 6.780560493469238,
      "learning_rate": 5.162094588166495e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8458901271224022,
      "num_tokens": 93177192.0,
      "step": 77480
    },
    {
      "entropy": 1.8438314184546472,
      "epoch": 0.2402122200100902,
      "grad_norm": 9.430864334106445,
      "learning_rate": 5.161761491782381e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8461054712533951,
      "num_tokens": 93189929.0,
      "step": 77490
    },
    {
      "entropy": 1.8030365750193595,
      "epoch": 0.2402432191351399,
      "grad_norm": 8.649742126464844,
      "learning_rate": 5.1614284598714455e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8661585509777069,
      "num_tokens": 93203184.0,
      "step": 77500
    },
    {
      "entropy": 1.8316243276000024,
      "epoch": 0.2402742182601896,
      "grad_norm": 8.923284530639648,
      "learning_rate": 5.1610954924128944e-06,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8588346287608146,
      "num_tokens": 93215255.0,
      "step": 77510
    },
    {
      "entropy": 1.9364321306347847,
      "epoch": 0.2403052173852393,
      "grad_norm": 8.266458511352539,
      "learning_rate": 5.160762589385941e-06,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8307982221245765,
      "num_tokens": 93226205.0,
      "step": 77520
    },
    {
      "entropy": 1.868191882967949,
      "epoch": 0.240336216510289,
      "grad_norm": 9.114846229553223,
      "learning_rate": 5.160429750769805e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8514704316854477,
      "num_tokens": 93238768.0,
      "step": 77530
    },
    {
      "entropy": 1.9055574059486389,
      "epoch": 0.2403672156353387,
      "grad_norm": 4.970535755157471,
      "learning_rate": 5.160096976543722e-06,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8339187651872635,
      "num_tokens": 93251174.0,
      "step": 77540
    },
    {
      "entropy": 1.8639071702957153,
      "epoch": 0.2403982147603884,
      "grad_norm": 9.217086791992188,
      "learning_rate": 5.159764266686933e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8411014005541801,
      "num_tokens": 93263487.0,
      "step": 77550
    },
    {
      "entropy": 1.9354406654834748,
      "epoch": 0.2404292138854381,
      "grad_norm": 7.856409072875977,
      "learning_rate": 5.159431621178688e-06,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8418198600411415,
      "num_tokens": 93275107.0,
      "step": 77560
    },
    {
      "entropy": 1.9459631145000458,
      "epoch": 0.24046021301048778,
      "grad_norm": 10.06895923614502,
      "learning_rate": 5.159099039998247e-06,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8353525027632713,
      "num_tokens": 93285373.0,
      "step": 77570
    },
    {
      "entropy": 1.8992941722273826,
      "epoch": 0.24049121213553748,
      "grad_norm": 9.065903663635254,
      "learning_rate": 5.158766523124879e-06,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8420209974050522,
      "num_tokens": 93297887.0,
      "step": 77580
    },
    {
      "entropy": 1.8735466703772545,
      "epoch": 0.24052221126058718,
      "grad_norm": 8.686766624450684,
      "learning_rate": 5.158434070537864e-06,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8445511654019355,
      "num_tokens": 93309645.0,
      "step": 77590
    },
    {
      "entropy": 1.9320282906293869,
      "epoch": 0.24055321038563687,
      "grad_norm": 4.7969770431518555,
      "learning_rate": 5.158101682216491e-06,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8482805505394936,
      "num_tokens": 93320996.0,
      "step": 77600
    },
    {
      "entropy": 1.8526810958981514,
      "epoch": 0.24058420951068657,
      "grad_norm": 3.4721367359161377,
      "learning_rate": 5.157769358140056e-06,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8560703948140145,
      "num_tokens": 93332605.0,
      "step": 77610
    },
    {
      "entropy": 1.8123348727822304,
      "epoch": 0.24061520863573627,
      "grad_norm": 4.217021465301514,
      "learning_rate": 5.157437098287867e-06,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8512080743908882,
      "num_tokens": 93344938.0,
      "step": 77620
    },
    {
      "entropy": 1.955248984694481,
      "epoch": 0.24064620776078596,
      "grad_norm": 9.06552505493164,
      "learning_rate": 5.157104902639239e-06,
      "loss": 0.5756,
      "mean_token_accuracy": 0.8257305637001991,
      "num_tokens": 93356260.0,
      "step": 77630
    },
    {
      "entropy": 1.8736144185066224,
      "epoch": 0.24067720688583566,
      "grad_norm": 8.679845809936523,
      "learning_rate": 5.156772771173499e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.835148498415947,
      "num_tokens": 93368497.0,
      "step": 77640
    },
    {
      "entropy": 1.785961812734604,
      "epoch": 0.24070820601088536,
      "grad_norm": 8.887287139892578,
      "learning_rate": 5.15644070386998e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8569586247205734,
      "num_tokens": 93381040.0,
      "step": 77650
    },
    {
      "entropy": 1.8760237216949462,
      "epoch": 0.24073920513593502,
      "grad_norm": 7.9106011390686035,
      "learning_rate": 5.15610870070803e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8385165154933929,
      "num_tokens": 93392906.0,
      "step": 77660
    },
    {
      "entropy": 1.9115961521863938,
      "epoch": 0.24077020426098472,
      "grad_norm": 7.96343469619751,
      "learning_rate": 5.155776761666998e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8567873150110245,
      "num_tokens": 93404492.0,
      "step": 77670
    },
    {
      "entropy": 1.8684505842626096,
      "epoch": 0.24080120338603442,
      "grad_norm": 8.68083667755127,
      "learning_rate": 5.15544488672625e-06,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8393196210265159,
      "num_tokens": 93417211.0,
      "step": 77680
    },
    {
      "entropy": 1.7916448466479777,
      "epoch": 0.24083220251108411,
      "grad_norm": 7.983458518981934,
      "learning_rate": 5.155113075865157e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8451001644134521,
      "num_tokens": 93430639.0,
      "step": 77690
    },
    {
      "entropy": 1.8966735377907753,
      "epoch": 0.2408632016361338,
      "grad_norm": 8.755640029907227,
      "learning_rate": 5.1547813290631e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8501760110259056,
      "num_tokens": 93441446.0,
      "step": 77700
    },
    {
      "entropy": 1.8320280969142915,
      "epoch": 0.2408942007611835,
      "grad_norm": 7.649133205413818,
      "learning_rate": 5.154449646299469e-06,
      "loss": 0.443,
      "mean_token_accuracy": 0.849604444205761,
      "num_tokens": 93453520.0,
      "step": 77710
    },
    {
      "entropy": 1.826886311173439,
      "epoch": 0.2409251998862332,
      "grad_norm": 3.657710313796997,
      "learning_rate": 5.154118027553669e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.849964652955532,
      "num_tokens": 93465485.0,
      "step": 77720
    },
    {
      "entropy": 1.84391158670187,
      "epoch": 0.2409561990112829,
      "grad_norm": 8.545806884765625,
      "learning_rate": 5.153786472805101e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8504501432180405,
      "num_tokens": 93478499.0,
      "step": 77730
    },
    {
      "entropy": 1.8420162439346313,
      "epoch": 0.2409871981363326,
      "grad_norm": 4.115400791168213,
      "learning_rate": 5.15345498203319e-06,
      "loss": 0.454,
      "mean_token_accuracy": 0.8493165418505668,
      "num_tokens": 93490375.0,
      "step": 77740
    },
    {
      "entropy": 1.897342699766159,
      "epoch": 0.2410181972613823,
      "grad_norm": 9.050869941711426,
      "learning_rate": 5.153123555217362e-06,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8383392289280891,
      "num_tokens": 93501729.0,
      "step": 77750
    },
    {
      "entropy": 1.8093791991472243,
      "epoch": 0.241049196386432,
      "grad_norm": 8.670755386352539,
      "learning_rate": 5.1527921923370536e-06,
      "loss": 0.4341,
      "mean_token_accuracy": 0.8506680727005005,
      "num_tokens": 93514603.0,
      "step": 77760
    },
    {
      "entropy": 1.7984491378068923,
      "epoch": 0.2410801955114817,
      "grad_norm": 2.556086540222168,
      "learning_rate": 5.15246089337171e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8423952326178551,
      "num_tokens": 93527732.0,
      "step": 77770
    },
    {
      "entropy": 1.857538816332817,
      "epoch": 0.24111119463653138,
      "grad_norm": 5.878166675567627,
      "learning_rate": 5.15212965830079e-06,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8417029067873955,
      "num_tokens": 93539767.0,
      "step": 77780
    },
    {
      "entropy": 1.9089462801814079,
      "epoch": 0.24114219376158108,
      "grad_norm": 8.4704008102417,
      "learning_rate": 5.151798487103755e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8439002588391304,
      "num_tokens": 93550984.0,
      "step": 77790
    },
    {
      "entropy": 1.8456274837255477,
      "epoch": 0.24117319288663078,
      "grad_norm": 4.18363618850708,
      "learning_rate": 5.151467379760081e-06,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8480399310588836,
      "num_tokens": 93563757.0,
      "step": 77800
    },
    {
      "entropy": 1.899016997218132,
      "epoch": 0.24120419201168047,
      "grad_norm": 7.410801887512207,
      "learning_rate": 5.1511363362492515e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8440502628684043,
      "num_tokens": 93575421.0,
      "step": 77810
    },
    {
      "entropy": 1.8831632077693938,
      "epoch": 0.24123519113673017,
      "grad_norm": 2.9362380504608154,
      "learning_rate": 5.150805356550758e-06,
      "loss": 0.475,
      "mean_token_accuracy": 0.842276705801487,
      "num_tokens": 93587824.0,
      "step": 77820
    },
    {
      "entropy": 1.9149335369467735,
      "epoch": 0.24126619026177987,
      "grad_norm": 9.272772789001465,
      "learning_rate": 5.150474440644102e-06,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8346376657485962,
      "num_tokens": 93599403.0,
      "step": 77830
    },
    {
      "entropy": 1.8213121712207794,
      "epoch": 0.24129718938682956,
      "grad_norm": 11.08968448638916,
      "learning_rate": 5.150143588508796e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8508032530546188,
      "num_tokens": 93612087.0,
      "step": 77840
    },
    {
      "entropy": 1.832365171611309,
      "epoch": 0.24132818851187926,
      "grad_norm": 8.193666458129883,
      "learning_rate": 5.149812800124359e-06,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8404935225844383,
      "num_tokens": 93623785.0,
      "step": 77850
    },
    {
      "entropy": 1.917369608581066,
      "epoch": 0.24135918763692896,
      "grad_norm": 8.672357559204102,
      "learning_rate": 5.149482075470319e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.8380819782614708,
      "num_tokens": 93635723.0,
      "step": 77860
    },
    {
      "entropy": 1.8775453761219978,
      "epoch": 0.24139018676197865,
      "grad_norm": 9.257630348205566,
      "learning_rate": 5.1491514145262174e-06,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8449780121445656,
      "num_tokens": 93647450.0,
      "step": 77870
    },
    {
      "entropy": 1.8986103981733322,
      "epoch": 0.24142118588702835,
      "grad_norm": 8.680018424987793,
      "learning_rate": 5.148820817271601e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8502163380384445,
      "num_tokens": 93659062.0,
      "step": 77880
    },
    {
      "entropy": 1.7778849676251411,
      "epoch": 0.24145218501207805,
      "grad_norm": 4.6502604484558105,
      "learning_rate": 5.148490283686026e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8509754404425621,
      "num_tokens": 93673049.0,
      "step": 77890
    },
    {
      "entropy": 1.83993206769228,
      "epoch": 0.24148318413712774,
      "grad_norm": 3.465973138809204,
      "learning_rate": 5.14815981374906e-06,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8415236309170723,
      "num_tokens": 93685767.0,
      "step": 77900
    },
    {
      "entropy": 1.863934588432312,
      "epoch": 0.2415141832621774,
      "grad_norm": 8.9150972366333,
      "learning_rate": 5.1478294074402756e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.844880360364914,
      "num_tokens": 93698133.0,
      "step": 77910
    },
    {
      "entropy": 1.8324093401432038,
      "epoch": 0.2415451823872271,
      "grad_norm": 4.7063188552856445,
      "learning_rate": 5.14749906473926e-06,
      "loss": 0.4428,
      "mean_token_accuracy": 0.851141095161438,
      "num_tokens": 93710437.0,
      "step": 77920
    },
    {
      "entropy": 1.9022783473134042,
      "epoch": 0.2415761815122768,
      "grad_norm": 8.489745140075684,
      "learning_rate": 5.147168785625606e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8330109208822251,
      "num_tokens": 93722370.0,
      "step": 77930
    },
    {
      "entropy": 1.837750643491745,
      "epoch": 0.2416071806373265,
      "grad_norm": 8.317510604858398,
      "learning_rate": 5.146838570078916e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.8525749742984772,
      "num_tokens": 93735012.0,
      "step": 77940
    },
    {
      "entropy": 1.8423487588763237,
      "epoch": 0.2416381797623762,
      "grad_norm": 3.3121819496154785,
      "learning_rate": 5.146508418078802e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8401482835412025,
      "num_tokens": 93747947.0,
      "step": 77950
    },
    {
      "entropy": 1.858578224480152,
      "epoch": 0.2416691788874259,
      "grad_norm": 4.246115207672119,
      "learning_rate": 5.146178329604885e-06,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8309834390878678,
      "num_tokens": 93760294.0,
      "step": 77960
    },
    {
      "entropy": 1.8050771802663803,
      "epoch": 0.2417001780124756,
      "grad_norm": 8.161648750305176,
      "learning_rate": 5.145848304636797e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8522293284535408,
      "num_tokens": 93773292.0,
      "step": 77970
    },
    {
      "entropy": 1.856981235742569,
      "epoch": 0.2417311771375253,
      "grad_norm": 9.97853946685791,
      "learning_rate": 5.1455183431541755e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8453084260225296,
      "num_tokens": 93785337.0,
      "step": 77980
    },
    {
      "entropy": 1.8165911972522735,
      "epoch": 0.24176217626257498,
      "grad_norm": 8.15949535369873,
      "learning_rate": 5.145188445136669e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8522542849183082,
      "num_tokens": 93798839.0,
      "step": 77990
    },
    {
      "entropy": 1.8631312981247903,
      "epoch": 0.24179317538762468,
      "grad_norm": 9.838682174682617,
      "learning_rate": 5.144858610563938e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8407356634736061,
      "num_tokens": 93811151.0,
      "step": 78000
    },
    {
      "entropy": 1.9338207572698594,
      "epoch": 0.24182417451267438,
      "grad_norm": 9.109156608581543,
      "learning_rate": 5.144528839415645e-06,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8407235726714134,
      "num_tokens": 93822175.0,
      "step": 78010
    },
    {
      "entropy": 1.8040778622031213,
      "epoch": 0.24185517363772407,
      "grad_norm": 4.23243522644043,
      "learning_rate": 5.1441991316714694e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8540575191378593,
      "num_tokens": 93834880.0,
      "step": 78020
    },
    {
      "entropy": 1.893469262123108,
      "epoch": 0.24188617276277377,
      "grad_norm": 8.863125801086426,
      "learning_rate": 5.143869487311095e-06,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8541948691010475,
      "num_tokens": 93846671.0,
      "step": 78030
    },
    {
      "entropy": 1.8411024257540702,
      "epoch": 0.24191717188782347,
      "grad_norm": 4.275692462921143,
      "learning_rate": 5.143539906314216e-06,
      "loss": 0.4935,
      "mean_token_accuracy": 0.8460016861557961,
      "num_tokens": 93859086.0,
      "step": 78040
    },
    {
      "entropy": 1.887072142958641,
      "epoch": 0.24194817101287316,
      "grad_norm": 3.7341136932373047,
      "learning_rate": 5.143210388660536e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8523519471287727,
      "num_tokens": 93870472.0,
      "step": 78050
    },
    {
      "entropy": 1.9175701081752776,
      "epoch": 0.24197917013792286,
      "grad_norm": 10.164228439331055,
      "learning_rate": 5.142880934329766e-06,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8337636604905129,
      "num_tokens": 93881653.0,
      "step": 78060
    },
    {
      "entropy": 1.9110116347670556,
      "epoch": 0.24201016926297256,
      "grad_norm": 8.151057243347168,
      "learning_rate": 5.142551543301631e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.843655027449131,
      "num_tokens": 93893107.0,
      "step": 78070
    },
    {
      "entropy": 1.8371216788887978,
      "epoch": 0.24204116838802225,
      "grad_norm": 4.863457679748535,
      "learning_rate": 5.142222215555856e-06,
      "loss": 0.4841,
      "mean_token_accuracy": 0.845946654677391,
      "num_tokens": 93905843.0,
      "step": 78080
    },
    {
      "entropy": 1.9443501621484756,
      "epoch": 0.24207216751307195,
      "grad_norm": 9.060688972473145,
      "learning_rate": 5.141892951072186e-06,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8199516415596009,
      "num_tokens": 93916955.0,
      "step": 78090
    },
    {
      "entropy": 1.8930419281125068,
      "epoch": 0.24210316663812165,
      "grad_norm": 3.962271213531494,
      "learning_rate": 5.141563749830367e-06,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8397519171237946,
      "num_tokens": 93928520.0,
      "step": 78100
    },
    {
      "entropy": 1.88976591527462,
      "epoch": 0.24213416576317134,
      "grad_norm": 8.086103439331055,
      "learning_rate": 5.141234611810158e-06,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8491416946053505,
      "num_tokens": 93939987.0,
      "step": 78110
    },
    {
      "entropy": 1.9858234167098998,
      "epoch": 0.24216516488822104,
      "grad_norm": 7.99221658706665,
      "learning_rate": 5.140905536991324e-06,
      "loss": 0.5678,
      "mean_token_accuracy": 0.8298144713044167,
      "num_tokens": 93950877.0,
      "step": 78120
    },
    {
      "entropy": 1.8197284460067749,
      "epoch": 0.24219616401327074,
      "grad_norm": 8.68156909942627,
      "learning_rate": 5.140576525353643e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8402933716773987,
      "num_tokens": 93963270.0,
      "step": 78130
    },
    {
      "entropy": 1.8739581018686295,
      "epoch": 0.24222716313832043,
      "grad_norm": 8.117925643920898,
      "learning_rate": 5.1402475768769e-06,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8324276804924011,
      "num_tokens": 93975569.0,
      "step": 78140
    },
    {
      "entropy": 1.9055853977799415,
      "epoch": 0.2422581622633701,
      "grad_norm": 8.360283851623535,
      "learning_rate": 5.139918691540887e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8316447660326958,
      "num_tokens": 93987298.0,
      "step": 78150
    },
    {
      "entropy": 2.0304242044687273,
      "epoch": 0.2422891613884198,
      "grad_norm": 8.648979187011719,
      "learning_rate": 5.13958986932541e-06,
      "loss": 0.6085,
      "mean_token_accuracy": 0.820050160586834,
      "num_tokens": 93997995.0,
      "step": 78160
    },
    {
      "entropy": 1.9047956734895706,
      "epoch": 0.2423201605134695,
      "grad_norm": 3.7517290115356445,
      "learning_rate": 5.139261110210278e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8566797718405723,
      "num_tokens": 94009266.0,
      "step": 78170
    },
    {
      "entropy": 1.8741468638181686,
      "epoch": 0.2423511596385192,
      "grad_norm": 8.909819602966309,
      "learning_rate": 5.138932414175315e-06,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8410193353891373,
      "num_tokens": 94021292.0,
      "step": 78180
    },
    {
      "entropy": 1.9001810044050216,
      "epoch": 0.2423821587635689,
      "grad_norm": 8.745471000671387,
      "learning_rate": 5.138603781200349e-06,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8419952884316444,
      "num_tokens": 94033054.0,
      "step": 78190
    },
    {
      "entropy": 1.9287573903799058,
      "epoch": 0.24241315788861859,
      "grad_norm": 3.352858543395996,
      "learning_rate": 5.138275211265221e-06,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8414346948266029,
      "num_tokens": 94045999.0,
      "step": 78200
    },
    {
      "entropy": 1.878933884203434,
      "epoch": 0.24244415701366828,
      "grad_norm": 8.681049346923828,
      "learning_rate": 5.137946704349778e-06,
      "loss": 0.494,
      "mean_token_accuracy": 0.8540082618594169,
      "num_tokens": 94057480.0,
      "step": 78210
    },
    {
      "entropy": 1.913780263066292,
      "epoch": 0.24247515613871798,
      "grad_norm": 9.51512336730957,
      "learning_rate": 5.137618260433878e-06,
      "loss": 0.5343,
      "mean_token_accuracy": 0.8367677152156829,
      "num_tokens": 94068651.0,
      "step": 78220
    },
    {
      "entropy": 1.8172692015767098,
      "epoch": 0.24250615526376768,
      "grad_norm": 10.392766952514648,
      "learning_rate": 5.137289879497387e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8513308480381966,
      "num_tokens": 94081805.0,
      "step": 78230
    },
    {
      "entropy": 1.9101085662841797,
      "epoch": 0.24253715438881737,
      "grad_norm": 4.027338027954102,
      "learning_rate": 5.136961561520181e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8364538699388504,
      "num_tokens": 94093434.0,
      "step": 78240
    },
    {
      "entropy": 1.9400910884141922,
      "epoch": 0.24256815351386707,
      "grad_norm": 8.5806303024292,
      "learning_rate": 5.1366333064821426e-06,
      "loss": 0.5406,
      "mean_token_accuracy": 0.8367814645171165,
      "num_tokens": 94104660.0,
      "step": 78250
    },
    {
      "entropy": 1.9483735159039497,
      "epoch": 0.24259915263891677,
      "grad_norm": 4.43411922454834,
      "learning_rate": 5.136305114363167e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8446143805980683,
      "num_tokens": 94116243.0,
      "step": 78260
    },
    {
      "entropy": 1.9460048735141755,
      "epoch": 0.24263015176396646,
      "grad_norm": 8.685762405395508,
      "learning_rate": 5.1359769851431565e-06,
      "loss": 0.5085,
      "mean_token_accuracy": 0.8413349702954293,
      "num_tokens": 94127332.0,
      "step": 78270
    },
    {
      "entropy": 1.9539130926132202,
      "epoch": 0.24266115088901616,
      "grad_norm": 9.817856788635254,
      "learning_rate": 5.13564891880202e-06,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8239889681339264,
      "num_tokens": 94138346.0,
      "step": 78280
    },
    {
      "entropy": 1.8358333230018615,
      "epoch": 0.24269215001406585,
      "grad_norm": 10.21696662902832,
      "learning_rate": 5.135320915319681e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8529067754745483,
      "num_tokens": 94151143.0,
      "step": 78290
    },
    {
      "entropy": 1.9505225792527199,
      "epoch": 0.24272314913911555,
      "grad_norm": 9.021811485290527,
      "learning_rate": 5.134992974676065e-06,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8432751014828682,
      "num_tokens": 94162767.0,
      "step": 78300
    },
    {
      "entropy": 1.927219384908676,
      "epoch": 0.24275414826416525,
      "grad_norm": 4.087057590484619,
      "learning_rate": 5.134665096851114e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8410610109567642,
      "num_tokens": 94174630.0,
      "step": 78310
    },
    {
      "entropy": 1.8358907088637353,
      "epoch": 0.24278514738921494,
      "grad_norm": 9.222640991210938,
      "learning_rate": 5.134337281824774e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8533089682459831,
      "num_tokens": 94186789.0,
      "step": 78320
    },
    {
      "entropy": 1.8353676095604896,
      "epoch": 0.24281614651426464,
      "grad_norm": 9.232024192810059,
      "learning_rate": 5.1340095295769985e-06,
      "loss": 0.4311,
      "mean_token_accuracy": 0.859758959710598,
      "num_tokens": 94198760.0,
      "step": 78330
    },
    {
      "entropy": 1.9933633387088776,
      "epoch": 0.24284714563931434,
      "grad_norm": 7.663358688354492,
      "learning_rate": 5.1336818400877575e-06,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8219748124480247,
      "num_tokens": 94209973.0,
      "step": 78340
    },
    {
      "entropy": 1.9356787115335465,
      "epoch": 0.24287814476436403,
      "grad_norm": 7.0271100997924805,
      "learning_rate": 5.1333542133370205e-06,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8462739869952202,
      "num_tokens": 94221268.0,
      "step": 78350
    },
    {
      "entropy": 1.9141018971800805,
      "epoch": 0.24290914388941373,
      "grad_norm": 8.173383712768555,
      "learning_rate": 5.133026649304772e-06,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8365029126405716,
      "num_tokens": 94233710.0,
      "step": 78360
    },
    {
      "entropy": 1.8921899944543839,
      "epoch": 0.24294014301446343,
      "grad_norm": 6.497464656829834,
      "learning_rate": 5.132699147971007e-06,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8493005961179734,
      "num_tokens": 94245927.0,
      "step": 78370
    },
    {
      "entropy": 1.8819943577051164,
      "epoch": 0.24297114213951312,
      "grad_norm": 8.804264068603516,
      "learning_rate": 5.132371709315721e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8414627909660339,
      "num_tokens": 94258498.0,
      "step": 78380
    },
    {
      "entropy": 1.9550851792097093,
      "epoch": 0.24300214126456282,
      "grad_norm": 8.61777400970459,
      "learning_rate": 5.1320443333189265e-06,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8282885104417801,
      "num_tokens": 94270225.0,
      "step": 78390
    },
    {
      "entropy": 1.8870521187782288,
      "epoch": 0.2430331403896125,
      "grad_norm": 10.647053718566895,
      "learning_rate": 5.131717019960643e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.847943240404129,
      "num_tokens": 94282222.0,
      "step": 78400
    },
    {
      "entropy": 1.942089530825615,
      "epoch": 0.2430641395146622,
      "grad_norm": 9.514321327209473,
      "learning_rate": 5.131389769220897e-06,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8450910165905953,
      "num_tokens": 94293068.0,
      "step": 78410
    },
    {
      "entropy": 1.9375642448663712,
      "epoch": 0.24309513863971188,
      "grad_norm": 10.78248405456543,
      "learning_rate": 5.131062581079726e-06,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8404152438044548,
      "num_tokens": 94304441.0,
      "step": 78420
    },
    {
      "entropy": 1.901486437022686,
      "epoch": 0.24312613776476158,
      "grad_norm": 10.459092140197754,
      "learning_rate": 5.130735455517173e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8432300522923469,
      "num_tokens": 94316598.0,
      "step": 78430
    },
    {
      "entropy": 1.9251017332077027,
      "epoch": 0.24315713688981128,
      "grad_norm": 10.798014640808105,
      "learning_rate": 5.130408392513295e-06,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8358462870121002,
      "num_tokens": 94328615.0,
      "step": 78440
    },
    {
      "entropy": 1.7331562623381616,
      "epoch": 0.24318813601486097,
      "grad_norm": 6.879677772521973,
      "learning_rate": 5.130081392048156e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8590050891041756,
      "num_tokens": 94342439.0,
      "step": 78450
    },
    {
      "entropy": 1.876169492304325,
      "epoch": 0.24321913513991067,
      "grad_norm": 6.808117389678955,
      "learning_rate": 5.129754454101825e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8520722165703773,
      "num_tokens": 94354438.0,
      "step": 78460
    },
    {
      "entropy": 1.9563359498977662,
      "epoch": 0.24325013426496037,
      "grad_norm": 8.930147171020508,
      "learning_rate": 5.129427578654386e-06,
      "loss": 0.5295,
      "mean_token_accuracy": 0.84179867208004,
      "num_tokens": 94365036.0,
      "step": 78470
    },
    {
      "entropy": 1.9348897516727448,
      "epoch": 0.24328113339001006,
      "grad_norm": 8.5611572265625,
      "learning_rate": 5.129100765685926e-06,
      "loss": 0.5008,
      "mean_token_accuracy": 0.851142629981041,
      "num_tokens": 94375811.0,
      "step": 78480
    },
    {
      "entropy": 1.9472255200147628,
      "epoch": 0.24331213251505976,
      "grad_norm": 10.067473411560059,
      "learning_rate": 5.1287740151765464e-06,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8335594087839127,
      "num_tokens": 94387094.0,
      "step": 78490
    },
    {
      "entropy": 1.8310412108898162,
      "epoch": 0.24334313164010946,
      "grad_norm": 8.096028327941895,
      "learning_rate": 5.128447327106353e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8536302044987678,
      "num_tokens": 94400414.0,
      "step": 78500
    },
    {
      "entropy": 1.8954817593097686,
      "epoch": 0.24337413076515915,
      "grad_norm": 8.50610637664795,
      "learning_rate": 5.128120701455464e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8491655126214027,
      "num_tokens": 94412241.0,
      "step": 78510
    },
    {
      "entropy": 1.8588109910488129,
      "epoch": 0.24340512989020885,
      "grad_norm": 8.121857643127441,
      "learning_rate": 5.127794138204003e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8583784848451614,
      "num_tokens": 94424289.0,
      "step": 78520
    },
    {
      "entropy": 1.9707530647516251,
      "epoch": 0.24343612901525855,
      "grad_norm": 11.943368911743164,
      "learning_rate": 5.127467637332106e-06,
      "loss": 0.5347,
      "mean_token_accuracy": 0.843015332520008,
      "num_tokens": 94435163.0,
      "step": 78530
    },
    {
      "entropy": 1.8756818160414697,
      "epoch": 0.24346712814030824,
      "grad_norm": 4.680079936981201,
      "learning_rate": 5.127141198819916e-06,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8448122501373291,
      "num_tokens": 94447563.0,
      "step": 78540
    },
    {
      "entropy": 1.9033118396997453,
      "epoch": 0.24349812726535794,
      "grad_norm": 7.533633708953857,
      "learning_rate": 5.126814822647584e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8476984471082687,
      "num_tokens": 94459089.0,
      "step": 78550
    },
    {
      "entropy": 1.9617197692394257,
      "epoch": 0.24352912639040764,
      "grad_norm": 9.499611854553223,
      "learning_rate": 5.126488508795272e-06,
      "loss": 0.5812,
      "mean_token_accuracy": 0.8220301568508148,
      "num_tokens": 94469681.0,
      "step": 78560
    },
    {
      "entropy": 1.8323184587061405,
      "epoch": 0.24356012551545733,
      "grad_norm": 9.460269927978516,
      "learning_rate": 5.126162257243148e-06,
      "loss": 0.4517,
      "mean_token_accuracy": 0.8555850937962532,
      "num_tokens": 94481974.0,
      "step": 78570
    },
    {
      "entropy": 1.8734732553362847,
      "epoch": 0.24359112464050703,
      "grad_norm": 8.973634719848633,
      "learning_rate": 5.1258360679713916e-06,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8513824939727783,
      "num_tokens": 94493390.0,
      "step": 78580
    },
    {
      "entropy": 1.935685896873474,
      "epoch": 0.24362212376555673,
      "grad_norm": 7.930929183959961,
      "learning_rate": 5.125509940960189e-06,
      "loss": 0.5584,
      "mean_token_accuracy": 0.8314585655927658,
      "num_tokens": 94504378.0,
      "step": 78590
    },
    {
      "entropy": 1.881124185025692,
      "epoch": 0.24365312289060642,
      "grad_norm": 4.80453634262085,
      "learning_rate": 5.12518387618974e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8595498457551003,
      "num_tokens": 94515741.0,
      "step": 78600
    },
    {
      "entropy": 1.867940901219845,
      "epoch": 0.24368412201565612,
      "grad_norm": 8.655344009399414,
      "learning_rate": 5.124857873640244e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8418323248624802,
      "num_tokens": 94528325.0,
      "step": 78610
    },
    {
      "entropy": 1.9250017136335373,
      "epoch": 0.24371512114070581,
      "grad_norm": 9.36817741394043,
      "learning_rate": 5.124531933291918e-06,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8363332346081733,
      "num_tokens": 94539198.0,
      "step": 78620
    },
    {
      "entropy": 1.782043893635273,
      "epoch": 0.2437461202657555,
      "grad_norm": 6.611782073974609,
      "learning_rate": 5.124206055124986e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8502190038561821,
      "num_tokens": 94553598.0,
      "step": 78630
    },
    {
      "entropy": 1.9345451429486276,
      "epoch": 0.2437771193908052,
      "grad_norm": 8.913650512695312,
      "learning_rate": 5.123880239119677e-06,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8417821496725082,
      "num_tokens": 94564564.0,
      "step": 78640
    },
    {
      "entropy": 1.900984486937523,
      "epoch": 0.24380811851585488,
      "grad_norm": 7.233798503875732,
      "learning_rate": 5.123554485256231e-06,
      "loss": 0.5571,
      "mean_token_accuracy": 0.835866068303585,
      "num_tokens": 94577049.0,
      "step": 78650
    },
    {
      "entropy": 1.898327873647213,
      "epoch": 0.24383911764090457,
      "grad_norm": 9.916297912597656,
      "learning_rate": 5.123228793514897e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8348048180341721,
      "num_tokens": 94588672.0,
      "step": 78660
    },
    {
      "entropy": 1.9103514924645424,
      "epoch": 0.24387011676595427,
      "grad_norm": 7.947087287902832,
      "learning_rate": 5.122903163875935e-06,
      "loss": 0.5346,
      "mean_token_accuracy": 0.8312714904546737,
      "num_tokens": 94600977.0,
      "step": 78670
    },
    {
      "entropy": 1.9822331309318542,
      "epoch": 0.24390111589100397,
      "grad_norm": 7.2357001304626465,
      "learning_rate": 5.1225775963196104e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8357154637575149,
      "num_tokens": 94612105.0,
      "step": 78680
    },
    {
      "entropy": 1.8677599892020225,
      "epoch": 0.24393211501605366,
      "grad_norm": 8.366761207580566,
      "learning_rate": 5.1222520908261965e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8575931638479233,
      "num_tokens": 94624016.0,
      "step": 78690
    },
    {
      "entropy": 1.9394556313753128,
      "epoch": 0.24396311414110336,
      "grad_norm": 10.68550968170166,
      "learning_rate": 5.12192664737598e-06,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8385277986526489,
      "num_tokens": 94635343.0,
      "step": 78700
    },
    {
      "entropy": 1.8500966012477875,
      "epoch": 0.24399411326615306,
      "grad_norm": 8.567591667175293,
      "learning_rate": 5.121601265949253e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8590314507484436,
      "num_tokens": 94647452.0,
      "step": 78710
    },
    {
      "entropy": 1.900219763815403,
      "epoch": 0.24402511239120275,
      "grad_norm": 9.341552734375,
      "learning_rate": 5.121275946526316e-06,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8450950443744659,
      "num_tokens": 94659220.0,
      "step": 78720
    },
    {
      "entropy": 1.8428411841392518,
      "epoch": 0.24405611151625245,
      "grad_norm": 5.176535129547119,
      "learning_rate": 5.120950689087481e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8477652072906494,
      "num_tokens": 94672010.0,
      "step": 78730
    },
    {
      "entropy": 1.960240864753723,
      "epoch": 0.24408711064130215,
      "grad_norm": 7.885514736175537,
      "learning_rate": 5.120625493613066e-06,
      "loss": 0.539,
      "mean_token_accuracy": 0.8431348979473114,
      "num_tokens": 94682844.0,
      "step": 78740
    },
    {
      "entropy": 1.9407020330429077,
      "epoch": 0.24411810976635184,
      "grad_norm": 9.455016136169434,
      "learning_rate": 5.1203003600834e-06,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8410814523696899,
      "num_tokens": 94694739.0,
      "step": 78750
    },
    {
      "entropy": 1.922006744146347,
      "epoch": 0.24414910889140154,
      "grad_norm": 4.940672874450684,
      "learning_rate": 5.119975288478818e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8352184310555458,
      "num_tokens": 94707168.0,
      "step": 78760
    },
    {
      "entropy": 1.8744517169892787,
      "epoch": 0.24418010801645124,
      "grad_norm": 8.631587982177734,
      "learning_rate": 5.119650278779667e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.841625614464283,
      "num_tokens": 94719921.0,
      "step": 78770
    },
    {
      "entropy": 1.8324996635317803,
      "epoch": 0.24421110714150093,
      "grad_norm": 7.340449333190918,
      "learning_rate": 5.119325330966301e-06,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8373197898268699,
      "num_tokens": 94732979.0,
      "step": 78780
    },
    {
      "entropy": 1.820642825961113,
      "epoch": 0.24424210626655063,
      "grad_norm": 3.907349109649658,
      "learning_rate": 5.119000445019081e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.853115190565586,
      "num_tokens": 94745715.0,
      "step": 78790
    },
    {
      "entropy": 1.992446595430374,
      "epoch": 0.24427310539160033,
      "grad_norm": 8.855254173278809,
      "learning_rate": 5.118675620918381e-06,
      "loss": 0.5751,
      "mean_token_accuracy": 0.8356500491499901,
      "num_tokens": 94757429.0,
      "step": 78800
    },
    {
      "entropy": 1.915099148452282,
      "epoch": 0.24430410451665002,
      "grad_norm": 8.259538650512695,
      "learning_rate": 5.11835085864458e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8401688992977142,
      "num_tokens": 94769751.0,
      "step": 78810
    },
    {
      "entropy": 1.999873200058937,
      "epoch": 0.24433510364169972,
      "grad_norm": 8.896110534667969,
      "learning_rate": 5.118026158178065e-06,
      "loss": 0.5863,
      "mean_token_accuracy": 0.8252399504184723,
      "num_tokens": 94781117.0,
      "step": 78820
    },
    {
      "entropy": 1.9294245585799217,
      "epoch": 0.24436610276674942,
      "grad_norm": 8.745575904846191,
      "learning_rate": 5.117701519499237e-06,
      "loss": 0.5684,
      "mean_token_accuracy": 0.8319064795970916,
      "num_tokens": 94793063.0,
      "step": 78830
    },
    {
      "entropy": 1.9275419771671296,
      "epoch": 0.2443971018917991,
      "grad_norm": 9.089315414428711,
      "learning_rate": 5.1173769425885015e-06,
      "loss": 0.456,
      "mean_token_accuracy": 0.8480811953544617,
      "num_tokens": 94804637.0,
      "step": 78840
    },
    {
      "entropy": 1.8522877663373947,
      "epoch": 0.2444281010168488,
      "grad_norm": 8.082670211791992,
      "learning_rate": 5.117052427426272e-06,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8565731942653656,
      "num_tokens": 94817168.0,
      "step": 78850
    },
    {
      "entropy": 1.9246091678738595,
      "epoch": 0.2444591001418985,
      "grad_norm": 7.712028980255127,
      "learning_rate": 5.116727973992975e-06,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8396865114569664,
      "num_tokens": 94828976.0,
      "step": 78860
    },
    {
      "entropy": 1.9487199917435647,
      "epoch": 0.2444900992669482,
      "grad_norm": 8.486215591430664,
      "learning_rate": 5.116403582269041e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8476333856582642,
      "num_tokens": 94840153.0,
      "step": 78870
    },
    {
      "entropy": 1.8542052239179612,
      "epoch": 0.2445210983919979,
      "grad_norm": 9.091276168823242,
      "learning_rate": 5.11607925223491e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8544638007879257,
      "num_tokens": 94852408.0,
      "step": 78880
    },
    {
      "entropy": 1.938251782208681,
      "epoch": 0.24455209751704757,
      "grad_norm": 8.672202110290527,
      "learning_rate": 5.115754983871035e-06,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8358112215995789,
      "num_tokens": 94864131.0,
      "step": 78890
    },
    {
      "entropy": 1.9440853491425514,
      "epoch": 0.24458309664209726,
      "grad_norm": 8.338716506958008,
      "learning_rate": 5.115430777157873e-06,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8359151259064674,
      "num_tokens": 94875953.0,
      "step": 78900
    },
    {
      "entropy": 1.9307004496455193,
      "epoch": 0.24461409576714696,
      "grad_norm": 8.26486587524414,
      "learning_rate": 5.11510663207589e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8382903128862381,
      "num_tokens": 94887663.0,
      "step": 78910
    },
    {
      "entropy": 1.88083965331316,
      "epoch": 0.24464509489219666,
      "grad_norm": 9.269944190979004,
      "learning_rate": 5.114782548605563e-06,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8377712652087211,
      "num_tokens": 94899958.0,
      "step": 78920
    },
    {
      "entropy": 1.872906593978405,
      "epoch": 0.24467609401724635,
      "grad_norm": 9.342957496643066,
      "learning_rate": 5.1144585267273764e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8415094420313836,
      "num_tokens": 94912389.0,
      "step": 78930
    },
    {
      "entropy": 2.0011187493801117,
      "epoch": 0.24470709314229605,
      "grad_norm": 9.931851387023926,
      "learning_rate": 5.114134566421823e-06,
      "loss": 0.5745,
      "mean_token_accuracy": 0.8279946282505989,
      "num_tokens": 94923039.0,
      "step": 78940
    },
    {
      "entropy": 1.8712818920612335,
      "epoch": 0.24473809226734575,
      "grad_norm": 8.743650436401367,
      "learning_rate": 5.113810667669406e-06,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8516668871045112,
      "num_tokens": 94935553.0,
      "step": 78950
    },
    {
      "entropy": 1.8285594224929809,
      "epoch": 0.24476909139239544,
      "grad_norm": 6.643009662628174,
      "learning_rate": 5.1134868304506335e-06,
      "loss": 0.3967,
      "mean_token_accuracy": 0.8515212833881378,
      "num_tokens": 94949056.0,
      "step": 78960
    },
    {
      "entropy": 1.8055023223161697,
      "epoch": 0.24480009051744514,
      "grad_norm": 8.94931411743164,
      "learning_rate": 5.1131630547460264e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8564708262681962,
      "num_tokens": 94962250.0,
      "step": 78970
    },
    {
      "entropy": 1.8580831050872804,
      "epoch": 0.24483108964249484,
      "grad_norm": 8.115015983581543,
      "learning_rate": 5.112839340536111e-06,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8429727420210839,
      "num_tokens": 94975092.0,
      "step": 78980
    },
    {
      "entropy": 1.8813864275813104,
      "epoch": 0.24486208876754453,
      "grad_norm": 3.8340635299682617,
      "learning_rate": 5.112515687801425e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8568834289908409,
      "num_tokens": 94986957.0,
      "step": 78990
    },
    {
      "entropy": 1.9341845840215683,
      "epoch": 0.24489308789259423,
      "grad_norm": 9.034078598022461,
      "learning_rate": 5.112192096522513e-06,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8297013834118843,
      "num_tokens": 94998820.0,
      "step": 79000
    },
    {
      "entropy": 1.8774139389395714,
      "epoch": 0.24492408701764393,
      "grad_norm": 8.676530838012695,
      "learning_rate": 5.1118685666799276e-06,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8356031611561775,
      "num_tokens": 95011150.0,
      "step": 79010
    },
    {
      "entropy": 1.7854979574680327,
      "epoch": 0.24495508614269362,
      "grad_norm": 4.426049709320068,
      "learning_rate": 5.111545098254231e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8604443162679672,
      "num_tokens": 95024853.0,
      "step": 79020
    },
    {
      "entropy": 1.9145899727940558,
      "epoch": 0.24498608526774332,
      "grad_norm": 3.6612796783447266,
      "learning_rate": 5.111221691225996e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8452530711889267,
      "num_tokens": 95036962.0,
      "step": 79030
    },
    {
      "entropy": 1.9246152505278586,
      "epoch": 0.24501708439279302,
      "grad_norm": 4.121425628662109,
      "learning_rate": 5.1108983455758e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8439809292554855,
      "num_tokens": 95048716.0,
      "step": 79040
    },
    {
      "entropy": 1.8734241530299187,
      "epoch": 0.2450480835178427,
      "grad_norm": 7.86496114730835,
      "learning_rate": 5.110575061284232e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8426591798663139,
      "num_tokens": 95061683.0,
      "step": 79050
    },
    {
      "entropy": 1.9878466486930848,
      "epoch": 0.2450790826428924,
      "grad_norm": 7.495734691619873,
      "learning_rate": 5.110251838331888e-06,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8370781674981117,
      "num_tokens": 95072706.0,
      "step": 79060
    },
    {
      "entropy": 1.8844273030757903,
      "epoch": 0.2451100817679421,
      "grad_norm": 8.038873672485352,
      "learning_rate": 5.109928676699374e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8393307909369468,
      "num_tokens": 95084183.0,
      "step": 79070
    },
    {
      "entropy": 1.9518539249897002,
      "epoch": 0.2451410808929918,
      "grad_norm": 8.668111801147461,
      "learning_rate": 5.109605576367302e-06,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8321204602718353,
      "num_tokens": 95095191.0,
      "step": 79080
    },
    {
      "entropy": 1.8990130126476288,
      "epoch": 0.2451720800180415,
      "grad_norm": 9.605693817138672,
      "learning_rate": 5.1092825373162965e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8330251052975655,
      "num_tokens": 95107290.0,
      "step": 79090
    },
    {
      "entropy": 1.853539851307869,
      "epoch": 0.2452030791430912,
      "grad_norm": 8.671045303344727,
      "learning_rate": 5.108959559526987e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8565944582223892,
      "num_tokens": 95119603.0,
      "step": 79100
    },
    {
      "entropy": 1.9775980800390243,
      "epoch": 0.2452340782681409,
      "grad_norm": 10.049057960510254,
      "learning_rate": 5.108636642980014e-06,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8332414567470551,
      "num_tokens": 95130785.0,
      "step": 79110
    },
    {
      "entropy": 1.8694923490285873,
      "epoch": 0.2452650773931906,
      "grad_norm": 5.189975738525391,
      "learning_rate": 5.108313787656024e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8414556428790092,
      "num_tokens": 95142468.0,
      "step": 79120
    },
    {
      "entropy": 1.8693754494190216,
      "epoch": 0.24529607651824029,
      "grad_norm": 9.248258590698242,
      "learning_rate": 5.107990993535676e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.848180778324604,
      "num_tokens": 95154856.0,
      "step": 79130
    },
    {
      "entropy": 1.8843225702643394,
      "epoch": 0.24532707564328995,
      "grad_norm": 8.56281852722168,
      "learning_rate": 5.107668260599633e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8529567748308182,
      "num_tokens": 95165960.0,
      "step": 79140
    },
    {
      "entropy": 1.8180608585476876,
      "epoch": 0.24535807476833965,
      "grad_norm": 3.8715312480926514,
      "learning_rate": 5.107345588828569e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.85515176653862,
      "num_tokens": 95178500.0,
      "step": 79150
    },
    {
      "entropy": 1.9070587247610091,
      "epoch": 0.24538907389338935,
      "grad_norm": 7.640878677368164,
      "learning_rate": 5.107022978203167e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8319504678249359,
      "num_tokens": 95190327.0,
      "step": 79160
    },
    {
      "entropy": 1.8788634032011031,
      "epoch": 0.24542007301843904,
      "grad_norm": 7.821987628936768,
      "learning_rate": 5.106700428704119e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8348909318447113,
      "num_tokens": 95202577.0,
      "step": 79170
    },
    {
      "entropy": 1.8948452711105346,
      "epoch": 0.24545107214348874,
      "grad_norm": 9.048386573791504,
      "learning_rate": 5.1063779403121214e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.837831036746502,
      "num_tokens": 95214436.0,
      "step": 79180
    },
    {
      "entropy": 1.889863994717598,
      "epoch": 0.24548207126853844,
      "grad_norm": 11.092171669006348,
      "learning_rate": 5.106055513007883e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8403749257326126,
      "num_tokens": 95225844.0,
      "step": 79190
    },
    {
      "entropy": 1.8009657189249992,
      "epoch": 0.24551307039358813,
      "grad_norm": 10.045639038085938,
      "learning_rate": 5.105733146772122e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8504743695259094,
      "num_tokens": 95239379.0,
      "step": 79200
    },
    {
      "entropy": 1.9157941043376923,
      "epoch": 0.24554406951863783,
      "grad_norm": 8.408306121826172,
      "learning_rate": 5.105410841585562e-06,
      "loss": 0.4829,
      "mean_token_accuracy": 0.837490102648735,
      "num_tokens": 95251567.0,
      "step": 79210
    },
    {
      "entropy": 1.8614981457591058,
      "epoch": 0.24557506864368753,
      "grad_norm": 8.256009101867676,
      "learning_rate": 5.1050885974289354e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8346313327550888,
      "num_tokens": 95264155.0,
      "step": 79220
    },
    {
      "entropy": 1.9501688122749328,
      "epoch": 0.24560606776873722,
      "grad_norm": 8.228469848632812,
      "learning_rate": 5.104766414282987e-06,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8341888338327408,
      "num_tokens": 95275602.0,
      "step": 79230
    },
    {
      "entropy": 1.8669834539294243,
      "epoch": 0.24563706689378692,
      "grad_norm": 7.618318557739258,
      "learning_rate": 5.1044442921284635e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8466882199048996,
      "num_tokens": 95288087.0,
      "step": 79240
    },
    {
      "entropy": 2.0039695501327515,
      "epoch": 0.24566806601883662,
      "grad_norm": 10.323431968688965,
      "learning_rate": 5.104122230946127e-06,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8389030545949936,
      "num_tokens": 95298644.0,
      "step": 79250
    },
    {
      "entropy": 1.9950080424547196,
      "epoch": 0.2456990651438863,
      "grad_norm": 8.642731666564941,
      "learning_rate": 5.103800230716744e-06,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8389644399285316,
      "num_tokens": 95309147.0,
      "step": 79260
    },
    {
      "entropy": 1.9348495423793792,
      "epoch": 0.245730064268936,
      "grad_norm": 9.167795181274414,
      "learning_rate": 5.10347829142109e-06,
      "loss": 0.5279,
      "mean_token_accuracy": 0.8349600344896316,
      "num_tokens": 95320103.0,
      "step": 79270
    },
    {
      "entropy": 1.925625742971897,
      "epoch": 0.2457610633939857,
      "grad_norm": 9.226808547973633,
      "learning_rate": 5.103156413039949e-06,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8468294337391853,
      "num_tokens": 95331516.0,
      "step": 79280
    },
    {
      "entropy": 1.890987327694893,
      "epoch": 0.2457920625190354,
      "grad_norm": 9.890280723571777,
      "learning_rate": 5.102834595554116e-06,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8447692885994911,
      "num_tokens": 95343047.0,
      "step": 79290
    },
    {
      "entropy": 1.9169722646474838,
      "epoch": 0.2458230616440851,
      "grad_norm": 7.960250377655029,
      "learning_rate": 5.102512838944389e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8344372197985649,
      "num_tokens": 95354554.0,
      "step": 79300
    },
    {
      "entropy": 1.9122997641563415,
      "epoch": 0.2458540607691348,
      "grad_norm": 7.443688869476318,
      "learning_rate": 5.102191143191582e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.844282315671444,
      "num_tokens": 95367170.0,
      "step": 79310
    },
    {
      "entropy": 1.989633098244667,
      "epoch": 0.2458850598941845,
      "grad_norm": 8.14206600189209,
      "learning_rate": 5.101869508276509e-06,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8317234337329864,
      "num_tokens": 95377788.0,
      "step": 79320
    },
    {
      "entropy": 1.8243707031011582,
      "epoch": 0.2459160590192342,
      "grad_norm": 8.314239501953125,
      "learning_rate": 5.10154793418e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.863274447619915,
      "num_tokens": 95391018.0,
      "step": 79330
    },
    {
      "entropy": 2.020896875858307,
      "epoch": 0.2459470581442839,
      "grad_norm": 9.07596492767334,
      "learning_rate": 5.10122642088289e-06,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8318601101636887,
      "num_tokens": 95401518.0,
      "step": 79340
    },
    {
      "entropy": 1.9158252328634262,
      "epoch": 0.24597805726933358,
      "grad_norm": 7.7173614501953125,
      "learning_rate": 5.100904968366021e-06,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8370094135403633,
      "num_tokens": 95413144.0,
      "step": 79350
    },
    {
      "entropy": 1.8825179994106294,
      "epoch": 0.24600905639438328,
      "grad_norm": 9.029237747192383,
      "learning_rate": 5.100583576610246e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8415081441402436,
      "num_tokens": 95426258.0,
      "step": 79360
    },
    {
      "entropy": 1.8851216346025468,
      "epoch": 0.24604005551943298,
      "grad_norm": 9.818262100219727,
      "learning_rate": 5.100262245596426e-06,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8332912772893906,
      "num_tokens": 95438222.0,
      "step": 79370
    },
    {
      "entropy": 1.8941182017326355,
      "epoch": 0.24607105464448267,
      "grad_norm": 8.11664867401123,
      "learning_rate": 5.099940975305429e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8558889240026474,
      "num_tokens": 95449373.0,
      "step": 79380
    },
    {
      "entropy": 1.870937429368496,
      "epoch": 0.24610205376953234,
      "grad_norm": 9.294564247131348,
      "learning_rate": 5.099619765718133e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8433859512209892,
      "num_tokens": 95461757.0,
      "step": 79390
    },
    {
      "entropy": 1.9412083461880685,
      "epoch": 0.24613305289458204,
      "grad_norm": 9.425738334655762,
      "learning_rate": 5.099298616815426e-06,
      "loss": 0.5611,
      "mean_token_accuracy": 0.8349045276641845,
      "num_tokens": 95473347.0,
      "step": 79400
    },
    {
      "entropy": 1.8201213255524635,
      "epoch": 0.24616405201963173,
      "grad_norm": 4.238409042358398,
      "learning_rate": 5.098977528578199e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8556158289313316,
      "num_tokens": 95486624.0,
      "step": 79410
    },
    {
      "entropy": 1.8622691087424754,
      "epoch": 0.24619505114468143,
      "grad_norm": 7.4858012199401855,
      "learning_rate": 5.098656500987356e-06,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8564441815018654,
      "num_tokens": 95499802.0,
      "step": 79420
    },
    {
      "entropy": 1.823809403181076,
      "epoch": 0.24622605026973113,
      "grad_norm": 2.4532666206359863,
      "learning_rate": 5.0983355340238096e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8558837413787842,
      "num_tokens": 95512410.0,
      "step": 79430
    },
    {
      "entropy": 1.8783535480499267,
      "epoch": 0.24625704939478082,
      "grad_norm": 7.133194446563721,
      "learning_rate": 5.0980146276684775e-06,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8473603904247284,
      "num_tokens": 95524775.0,
      "step": 79440
    },
    {
      "entropy": 1.8303729638457298,
      "epoch": 0.24628804851983052,
      "grad_norm": 8.516709327697754,
      "learning_rate": 5.097693781902286e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8431586921215057,
      "num_tokens": 95537638.0,
      "step": 79450
    },
    {
      "entropy": 1.9185088947415352,
      "epoch": 0.24631904764488022,
      "grad_norm": 10.453644752502441,
      "learning_rate": 5.097372996706177e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.84544235765934,
      "num_tokens": 95549282.0,
      "step": 79460
    },
    {
      "entropy": 1.8807260736823082,
      "epoch": 0.24635004676992991,
      "grad_norm": 11.140459060668945,
      "learning_rate": 5.09705227206109e-06,
      "loss": 0.4641,
      "mean_token_accuracy": 0.8533786773681641,
      "num_tokens": 95561550.0,
      "step": 79470
    },
    {
      "entropy": 1.9539588153362275,
      "epoch": 0.2463810458949796,
      "grad_norm": 9.068120002746582,
      "learning_rate": 5.096731607947981e-06,
      "loss": 0.5577,
      "mean_token_accuracy": 0.8324769794940948,
      "num_tokens": 95572478.0,
      "step": 79480
    },
    {
      "entropy": 1.803953130543232,
      "epoch": 0.2464120450200293,
      "grad_norm": 3.0993998050689697,
      "learning_rate": 5.096411004347811e-06,
      "loss": 0.4285,
      "mean_token_accuracy": 0.8539158910512924,
      "num_tokens": 95585431.0,
      "step": 79490
    },
    {
      "entropy": 1.9023709833621978,
      "epoch": 0.246443044145079,
      "grad_norm": 8.446845054626465,
      "learning_rate": 5.096090461241549e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8470258235931396,
      "num_tokens": 95596888.0,
      "step": 79500
    },
    {
      "entropy": 1.9057755261659621,
      "epoch": 0.2464740432701287,
      "grad_norm": 3.741781711578369,
      "learning_rate": 5.095769978610174e-06,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8380037263035774,
      "num_tokens": 95608611.0,
      "step": 79510
    },
    {
      "entropy": 1.8339520961046218,
      "epoch": 0.2465050423951784,
      "grad_norm": 3.7660844326019287,
      "learning_rate": 5.095449556434673e-06,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8561963886022568,
      "num_tokens": 95621421.0,
      "step": 79520
    },
    {
      "entropy": 1.8732254639267922,
      "epoch": 0.2465360415202281,
      "grad_norm": 8.025121688842773,
      "learning_rate": 5.09512919469604e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8454299658536911,
      "num_tokens": 95633799.0,
      "step": 79530
    },
    {
      "entropy": 1.9583280727267265,
      "epoch": 0.2465670406452778,
      "grad_norm": 8.794914245605469,
      "learning_rate": 5.0948088933752795e-06,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8344408705830574,
      "num_tokens": 95645176.0,
      "step": 79540
    },
    {
      "entropy": 1.8870036020874976,
      "epoch": 0.2465980397703275,
      "grad_norm": 3.5362777709960938,
      "learning_rate": 5.094488652453403e-06,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8496022373437881,
      "num_tokens": 95657347.0,
      "step": 79550
    },
    {
      "entropy": 1.9044120475649833,
      "epoch": 0.24662903889537718,
      "grad_norm": 9.073451042175293,
      "learning_rate": 5.094168471911431e-06,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8413415655493737,
      "num_tokens": 95670276.0,
      "step": 79560
    },
    {
      "entropy": 1.8711192563176156,
      "epoch": 0.24666003802042688,
      "grad_norm": 7.03518533706665,
      "learning_rate": 5.0938483517303914e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8590395718812942,
      "num_tokens": 95682420.0,
      "step": 79570
    },
    {
      "entropy": 1.8627604782581328,
      "epoch": 0.24669103714547658,
      "grad_norm": 8.095098495483398,
      "learning_rate": 5.093528291891321e-06,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8616518348455429,
      "num_tokens": 95695131.0,
      "step": 79580
    },
    {
      "entropy": 1.8479859337210656,
      "epoch": 0.24672203627052627,
      "grad_norm": 10.04355525970459,
      "learning_rate": 5.093208292375264e-06,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8598237797617913,
      "num_tokens": 95707350.0,
      "step": 79590
    },
    {
      "entropy": 1.8836325496435165,
      "epoch": 0.24675303539557597,
      "grad_norm": 7.801244735717773,
      "learning_rate": 5.092888353163278e-06,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8486521244049072,
      "num_tokens": 95719130.0,
      "step": 79600
    },
    {
      "entropy": 1.8783444881439209,
      "epoch": 0.24678403452062567,
      "grad_norm": 7.576953411102295,
      "learning_rate": 5.092568474236419e-06,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8356266215443611,
      "num_tokens": 95730894.0,
      "step": 79610
    },
    {
      "entropy": 1.778034082055092,
      "epoch": 0.24681503364567536,
      "grad_norm": 8.774757385253906,
      "learning_rate": 5.0922486555757615e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8508021235466003,
      "num_tokens": 95743797.0,
      "step": 79620
    },
    {
      "entropy": 1.893149121105671,
      "epoch": 0.24684603277072506,
      "grad_norm": 10.234265327453613,
      "learning_rate": 5.09192889716238e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8557818323373795,
      "num_tokens": 95756512.0,
      "step": 79630
    },
    {
      "entropy": 1.9457867011427878,
      "epoch": 0.24687703189577473,
      "grad_norm": 8.266569137573242,
      "learning_rate": 5.091609198977366e-06,
      "loss": 0.5249,
      "mean_token_accuracy": 0.832012552022934,
      "num_tokens": 95767848.0,
      "step": 79640
    },
    {
      "entropy": 1.8916153132915496,
      "epoch": 0.24690803102082443,
      "grad_norm": 8.928706169128418,
      "learning_rate": 5.091289561001813e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8554543048143387,
      "num_tokens": 95779825.0,
      "step": 79650
    },
    {
      "entropy": 1.843326412141323,
      "epoch": 0.24693903014587412,
      "grad_norm": 7.779528617858887,
      "learning_rate": 5.090969983216823e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8484487190842629,
      "num_tokens": 95792518.0,
      "step": 79660
    },
    {
      "entropy": 1.8858656302094459,
      "epoch": 0.24697002927092382,
      "grad_norm": 4.888106822967529,
      "learning_rate": 5.090650465603507e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8411580830812454,
      "num_tokens": 95805316.0,
      "step": 79670
    },
    {
      "entropy": 1.9305724531412125,
      "epoch": 0.24700102839597352,
      "grad_norm": 8.88187026977539,
      "learning_rate": 5.090331008142988e-06,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8380944326519966,
      "num_tokens": 95816570.0,
      "step": 79680
    },
    {
      "entropy": 1.944003912806511,
      "epoch": 0.2470320275210232,
      "grad_norm": 8.070425033569336,
      "learning_rate": 5.090011610816392e-06,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8433306530117989,
      "num_tokens": 95827116.0,
      "step": 79690
    },
    {
      "entropy": 1.8563833236694336,
      "epoch": 0.2470630266460729,
      "grad_norm": 9.2925386428833,
      "learning_rate": 5.089692273604857e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8484684824943542,
      "num_tokens": 95839368.0,
      "step": 79700
    },
    {
      "entropy": 1.8929186090826988,
      "epoch": 0.2470940257711226,
      "grad_norm": 7.603442668914795,
      "learning_rate": 5.089372996489528e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.844200924038887,
      "num_tokens": 95850482.0,
      "step": 79710
    },
    {
      "entropy": 1.9358657032251358,
      "epoch": 0.2471250248961723,
      "grad_norm": 10.141935348510742,
      "learning_rate": 5.089053779451555e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8503325775265693,
      "num_tokens": 95861937.0,
      "step": 79720
    },
    {
      "entropy": 1.7907779216766357,
      "epoch": 0.247156024021222,
      "grad_norm": 4.321413040161133,
      "learning_rate": 5.088734622472102e-06,
      "loss": 0.4334,
      "mean_token_accuracy": 0.846727529168129,
      "num_tokens": 95875553.0,
      "step": 79730
    },
    {
      "entropy": 1.889219006896019,
      "epoch": 0.2471870231462717,
      "grad_norm": 4.462600231170654,
      "learning_rate": 5.0884155255323405e-06,
      "loss": 0.519,
      "mean_token_accuracy": 0.8342496454715729,
      "num_tokens": 95888042.0,
      "step": 79740
    },
    {
      "entropy": 1.9225512325763703,
      "epoch": 0.2472180222713214,
      "grad_norm": 8.681279182434082,
      "learning_rate": 5.088096488613445e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8418059065937996,
      "num_tokens": 95899096.0,
      "step": 79750
    },
    {
      "entropy": 1.9431275591254233,
      "epoch": 0.2472490213963711,
      "grad_norm": 10.247178077697754,
      "learning_rate": 5.087777511696603e-06,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8435952246189118,
      "num_tokens": 95910960.0,
      "step": 79760
    },
    {
      "entropy": 1.9674682512879371,
      "epoch": 0.24728002052142078,
      "grad_norm": 10.39339542388916,
      "learning_rate": 5.087458594763011e-06,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8337034076452255,
      "num_tokens": 95922652.0,
      "step": 79770
    },
    {
      "entropy": 1.9188352286815644,
      "epoch": 0.24731101964647048,
      "grad_norm": 8.393038749694824,
      "learning_rate": 5.087139737793868e-06,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8489988818764687,
      "num_tokens": 95933179.0,
      "step": 79780
    },
    {
      "entropy": 1.9350088462233543,
      "epoch": 0.24734201877152018,
      "grad_norm": 7.100698471069336,
      "learning_rate": 5.086820940770387e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8473016381263733,
      "num_tokens": 95944650.0,
      "step": 79790
    },
    {
      "entropy": 1.9587112590670586,
      "epoch": 0.24737301789656987,
      "grad_norm": 10.05678653717041,
      "learning_rate": 5.0865022036737876e-06,
      "loss": 0.5308,
      "mean_token_accuracy": 0.8289535716176033,
      "num_tokens": 95956214.0,
      "step": 79800
    },
    {
      "entropy": 1.854060137271881,
      "epoch": 0.24740401702161957,
      "grad_norm": 7.256962776184082,
      "learning_rate": 5.086183526485297e-06,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8500014215707778,
      "num_tokens": 95968951.0,
      "step": 79810
    },
    {
      "entropy": 1.9160088315606116,
      "epoch": 0.24743501614666927,
      "grad_norm": 8.998300552368164,
      "learning_rate": 5.08586490918615e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8392581641674042,
      "num_tokens": 95980577.0,
      "step": 79820
    },
    {
      "entropy": 1.88340106010437,
      "epoch": 0.24746601527171896,
      "grad_norm": 5.0286030769348145,
      "learning_rate": 5.08554635175759e-06,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8310941204428672,
      "num_tokens": 95993600.0,
      "step": 79830
    },
    {
      "entropy": 1.831264679133892,
      "epoch": 0.24749701439676866,
      "grad_norm": 9.444331169128418,
      "learning_rate": 5.085227854180872e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.847031046450138,
      "num_tokens": 96006803.0,
      "step": 79840
    },
    {
      "entropy": 1.895230557024479,
      "epoch": 0.24752801352181836,
      "grad_norm": 9.098989486694336,
      "learning_rate": 5.0849094164372525e-06,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8438019439578056,
      "num_tokens": 96018578.0,
      "step": 79850
    },
    {
      "entropy": 1.875000685453415,
      "epoch": 0.24755901264686805,
      "grad_norm": 8.089691162109375,
      "learning_rate": 5.084591038508003e-06,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8493760719895362,
      "num_tokens": 96030419.0,
      "step": 79860
    },
    {
      "entropy": 1.9285678580403327,
      "epoch": 0.24759001177191775,
      "grad_norm": 8.030923843383789,
      "learning_rate": 5.0842727203744e-06,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8497920066118241,
      "num_tokens": 96042124.0,
      "step": 79870
    },
    {
      "entropy": 1.9058194294571877,
      "epoch": 0.24762101089696742,
      "grad_norm": 6.725025653839111,
      "learning_rate": 5.083954462017727e-06,
      "loss": 0.45,
      "mean_token_accuracy": 0.8582265987992287,
      "num_tokens": 96053579.0,
      "step": 79880
    },
    {
      "entropy": 1.9877204060554505,
      "epoch": 0.24765201002201712,
      "grad_norm": 7.7386555671691895,
      "learning_rate": 5.083636263419278e-06,
      "loss": 0.5676,
      "mean_token_accuracy": 0.8407276496291161,
      "num_tokens": 96064626.0,
      "step": 79890
    },
    {
      "entropy": 1.9019395023584367,
      "epoch": 0.2476830091470668,
      "grad_norm": 8.415245056152344,
      "learning_rate": 5.083318124560355e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8463913604617119,
      "num_tokens": 96075407.0,
      "step": 79900
    },
    {
      "entropy": 1.9611292690038682,
      "epoch": 0.2477140082721165,
      "grad_norm": 7.75224494934082,
      "learning_rate": 5.083000045422266e-06,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8489187434315681,
      "num_tokens": 96086058.0,
      "step": 79910
    },
    {
      "entropy": 1.8318257644772529,
      "epoch": 0.2477450073971662,
      "grad_norm": 4.178698539733887,
      "learning_rate": 5.082682025986331e-06,
      "loss": 0.4311,
      "mean_token_accuracy": 0.8574068948626519,
      "num_tokens": 96098208.0,
      "step": 79920
    },
    {
      "entropy": 1.8961550071835518,
      "epoch": 0.2477760065222159,
      "grad_norm": 9.65478515625,
      "learning_rate": 5.082364066233872e-06,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8370327904820443,
      "num_tokens": 96109735.0,
      "step": 79930
    },
    {
      "entropy": 1.8611363500356675,
      "epoch": 0.2478070056472656,
      "grad_norm": 4.771914482116699,
      "learning_rate": 5.082046166146227e-06,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8574331164360046,
      "num_tokens": 96121606.0,
      "step": 79940
    },
    {
      "entropy": 1.8779680162668229,
      "epoch": 0.2478380047723153,
      "grad_norm": 10.091143608093262,
      "learning_rate": 5.0817283257047375e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8469004735350609,
      "num_tokens": 96133832.0,
      "step": 79950
    },
    {
      "entropy": 1.8258475840091706,
      "epoch": 0.247869003897365,
      "grad_norm": 8.701608657836914,
      "learning_rate": 5.081410544890754e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.84885775744915,
      "num_tokens": 96146484.0,
      "step": 79960
    },
    {
      "entropy": 1.856001165509224,
      "epoch": 0.2479000030224147,
      "grad_norm": 7.74146842956543,
      "learning_rate": 5.081092823685633e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8504522681236267,
      "num_tokens": 96159276.0,
      "step": 79970
    },
    {
      "entropy": 1.8811848238110542,
      "epoch": 0.24793100214746439,
      "grad_norm": 4.684483051300049,
      "learning_rate": 5.0807751620707425e-06,
      "loss": 0.4806,
      "mean_token_accuracy": 0.841371999680996,
      "num_tokens": 96171602.0,
      "step": 79980
    },
    {
      "entropy": 1.9496762931346894,
      "epoch": 0.24796200127251408,
      "grad_norm": 9.64576530456543,
      "learning_rate": 5.0804575600274575e-06,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8369770348072052,
      "num_tokens": 96182516.0,
      "step": 79990
    },
    {
      "entropy": 1.8256424218416214,
      "epoch": 0.24799300039756378,
      "grad_norm": 7.734647750854492,
      "learning_rate": 5.080140017537162e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8576287388801574,
      "num_tokens": 96195002.0,
      "step": 80000
    },
    {
      "entropy": 1.8538661286234857,
      "epoch": 0.24802399952261348,
      "grad_norm": 4.36631441116333,
      "learning_rate": 5.079822534581246e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8439467817544937,
      "num_tokens": 96207489.0,
      "step": 80010
    },
    {
      "entropy": 1.9011688023805617,
      "epoch": 0.24805499864766317,
      "grad_norm": 10.34740924835205,
      "learning_rate": 5.07950511114111e-06,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8319029092788697,
      "num_tokens": 96218430.0,
      "step": 80020
    },
    {
      "entropy": 1.8638526298105718,
      "epoch": 0.24808599777271287,
      "grad_norm": 3.3797831535339355,
      "learning_rate": 5.07918774719816e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8378943756222725,
      "num_tokens": 96231778.0,
      "step": 80030
    },
    {
      "entropy": 1.861025558412075,
      "epoch": 0.24811699689776257,
      "grad_norm": 8.062844276428223,
      "learning_rate": 5.078870442733811e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8469403609633446,
      "num_tokens": 96243932.0,
      "step": 80040
    },
    {
      "entropy": 1.835094903409481,
      "epoch": 0.24814799602281226,
      "grad_norm": 4.131919860839844,
      "learning_rate": 5.07855319772949e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8548068985342979,
      "num_tokens": 96256765.0,
      "step": 80050
    },
    {
      "entropy": 1.9781760573387146,
      "epoch": 0.24817899514786196,
      "grad_norm": 8.877656936645508,
      "learning_rate": 5.078236012166626e-06,
      "loss": 0.546,
      "mean_token_accuracy": 0.8398910224437713,
      "num_tokens": 96267560.0,
      "step": 80060
    },
    {
      "entropy": 1.8965528056025505,
      "epoch": 0.24820999427291165,
      "grad_norm": 9.912452697753906,
      "learning_rate": 5.077918886026659e-06,
      "loss": 0.521,
      "mean_token_accuracy": 0.8338564082980156,
      "num_tokens": 96279157.0,
      "step": 80070
    },
    {
      "entropy": 1.8625015437602996,
      "epoch": 0.24824099339796135,
      "grad_norm": 7.917440891265869,
      "learning_rate": 5.077601819291041e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8532622784376145,
      "num_tokens": 96291601.0,
      "step": 80080
    },
    {
      "entropy": 1.8780291944742202,
      "epoch": 0.24827199252301105,
      "grad_norm": 4.1151123046875,
      "learning_rate": 5.077284811941222e-06,
      "loss": 0.5087,
      "mean_token_accuracy": 0.837240794301033,
      "num_tokens": 96303685.0,
      "step": 80090
    },
    {
      "entropy": 1.9204677224159241,
      "epoch": 0.24830299164806074,
      "grad_norm": 8.22902774810791,
      "learning_rate": 5.076967863958671e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8477205768227577,
      "num_tokens": 96314313.0,
      "step": 80100
    },
    {
      "entropy": 1.8754928812384606,
      "epoch": 0.24833399077311044,
      "grad_norm": 3.904203176498413,
      "learning_rate": 5.076650975324857e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.852228082716465,
      "num_tokens": 96326655.0,
      "step": 80110
    },
    {
      "entropy": 1.8840164422988892,
      "epoch": 0.24836498989816014,
      "grad_norm": 7.1997504234313965,
      "learning_rate": 5.076334146021265e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.850933963060379,
      "num_tokens": 96338425.0,
      "step": 80120
    },
    {
      "entropy": 1.923794236779213,
      "epoch": 0.2483959890232098,
      "grad_norm": 9.449333190917969,
      "learning_rate": 5.076017376029378e-06,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8383274003863335,
      "num_tokens": 96349766.0,
      "step": 80130
    },
    {
      "entropy": 1.9393807545304298,
      "epoch": 0.2484269881482595,
      "grad_norm": 7.0489397048950195,
      "learning_rate": 5.0757006653306975e-06,
      "loss": 0.488,
      "mean_token_accuracy": 0.8506465151906013,
      "num_tokens": 96360918.0,
      "step": 80140
    },
    {
      "entropy": 1.9307123482227326,
      "epoch": 0.2484579872733092,
      "grad_norm": 8.270514488220215,
      "learning_rate": 5.075384013906726e-06,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8430813983082771,
      "num_tokens": 96371709.0,
      "step": 80150
    },
    {
      "entropy": 1.8894588977098465,
      "epoch": 0.2484889863983589,
      "grad_norm": 8.323099136352539,
      "learning_rate": 5.075067421738976e-06,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8491836950182915,
      "num_tokens": 96384161.0,
      "step": 80160
    },
    {
      "entropy": 1.9335136204957961,
      "epoch": 0.2485199855234086,
      "grad_norm": 8.298654556274414,
      "learning_rate": 5.074750888808969e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8408285826444626,
      "num_tokens": 96395080.0,
      "step": 80170
    },
    {
      "entropy": 1.9049769386649131,
      "epoch": 0.2485509846484583,
      "grad_norm": 8.685126304626465,
      "learning_rate": 5.074434415098235e-06,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8457057103514671,
      "num_tokens": 96406602.0,
      "step": 80180
    },
    {
      "entropy": 1.8694998741149902,
      "epoch": 0.248581983773508,
      "grad_norm": 9.182705879211426,
      "learning_rate": 5.07411800058831e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8451626107096673,
      "num_tokens": 96418323.0,
      "step": 80190
    },
    {
      "entropy": 1.8484734997153283,
      "epoch": 0.24861298289855768,
      "grad_norm": 8.897099494934082,
      "learning_rate": 5.0738016452607374e-06,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8333059296011924,
      "num_tokens": 96430631.0,
      "step": 80200
    },
    {
      "entropy": 1.9490864470601081,
      "epoch": 0.24864398202360738,
      "grad_norm": 8.925657272338867,
      "learning_rate": 5.073485349097073e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8292359367012978,
      "num_tokens": 96442572.0,
      "step": 80210
    },
    {
      "entropy": 1.9751644432544708,
      "epoch": 0.24867498114865708,
      "grad_norm": 8.735760688781738,
      "learning_rate": 5.073169112078877e-06,
      "loss": 0.5984,
      "mean_token_accuracy": 0.8311158329248428,
      "num_tokens": 96453557.0,
      "step": 80220
    },
    {
      "entropy": 1.8875069230794908,
      "epoch": 0.24870598027370677,
      "grad_norm": 3.8038547039031982,
      "learning_rate": 5.072852934187719e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8413284301757813,
      "num_tokens": 96465184.0,
      "step": 80230
    },
    {
      "entropy": 1.9452220991253852,
      "epoch": 0.24873697939875647,
      "grad_norm": 10.38296127319336,
      "learning_rate": 5.072536815405176e-06,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8376503065228462,
      "num_tokens": 96476422.0,
      "step": 80240
    },
    {
      "entropy": 1.9426960051059723,
      "epoch": 0.24876797852380617,
      "grad_norm": 8.761679649353027,
      "learning_rate": 5.072220755712832e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8422357112169265,
      "num_tokens": 96487607.0,
      "step": 80250
    },
    {
      "entropy": 1.968003484606743,
      "epoch": 0.24879897764885586,
      "grad_norm": 9.588860511779785,
      "learning_rate": 5.071904755092282e-06,
      "loss": 0.537,
      "mean_token_accuracy": 0.8418572053313256,
      "num_tokens": 96498517.0,
      "step": 80260
    },
    {
      "entropy": 1.798211957514286,
      "epoch": 0.24882997677390556,
      "grad_norm": 4.688083648681641,
      "learning_rate": 5.071588813525126e-06,
      "loss": 0.383,
      "mean_token_accuracy": 0.8539124757051468,
      "num_tokens": 96511459.0,
      "step": 80270
    },
    {
      "entropy": 1.9249870374798774,
      "epoch": 0.24886097589895526,
      "grad_norm": 9.468367576599121,
      "learning_rate": 5.071272930992976e-06,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8211545005440712,
      "num_tokens": 96523206.0,
      "step": 80280
    },
    {
      "entropy": 1.9651254445314408,
      "epoch": 0.24889197502400495,
      "grad_norm": 8.676048278808594,
      "learning_rate": 5.070957107477445e-06,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8506427973508834,
      "num_tokens": 96534006.0,
      "step": 80290
    },
    {
      "entropy": 1.9338118746876716,
      "epoch": 0.24892297414905465,
      "grad_norm": 10.065473556518555,
      "learning_rate": 5.070641342960163e-06,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8366307452321052,
      "num_tokens": 96545681.0,
      "step": 80300
    },
    {
      "entropy": 1.9056643381714822,
      "epoch": 0.24895397327410435,
      "grad_norm": 7.843869686126709,
      "learning_rate": 5.070325637422762e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8389701470732689,
      "num_tokens": 96557894.0,
      "step": 80310
    },
    {
      "entropy": 1.766680945456028,
      "epoch": 0.24898497239915404,
      "grad_norm": 8.891898155212402,
      "learning_rate": 5.070009990846881e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8574616640806199,
      "num_tokens": 96572023.0,
      "step": 80320
    },
    {
      "entropy": 1.873360113799572,
      "epoch": 0.24901597152420374,
      "grad_norm": 8.903762817382812,
      "learning_rate": 5.069694403214172e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8422929286956787,
      "num_tokens": 96584169.0,
      "step": 80330
    },
    {
      "entropy": 1.949498575925827,
      "epoch": 0.24904697064925344,
      "grad_norm": 5.78523063659668,
      "learning_rate": 5.069378874506292e-06,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8299229055643081,
      "num_tokens": 96596779.0,
      "step": 80340
    },
    {
      "entropy": 1.8998542904853821,
      "epoch": 0.24907796977430313,
      "grad_norm": 8.483941078186035,
      "learning_rate": 5.069063404704906e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8469171851873398,
      "num_tokens": 96608569.0,
      "step": 80350
    },
    {
      "entropy": 1.8074628964066506,
      "epoch": 0.24910896889935283,
      "grad_norm": 8.821971893310547,
      "learning_rate": 5.068747993791688e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.850117315351963,
      "num_tokens": 96620814.0,
      "step": 80360
    },
    {
      "entropy": 1.818614599108696,
      "epoch": 0.24913996802440252,
      "grad_norm": 8.12628173828125,
      "learning_rate": 5.068432641748318e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8480607718229294,
      "num_tokens": 96634266.0,
      "step": 80370
    },
    {
      "entropy": 1.865596404671669,
      "epoch": 0.2491709671494522,
      "grad_norm": 7.75429630279541,
      "learning_rate": 5.068117348556486e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8603037342429161,
      "num_tokens": 96646716.0,
      "step": 80380
    },
    {
      "entropy": 1.9810962960124017,
      "epoch": 0.2492019662745019,
      "grad_norm": 9.098136901855469,
      "learning_rate": 5.06780211419789e-06,
      "loss": 0.508,
      "mean_token_accuracy": 0.841071504354477,
      "num_tokens": 96657927.0,
      "step": 80390
    },
    {
      "entropy": 1.8735760763287543,
      "epoch": 0.2492329653995516,
      "grad_norm": 8.33572769165039,
      "learning_rate": 5.067486938654235e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8473378553986549,
      "num_tokens": 96670616.0,
      "step": 80400
    },
    {
      "entropy": 1.9213013619184494,
      "epoch": 0.24926396452460128,
      "grad_norm": 8.421183586120605,
      "learning_rate": 5.067171821907233e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8397013619542122,
      "num_tokens": 96682228.0,
      "step": 80410
    },
    {
      "entropy": 1.8531699359416962,
      "epoch": 0.24929496364965098,
      "grad_norm": 7.683859825134277,
      "learning_rate": 5.066856763938607e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8473792493343353,
      "num_tokens": 96695485.0,
      "step": 80420
    },
    {
      "entropy": 1.8864355862140656,
      "epoch": 0.24932596277470068,
      "grad_norm": 8.114187240600586,
      "learning_rate": 5.066541764730085e-06,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8485756784677505,
      "num_tokens": 96707903.0,
      "step": 80430
    },
    {
      "entropy": 1.9324491962790489,
      "epoch": 0.24935696189975037,
      "grad_norm": 7.229660987854004,
      "learning_rate": 5.066226824263405e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.8466485366225243,
      "num_tokens": 96719778.0,
      "step": 80440
    },
    {
      "entropy": 1.9333231955766679,
      "epoch": 0.24938796102480007,
      "grad_norm": 10.796730041503906,
      "learning_rate": 5.0659119425203116e-06,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8371896788477897,
      "num_tokens": 96731297.0,
      "step": 80450
    },
    {
      "entropy": 1.8949188530445098,
      "epoch": 0.24941896014984977,
      "grad_norm": 8.327264785766602,
      "learning_rate": 5.0655971194825586e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.840109933912754,
      "num_tokens": 96743476.0,
      "step": 80460
    },
    {
      "entropy": 1.9443549513816833,
      "epoch": 0.24944995927489946,
      "grad_norm": 8.616156578063965,
      "learning_rate": 5.065282355131904e-06,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8376484885811806,
      "num_tokens": 96755299.0,
      "step": 80470
    },
    {
      "entropy": 1.8526927456259727,
      "epoch": 0.24948095839994916,
      "grad_norm": 4.920670032501221,
      "learning_rate": 5.06496764945012e-06,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8482913061976433,
      "num_tokens": 96768603.0,
      "step": 80480
    },
    {
      "entropy": 1.9240273088216782,
      "epoch": 0.24951195752499886,
      "grad_norm": 7.758111000061035,
      "learning_rate": 5.064653002418982e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8401697605848313,
      "num_tokens": 96780005.0,
      "step": 80490
    },
    {
      "entropy": 1.8932354971766472,
      "epoch": 0.24954295665004855,
      "grad_norm": 8.145684242248535,
      "learning_rate": 5.064338414020274e-06,
      "loss": 0.449,
      "mean_token_accuracy": 0.8508092865347863,
      "num_tokens": 96792344.0,
      "step": 80500
    },
    {
      "entropy": 1.8841583669185638,
      "epoch": 0.24957395577509825,
      "grad_norm": 7.790008068084717,
      "learning_rate": 5.064023884235791e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.8376784399151802,
      "num_tokens": 96804429.0,
      "step": 80510
    },
    {
      "entropy": 1.905587163567543,
      "epoch": 0.24960495490014795,
      "grad_norm": 8.785361289978027,
      "learning_rate": 5.063709413047332e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8395880877971649,
      "num_tokens": 96815913.0,
      "step": 80520
    },
    {
      "entropy": 1.884190782904625,
      "epoch": 0.24963595402519764,
      "grad_norm": 7.5868659019470215,
      "learning_rate": 5.063395000436705e-06,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8370744809508324,
      "num_tokens": 96828062.0,
      "step": 80530
    },
    {
      "entropy": 1.9852372884750367,
      "epoch": 0.24966695315024734,
      "grad_norm": 7.919397354125977,
      "learning_rate": 5.063080646385727e-06,
      "loss": 0.5456,
      "mean_token_accuracy": 0.8372212365269661,
      "num_tokens": 96838879.0,
      "step": 80540
    },
    {
      "entropy": 1.8885437712073325,
      "epoch": 0.24969795227529704,
      "grad_norm": 8.829280853271484,
      "learning_rate": 5.062766350876223e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.8520119696855545,
      "num_tokens": 96851447.0,
      "step": 80550
    },
    {
      "entropy": 1.9246813550591468,
      "epoch": 0.24972895140034673,
      "grad_norm": 9.468807220458984,
      "learning_rate": 5.062452113890023e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8505311653017997,
      "num_tokens": 96863095.0,
      "step": 80560
    },
    {
      "entropy": 1.9204016119241714,
      "epoch": 0.24975995052539643,
      "grad_norm": 9.094423294067383,
      "learning_rate": 5.06213793540897e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8493364602327347,
      "num_tokens": 96874536.0,
      "step": 80570
    },
    {
      "entropy": 1.8319402411580086,
      "epoch": 0.24979094965044613,
      "grad_norm": 6.896250247955322,
      "learning_rate": 5.061823815414909e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8573963329195976,
      "num_tokens": 96887175.0,
      "step": 80580
    },
    {
      "entropy": 1.8221399798989295,
      "epoch": 0.24982194877549582,
      "grad_norm": 8.107512474060059,
      "learning_rate": 5.061509753889697e-06,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8606969341635704,
      "num_tokens": 96899578.0,
      "step": 80590
    },
    {
      "entropy": 1.8691432937979697,
      "epoch": 0.24985294790054552,
      "grad_norm": 4.902885913848877,
      "learning_rate": 5.0611957508152e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8467679604887962,
      "num_tokens": 96911585.0,
      "step": 80600
    },
    {
      "entropy": 1.8758586004376412,
      "epoch": 0.24988394702559522,
      "grad_norm": 9.201493263244629,
      "learning_rate": 5.0608818061732855e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8440716713666916,
      "num_tokens": 96923853.0,
      "step": 80610
    },
    {
      "entropy": 1.9490404814481734,
      "epoch": 0.24991494615064488,
      "grad_norm": 9.697155952453613,
      "learning_rate": 5.0605679199458365e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8389123737812042,
      "num_tokens": 96935124.0,
      "step": 80620
    },
    {
      "entropy": 1.86717938631773,
      "epoch": 0.24994594527569458,
      "grad_norm": 8.20489501953125,
      "learning_rate": 5.060254092114738e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8462685376405716,
      "num_tokens": 96946886.0,
      "step": 80630
    },
    {
      "entropy": 1.856597825884819,
      "epoch": 0.24997694440074428,
      "grad_norm": 8.653462409973145,
      "learning_rate": 5.059940322661886e-06,
      "loss": 0.4763,
      "mean_token_accuracy": 0.8464651450514793,
      "num_tokens": 96958754.0,
      "step": 80640
    },
    {
      "entropy": 1.9769204616546632,
      "epoch": 0.250007943525794,
      "grad_norm": 8.563497543334961,
      "learning_rate": 5.059626611569183e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8379195779561996,
      "num_tokens": 96969763.0,
      "step": 80650
    },
    {
      "entropy": 1.8927036389708518,
      "epoch": 0.2500389426508437,
      "grad_norm": 10.38947868347168,
      "learning_rate": 5.059312958818542e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8479386597871781,
      "num_tokens": 96981550.0,
      "step": 80660
    },
    {
      "entropy": 1.9301312297582627,
      "epoch": 0.25006994177589337,
      "grad_norm": 10.17356014251709,
      "learning_rate": 5.058999364391879e-06,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8410406053066254,
      "num_tokens": 96992490.0,
      "step": 80670
    },
    {
      "entropy": 1.89289371073246,
      "epoch": 0.2501009409009431,
      "grad_norm": 7.985718250274658,
      "learning_rate": 5.058685828271122e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8415110245347023,
      "num_tokens": 97004316.0,
      "step": 80680
    },
    {
      "entropy": 1.9630913496017457,
      "epoch": 0.25013194002599276,
      "grad_norm": 7.961121559143066,
      "learning_rate": 5.0583723504382044e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8376610234379769,
      "num_tokens": 97015485.0,
      "step": 80690
    },
    {
      "entropy": 1.934187889099121,
      "epoch": 0.2501629391510425,
      "grad_norm": 8.046591758728027,
      "learning_rate": 5.05805893087507e-06,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8271699488162995,
      "num_tokens": 97027583.0,
      "step": 80700
    },
    {
      "entropy": 1.9274443075060845,
      "epoch": 0.25019393827609215,
      "grad_norm": 3.1628236770629883,
      "learning_rate": 5.057745569563669e-06,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8287164881825447,
      "num_tokens": 97039330.0,
      "step": 80710
    },
    {
      "entropy": 1.9359917491674423,
      "epoch": 0.2502249374011419,
      "grad_norm": 7.84744119644165,
      "learning_rate": 5.057432266485958e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8385084256529808,
      "num_tokens": 97050574.0,
      "step": 80720
    },
    {
      "entropy": 1.9594187870621682,
      "epoch": 0.25025593652619155,
      "grad_norm": 7.170385837554932,
      "learning_rate": 5.057119021623903e-06,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8456411883234978,
      "num_tokens": 97062756.0,
      "step": 80730
    },
    {
      "entropy": 1.9243200287222861,
      "epoch": 0.25028693565124127,
      "grad_norm": 7.072485446929932,
      "learning_rate": 5.056805834959478e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8350550681352615,
      "num_tokens": 97074540.0,
      "step": 80740
    },
    {
      "entropy": 1.9237784013152122,
      "epoch": 0.25031793477629094,
      "grad_norm": 8.005240440368652,
      "learning_rate": 5.056492706474664e-06,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8389381229877472,
      "num_tokens": 97086382.0,
      "step": 80750
    },
    {
      "entropy": 1.939412146806717,
      "epoch": 0.2503489339013406,
      "grad_norm": 8.643763542175293,
      "learning_rate": 5.056179636151449e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.847502426803112,
      "num_tokens": 97097391.0,
      "step": 80760
    },
    {
      "entropy": 1.9007414281368256,
      "epoch": 0.25037993302639033,
      "grad_norm": 5.617809295654297,
      "learning_rate": 5.055866623971834e-06,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8450678214430809,
      "num_tokens": 97109904.0,
      "step": 80770
    },
    {
      "entropy": 1.9128052070736885,
      "epoch": 0.25041093215144,
      "grad_norm": 10.466955184936523,
      "learning_rate": 5.05555366991782e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8468824103474617,
      "num_tokens": 97121730.0,
      "step": 80780
    },
    {
      "entropy": 1.9782243341207504,
      "epoch": 0.2504419312764897,
      "grad_norm": 9.422632217407227,
      "learning_rate": 5.0552407739714205e-06,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8376658350229264,
      "num_tokens": 97133221.0,
      "step": 80790
    },
    {
      "entropy": 1.8325415551662445,
      "epoch": 0.2504729304015394,
      "grad_norm": 4.565164089202881,
      "learning_rate": 5.0549279361146554e-06,
      "loss": 0.3966,
      "mean_token_accuracy": 0.8581367552280426,
      "num_tokens": 97145983.0,
      "step": 80800
    },
    {
      "entropy": 1.8578097239136695,
      "epoch": 0.2505039295265891,
      "grad_norm": 8.166582107543945,
      "learning_rate": 5.0546151563295545e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8440769106149674,
      "num_tokens": 97159728.0,
      "step": 80810
    },
    {
      "entropy": 1.8590758338570594,
      "epoch": 0.2505349286516388,
      "grad_norm": 10.394712448120117,
      "learning_rate": 5.054302434598153e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8476429939270019,
      "num_tokens": 97172401.0,
      "step": 80820
    },
    {
      "entropy": 1.8516904532909393,
      "epoch": 0.2505659277766885,
      "grad_norm": 4.466270923614502,
      "learning_rate": 5.053989770902494e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8435142487287521,
      "num_tokens": 97184703.0,
      "step": 80830
    },
    {
      "entropy": 1.9154213652014733,
      "epoch": 0.2505969269017382,
      "grad_norm": 3.9221031665802,
      "learning_rate": 5.053677165224629e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8629103854298592,
      "num_tokens": 97195885.0,
      "step": 80840
    },
    {
      "entropy": 1.8582031592726707,
      "epoch": 0.2506279260267879,
      "grad_norm": 4.536048412322998,
      "learning_rate": 5.053364617546619e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8564950287342071,
      "num_tokens": 97207732.0,
      "step": 80850
    },
    {
      "entropy": 1.953849881887436,
      "epoch": 0.2506589251518376,
      "grad_norm": 7.936028003692627,
      "learning_rate": 5.05305212785053e-06,
      "loss": 0.5772,
      "mean_token_accuracy": 0.838682298362255,
      "num_tokens": 97218464.0,
      "step": 80860
    },
    {
      "entropy": 1.91454386562109,
      "epoch": 0.2506899242768873,
      "grad_norm": 10.065215110778809,
      "learning_rate": 5.052739696118435e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.837799771130085,
      "num_tokens": 97230250.0,
      "step": 80870
    },
    {
      "entropy": 1.8838393643498421,
      "epoch": 0.25072092340193697,
      "grad_norm": 9.204728126525879,
      "learning_rate": 5.05242732233242e-06,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8544922456145286,
      "num_tokens": 97242703.0,
      "step": 80880
    },
    {
      "entropy": 1.7693653479218483,
      "epoch": 0.2507519225269867,
      "grad_norm": 2.456925392150879,
      "learning_rate": 5.052115006474571e-06,
      "loss": 0.38,
      "mean_token_accuracy": 0.8611262261867523,
      "num_tokens": 97256526.0,
      "step": 80890
    },
    {
      "entropy": 1.830974417924881,
      "epoch": 0.25078292165203636,
      "grad_norm": 9.351210594177246,
      "learning_rate": 5.051802748526991e-06,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8481344610452652,
      "num_tokens": 97268711.0,
      "step": 80900
    },
    {
      "entropy": 1.8810481294989585,
      "epoch": 0.2508139207770861,
      "grad_norm": 7.888736724853516,
      "learning_rate": 5.051490548471781e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8467276513576507,
      "num_tokens": 97281167.0,
      "step": 80910
    },
    {
      "entropy": 1.9074201628565788,
      "epoch": 0.25084491990213575,
      "grad_norm": 9.632227897644043,
      "learning_rate": 5.051178406291058e-06,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8443691492080688,
      "num_tokens": 97293034.0,
      "step": 80920
    },
    {
      "entropy": 1.9555223256349563,
      "epoch": 0.2508759190271855,
      "grad_norm": 9.613581657409668,
      "learning_rate": 5.050866321966943e-06,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8227898702025414,
      "num_tokens": 97304238.0,
      "step": 80930
    },
    {
      "entropy": 1.852898570895195,
      "epoch": 0.25090691815223515,
      "grad_norm": 4.1173553466796875,
      "learning_rate": 5.050554295481563e-06,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8532208919525146,
      "num_tokens": 97317260.0,
      "step": 80940
    },
    {
      "entropy": 1.9421288311481475,
      "epoch": 0.25093791727728487,
      "grad_norm": 8.63619613647461,
      "learning_rate": 5.0502423268170556e-06,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8310228928923606,
      "num_tokens": 97329358.0,
      "step": 80950
    },
    {
      "entropy": 1.9374548494815826,
      "epoch": 0.25096891640233454,
      "grad_norm": 8.348370552062988,
      "learning_rate": 5.049930415955566e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8462329894304276,
      "num_tokens": 97340759.0,
      "step": 80960
    },
    {
      "entropy": 1.9339235559105874,
      "epoch": 0.25099991552738427,
      "grad_norm": 3.9944283962249756,
      "learning_rate": 5.049618562879247e-06,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8473946884274483,
      "num_tokens": 97351791.0,
      "step": 80970
    },
    {
      "entropy": 1.936272232234478,
      "epoch": 0.25103091465243393,
      "grad_norm": 3.7326107025146484,
      "learning_rate": 5.049306767570257e-06,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8310480803251267,
      "num_tokens": 97363291.0,
      "step": 80980
    },
    {
      "entropy": 1.924396450817585,
      "epoch": 0.2510619137774836,
      "grad_norm": 8.854395866394043,
      "learning_rate": 5.048995030010763e-06,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8389732867479325,
      "num_tokens": 97374667.0,
      "step": 80990
    },
    {
      "entropy": 1.9151296749711038,
      "epoch": 0.2510929129025333,
      "grad_norm": 8.511651992797852,
      "learning_rate": 5.048683350182941e-06,
      "loss": 0.474,
      "mean_token_accuracy": 0.8510990381240845,
      "num_tokens": 97385441.0,
      "step": 81000
    },
    {
      "entropy": 1.9651855736970902,
      "epoch": 0.251123912027583,
      "grad_norm": 8.566028594970703,
      "learning_rate": 5.048371728068976e-06,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8424704790115356,
      "num_tokens": 97396228.0,
      "step": 81010
    },
    {
      "entropy": 1.8919865861535072,
      "epoch": 0.2511549111526327,
      "grad_norm": 7.970651149749756,
      "learning_rate": 5.048060163651056e-06,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8576287001371383,
      "num_tokens": 97408337.0,
      "step": 81020
    },
    {
      "entropy": 1.8696946695446968,
      "epoch": 0.2511859102776824,
      "grad_norm": 8.631657600402832,
      "learning_rate": 5.047748656911381e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8475007191300392,
      "num_tokens": 97420330.0,
      "step": 81030
    },
    {
      "entropy": 1.8537717416882515,
      "epoch": 0.2512169094027321,
      "grad_norm": 8.173912048339844,
      "learning_rate": 5.047437207832157e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8481451213359833,
      "num_tokens": 97432563.0,
      "step": 81040
    },
    {
      "entropy": 1.8485816575586795,
      "epoch": 0.2512479085277818,
      "grad_norm": 9.571538925170898,
      "learning_rate": 5.047125816395597e-06,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8497818827629089,
      "num_tokens": 97445205.0,
      "step": 81050
    },
    {
      "entropy": 1.9077823638916016,
      "epoch": 0.2512789076528315,
      "grad_norm": 9.712011337280273,
      "learning_rate": 5.046814482583923e-06,
      "loss": 0.532,
      "mean_token_accuracy": 0.8364897713065147,
      "num_tokens": 97456191.0,
      "step": 81060
    },
    {
      "entropy": 1.9123283997178078,
      "epoch": 0.2513099067778812,
      "grad_norm": 3.9969263076782227,
      "learning_rate": 5.046503206379363e-06,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8428677946329117,
      "num_tokens": 97468188.0,
      "step": 81070
    },
    {
      "entropy": 1.8334277987480163,
      "epoch": 0.2513409059029309,
      "grad_norm": 8.850257873535156,
      "learning_rate": 5.046191987764155e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8447373628616333,
      "num_tokens": 97481082.0,
      "step": 81080
    },
    {
      "entropy": 1.9028025731444358,
      "epoch": 0.25137190502798057,
      "grad_norm": 8.100180625915527,
      "learning_rate": 5.045880826720544e-06,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8357649400830269,
      "num_tokens": 97492640.0,
      "step": 81090
    },
    {
      "entropy": 1.9044124081730842,
      "epoch": 0.2514029041530303,
      "grad_norm": 7.797017574310303,
      "learning_rate": 5.045569723230781e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.8423758924007416,
      "num_tokens": 97503638.0,
      "step": 81100
    },
    {
      "entropy": 1.9882340669631957,
      "epoch": 0.25143390327807996,
      "grad_norm": 7.203917980194092,
      "learning_rate": 5.045258677277125e-06,
      "loss": 0.557,
      "mean_token_accuracy": 0.8339864879846572,
      "num_tokens": 97514886.0,
      "step": 81110
    },
    {
      "entropy": 1.9591509833931924,
      "epoch": 0.2514649024031297,
      "grad_norm": 7.691939830780029,
      "learning_rate": 5.044947688841846e-06,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8347377508878708,
      "num_tokens": 97526189.0,
      "step": 81120
    },
    {
      "entropy": 1.8668823108077048,
      "epoch": 0.25149590152817936,
      "grad_norm": 4.000504970550537,
      "learning_rate": 5.044636757907217e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8517090499401092,
      "num_tokens": 97538527.0,
      "step": 81130
    },
    {
      "entropy": 1.8323442935943604,
      "epoch": 0.2515269006532291,
      "grad_norm": 9.106334686279297,
      "learning_rate": 5.044325884455522e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8402392759919166,
      "num_tokens": 97551259.0,
      "step": 81140
    },
    {
      "entropy": 1.8727603957057,
      "epoch": 0.25155789977827875,
      "grad_norm": 4.572244167327881,
      "learning_rate": 5.04401506846905e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8374785885214806,
      "num_tokens": 97563511.0,
      "step": 81150
    },
    {
      "entropy": 1.858438329398632,
      "epoch": 0.2515888989033285,
      "grad_norm": 7.100371360778809,
      "learning_rate": 5.0437043099301006e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8520154163241387,
      "num_tokens": 97575837.0,
      "step": 81160
    },
    {
      "entropy": 1.8653832495212554,
      "epoch": 0.25161989802837814,
      "grad_norm": 7.727576732635498,
      "learning_rate": 5.043393608820979e-06,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8588355794548989,
      "num_tokens": 97587041.0,
      "step": 81170
    },
    {
      "entropy": 1.8832016855478286,
      "epoch": 0.25165089715342787,
      "grad_norm": 9.95567512512207,
      "learning_rate": 5.043082965123996e-06,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8414867803454399,
      "num_tokens": 97598721.0,
      "step": 81180
    },
    {
      "entropy": 1.9306304231286049,
      "epoch": 0.25168189627847753,
      "grad_norm": 9.012680053710938,
      "learning_rate": 5.042772378821477e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8460228756070137,
      "num_tokens": 97609705.0,
      "step": 81190
    },
    {
      "entropy": 1.8687114715576172,
      "epoch": 0.25171289540352726,
      "grad_norm": 7.753897190093994,
      "learning_rate": 5.042461849895747e-06,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8424293950200081,
      "num_tokens": 97622075.0,
      "step": 81200
    },
    {
      "entropy": 1.8677958205342293,
      "epoch": 0.25174389452857693,
      "grad_norm": 9.426398277282715,
      "learning_rate": 5.0421513783291445e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8376675888895988,
      "num_tokens": 97634437.0,
      "step": 81210
    },
    {
      "entropy": 1.85673858076334,
      "epoch": 0.25177489365362665,
      "grad_norm": 3.3894803524017334,
      "learning_rate": 5.04184096410401e-06,
      "loss": 0.449,
      "mean_token_accuracy": 0.849704897403717,
      "num_tokens": 97647188.0,
      "step": 81220
    },
    {
      "entropy": 1.8845996797084807,
      "epoch": 0.2518058927786763,
      "grad_norm": 4.1136322021484375,
      "learning_rate": 5.041530607202698e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8388149186968803,
      "num_tokens": 97659181.0,
      "step": 81230
    },
    {
      "entropy": 1.8818348929286004,
      "epoch": 0.251836891903726,
      "grad_norm": 8.5837984085083,
      "learning_rate": 5.041220307607568e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8552572011947632,
      "num_tokens": 97670943.0,
      "step": 81240
    },
    {
      "entropy": 1.9214624166488647,
      "epoch": 0.2518678910287757,
      "grad_norm": 8.553311347961426,
      "learning_rate": 5.040910065300984e-06,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8412210613489151,
      "num_tokens": 97682965.0,
      "step": 81250
    },
    {
      "entropy": 1.7995470568537713,
      "epoch": 0.2518988901538254,
      "grad_norm": 10.36379623413086,
      "learning_rate": 5.04059988026532e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8609731838107109,
      "num_tokens": 97696313.0,
      "step": 81260
    },
    {
      "entropy": 1.9689121127128602,
      "epoch": 0.2519298892788751,
      "grad_norm": 7.462158679962158,
      "learning_rate": 5.0402897524829595e-06,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8380709275603294,
      "num_tokens": 97708096.0,
      "step": 81270
    },
    {
      "entropy": 1.9260016784071923,
      "epoch": 0.2519608884039248,
      "grad_norm": 3.709622859954834,
      "learning_rate": 5.039979681936291e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8301331534981727,
      "num_tokens": 97720285.0,
      "step": 81280
    },
    {
      "entropy": 1.868030358850956,
      "epoch": 0.2519918875289745,
      "grad_norm": 9.611808776855469,
      "learning_rate": 5.039669668607713e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8606329753994941,
      "num_tokens": 97733109.0,
      "step": 81290
    },
    {
      "entropy": 1.8924655795097352,
      "epoch": 0.25202288665402417,
      "grad_norm": 9.830230712890625,
      "learning_rate": 5.039359712479628e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8400240853428841,
      "num_tokens": 97745924.0,
      "step": 81300
    },
    {
      "entropy": 1.900289523601532,
      "epoch": 0.2520538857790739,
      "grad_norm": 7.719475746154785,
      "learning_rate": 5.039049813534448e-06,
      "loss": 0.4758,
      "mean_token_accuracy": 0.84294343739748,
      "num_tokens": 97758200.0,
      "step": 81310
    },
    {
      "entropy": 1.8474449053406716,
      "epoch": 0.25208488490412356,
      "grad_norm": 4.5430684089660645,
      "learning_rate": 5.0387399717545945e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8552279889583587,
      "num_tokens": 97770861.0,
      "step": 81320
    },
    {
      "entropy": 1.965096142888069,
      "epoch": 0.2521158840291733,
      "grad_norm": 10.388503074645996,
      "learning_rate": 5.038430187122494e-06,
      "loss": 0.4992,
      "mean_token_accuracy": 0.8290927574038506,
      "num_tokens": 97783190.0,
      "step": 81330
    },
    {
      "entropy": 1.9893094927072525,
      "epoch": 0.25214688315422296,
      "grad_norm": 7.405788898468018,
      "learning_rate": 5.03812045962058e-06,
      "loss": 0.5311,
      "mean_token_accuracy": 0.8353447288274765,
      "num_tokens": 97794623.0,
      "step": 81340
    },
    {
      "entropy": 1.9728745833039283,
      "epoch": 0.2521778822792727,
      "grad_norm": 7.6129045486450195,
      "learning_rate": 5.037810789231295e-06,
      "loss": 0.5364,
      "mean_token_accuracy": 0.8378026425838471,
      "num_tokens": 97806542.0,
      "step": 81350
    },
    {
      "entropy": 1.9820994019508362,
      "epoch": 0.25220888140432235,
      "grad_norm": 8.426169395446777,
      "learning_rate": 5.0375011759370905e-06,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8454423397779465,
      "num_tokens": 97817791.0,
      "step": 81360
    },
    {
      "entropy": 1.9337128177285194,
      "epoch": 0.2522398805293721,
      "grad_norm": 9.23006534576416,
      "learning_rate": 5.037191619720424e-06,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8424638077616692,
      "num_tokens": 97830448.0,
      "step": 81370
    },
    {
      "entropy": 1.8902158245444298,
      "epoch": 0.25227087965442174,
      "grad_norm": 8.688264846801758,
      "learning_rate": 5.036882120563758e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8471032664179802,
      "num_tokens": 97842561.0,
      "step": 81380
    },
    {
      "entropy": 1.7683439910411836,
      "epoch": 0.25230187877947147,
      "grad_norm": 7.763362407684326,
      "learning_rate": 5.036572678449568e-06,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8493596822023392,
      "num_tokens": 97856622.0,
      "step": 81390
    },
    {
      "entropy": 1.9423942849040032,
      "epoch": 0.25233287790452114,
      "grad_norm": 7.845071792602539,
      "learning_rate": 5.036263293360331e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8349541559815407,
      "num_tokens": 97867976.0,
      "step": 81400
    },
    {
      "entropy": 1.9156456768512726,
      "epoch": 0.25236387702957086,
      "grad_norm": 9.051328659057617,
      "learning_rate": 5.035953965278539e-06,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8346843421459198,
      "num_tokens": 97880411.0,
      "step": 81410
    },
    {
      "entropy": 1.9632582008838653,
      "epoch": 0.25239487615462053,
      "grad_norm": 8.78569507598877,
      "learning_rate": 5.035644694186681e-06,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8315117686986924,
      "num_tokens": 97891776.0,
      "step": 81420
    },
    {
      "entropy": 1.9228404372930528,
      "epoch": 0.25242587527967025,
      "grad_norm": 3.864513397216797,
      "learning_rate": 5.035335480067265e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8493173897266388,
      "num_tokens": 97903613.0,
      "step": 81430
    },
    {
      "entropy": 1.8714887380599976,
      "epoch": 0.2524568744047199,
      "grad_norm": 2.324655294418335,
      "learning_rate": 5.035026322902799e-06,
      "loss": 0.4154,
      "mean_token_accuracy": 0.8534365460276604,
      "num_tokens": 97915252.0,
      "step": 81440
    },
    {
      "entropy": 1.9306543365120887,
      "epoch": 0.25248787352976965,
      "grad_norm": 3.5631608963012695,
      "learning_rate": 5.0347172226758e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8418522760272026,
      "num_tokens": 97926415.0,
      "step": 81450
    },
    {
      "entropy": 1.9040459290146827,
      "epoch": 0.2525188726548193,
      "grad_norm": 8.15766429901123,
      "learning_rate": 5.034408179368794e-06,
      "loss": 0.5042,
      "mean_token_accuracy": 0.8509362369775773,
      "num_tokens": 97937394.0,
      "step": 81460
    },
    {
      "entropy": 1.8772794365882874,
      "epoch": 0.25254987177986904,
      "grad_norm": 4.18629789352417,
      "learning_rate": 5.034099192964314e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8516110569238663,
      "num_tokens": 97949121.0,
      "step": 81470
    },
    {
      "entropy": 1.9287880495190621,
      "epoch": 0.2525808709049187,
      "grad_norm": 9.347918510437012,
      "learning_rate": 5.033790263444901e-06,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8376461073756218,
      "num_tokens": 97960754.0,
      "step": 81480
    },
    {
      "entropy": 1.9006264075636863,
      "epoch": 0.2526118700299684,
      "grad_norm": 4.023174285888672,
      "learning_rate": 5.0334813907931005e-06,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8348308220505715,
      "num_tokens": 97972812.0,
      "step": 81490
    },
    {
      "entropy": 1.983270612359047,
      "epoch": 0.2526428691550181,
      "grad_norm": 10.707830429077148,
      "learning_rate": 5.033172574991469e-06,
      "loss": 0.5652,
      "mean_token_accuracy": 0.8273524597287178,
      "num_tokens": 97984573.0,
      "step": 81500
    },
    {
      "entropy": 1.8726325988769532,
      "epoch": 0.25267386828006777,
      "grad_norm": 9.031349182128906,
      "learning_rate": 5.03286381602257e-06,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8485777363181114,
      "num_tokens": 97996152.0,
      "step": 81510
    },
    {
      "entropy": 1.9527307838201522,
      "epoch": 0.2527048674051175,
      "grad_norm": 8.993306159973145,
      "learning_rate": 5.032555113868971e-06,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8380294933915138,
      "num_tokens": 98007120.0,
      "step": 81520
    },
    {
      "entropy": 1.9475104868412019,
      "epoch": 0.25273586653016716,
      "grad_norm": 8.436033248901367,
      "learning_rate": 5.032246468513252e-06,
      "loss": 0.5843,
      "mean_token_accuracy": 0.8303594037890434,
      "num_tokens": 98018955.0,
      "step": 81530
    },
    {
      "entropy": 1.8729032024741172,
      "epoch": 0.2527668656552169,
      "grad_norm": 7.274577617645264,
      "learning_rate": 5.031937879937998e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8430945709347725,
      "num_tokens": 98030512.0,
      "step": 81540
    },
    {
      "entropy": 1.8622186571359634,
      "epoch": 0.25279786478026656,
      "grad_norm": 7.648394584655762,
      "learning_rate": 5.031629348125801e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8509231805801392,
      "num_tokens": 98043454.0,
      "step": 81550
    },
    {
      "entropy": 1.8595874547958373,
      "epoch": 0.2528288639053163,
      "grad_norm": 8.356622695922852,
      "learning_rate": 5.031320873059261e-06,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8488967612385749,
      "num_tokens": 98055991.0,
      "step": 81560
    },
    {
      "entropy": 1.8365911930799483,
      "epoch": 0.25285986303036595,
      "grad_norm": 3.591733932495117,
      "learning_rate": 5.031012454720986e-06,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8576977401971817,
      "num_tokens": 98067770.0,
      "step": 81570
    },
    {
      "entropy": 1.7719722762703896,
      "epoch": 0.2528908621554157,
      "grad_norm": 3.6505837440490723,
      "learning_rate": 5.03070409309359e-06,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8484556913375855,
      "num_tokens": 98081361.0,
      "step": 81580
    },
    {
      "entropy": 1.801680639386177,
      "epoch": 0.25292186128046534,
      "grad_norm": 8.050759315490723,
      "learning_rate": 5.030395788159697e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8547103390097618,
      "num_tokens": 98094649.0,
      "step": 81590
    },
    {
      "entropy": 1.9035234346985817,
      "epoch": 0.25295286040551507,
      "grad_norm": 9.848466873168945,
      "learning_rate": 5.030087539901935e-06,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8474085479974747,
      "num_tokens": 98106101.0,
      "step": 81600
    },
    {
      "entropy": 1.925628274679184,
      "epoch": 0.25298385953056474,
      "grad_norm": 7.4657206535339355,
      "learning_rate": 5.0297793483029445e-06,
      "loss": 0.5281,
      "mean_token_accuracy": 0.8374746307730675,
      "num_tokens": 98116953.0,
      "step": 81610
    },
    {
      "entropy": 1.8128420755267143,
      "epoch": 0.25301485865561446,
      "grad_norm": 8.32465934753418,
      "learning_rate": 5.029471213345367e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.856282414495945,
      "num_tokens": 98129159.0,
      "step": 81620
    },
    {
      "entropy": 1.8045406460762023,
      "epoch": 0.25304585778066413,
      "grad_norm": 8.122908592224121,
      "learning_rate": 5.029163135011857e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8475161448121071,
      "num_tokens": 98142941.0,
      "step": 81630
    },
    {
      "entropy": 1.9032706007361413,
      "epoch": 0.25307685690571385,
      "grad_norm": 7.461269855499268,
      "learning_rate": 5.028855113285072e-06,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8353031173348426,
      "num_tokens": 98154626.0,
      "step": 81640
    },
    {
      "entropy": 1.8948104843497275,
      "epoch": 0.2531078560307635,
      "grad_norm": 3.860278606414795,
      "learning_rate": 5.02854714814768e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8362452149391174,
      "num_tokens": 98166557.0,
      "step": 81650
    },
    {
      "entropy": 1.9045625925064087,
      "epoch": 0.25313885515581325,
      "grad_norm": 8.716605186462402,
      "learning_rate": 5.028239239582357e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8562442421913147,
      "num_tokens": 98177738.0,
      "step": 81660
    },
    {
      "entropy": 1.7965163722634316,
      "epoch": 0.2531698542808629,
      "grad_norm": 7.601503849029541,
      "learning_rate": 5.027931387571784e-06,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8531433448195458,
      "num_tokens": 98190832.0,
      "step": 81670
    },
    {
      "entropy": 1.9607613369822503,
      "epoch": 0.25320085340591264,
      "grad_norm": 10.182971000671387,
      "learning_rate": 5.0276235920986505e-06,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8343812167644501,
      "num_tokens": 98202527.0,
      "step": 81680
    },
    {
      "entropy": 1.8433429718017578,
      "epoch": 0.2532318525309623,
      "grad_norm": 10.645054817199707,
      "learning_rate": 5.027315853145653e-06,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8503381237387657,
      "num_tokens": 98215514.0,
      "step": 81690
    },
    {
      "entropy": 1.8184980183839798,
      "epoch": 0.25326285165601203,
      "grad_norm": 8.565644264221191,
      "learning_rate": 5.0270081706954955e-06,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8563423335552216,
      "num_tokens": 98228449.0,
      "step": 81700
    },
    {
      "entropy": 1.9226618602871894,
      "epoch": 0.2532938507810617,
      "grad_norm": 8.709757804870605,
      "learning_rate": 5.02670054473089e-06,
      "loss": 0.5179,
      "mean_token_accuracy": 0.841064678132534,
      "num_tokens": 98240774.0,
      "step": 81710
    },
    {
      "entropy": 1.8866528853774072,
      "epoch": 0.2533248499061114,
      "grad_norm": 8.839398384094238,
      "learning_rate": 5.0263929752345564e-06,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8352084383368492,
      "num_tokens": 98253172.0,
      "step": 81720
    },
    {
      "entropy": 1.9001985654234885,
      "epoch": 0.2533558490311611,
      "grad_norm": 7.948025226593018,
      "learning_rate": 5.0260854621892196e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8508359596133233,
      "num_tokens": 98265024.0,
      "step": 81730
    },
    {
      "entropy": 1.9735320836305619,
      "epoch": 0.25338684815621076,
      "grad_norm": 8.330632209777832,
      "learning_rate": 5.0257780055776154e-06,
      "loss": 0.5311,
      "mean_token_accuracy": 0.8348188728094101,
      "num_tokens": 98275695.0,
      "step": 81740
    },
    {
      "entropy": 1.852024681866169,
      "epoch": 0.2534178472812605,
      "grad_norm": 7.069023132324219,
      "learning_rate": 5.025470605382483e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8471727296710014,
      "num_tokens": 98288269.0,
      "step": 81750
    },
    {
      "entropy": 1.8877501636743546,
      "epoch": 0.25344884640631016,
      "grad_norm": 7.931328296661377,
      "learning_rate": 5.0251632615865705e-06,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8467486530542374,
      "num_tokens": 98299962.0,
      "step": 81760
    },
    {
      "entropy": 1.9214854270219803,
      "epoch": 0.2534798455313599,
      "grad_norm": 8.707962989807129,
      "learning_rate": 5.024855974172638e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8390247240662575,
      "num_tokens": 98311678.0,
      "step": 81770
    },
    {
      "entropy": 1.88422030210495,
      "epoch": 0.25351084465640955,
      "grad_norm": 8.946457862854004,
      "learning_rate": 5.024548743123444e-06,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8396681264042855,
      "num_tokens": 98324228.0,
      "step": 81780
    },
    {
      "entropy": 1.8994328901171684,
      "epoch": 0.2535418437814593,
      "grad_norm": 8.10692024230957,
      "learning_rate": 5.024241568421762e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8410519272089004,
      "num_tokens": 98335294.0,
      "step": 81790
    },
    {
      "entropy": 1.9097337901592255,
      "epoch": 0.25357284290650894,
      "grad_norm": 3.829202175140381,
      "learning_rate": 5.02393445005037e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8425879299640655,
      "num_tokens": 98347548.0,
      "step": 81800
    },
    {
      "entropy": 1.864240688085556,
      "epoch": 0.25360384203155867,
      "grad_norm": 4.576921463012695,
      "learning_rate": 5.0236273879920534e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.8558398082852363,
      "num_tokens": 98359543.0,
      "step": 81810
    },
    {
      "entropy": 1.885629440844059,
      "epoch": 0.25363484115660834,
      "grad_norm": 8.460371017456055,
      "learning_rate": 5.023320382229604e-06,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8493375033140182,
      "num_tokens": 98370400.0,
      "step": 81820
    },
    {
      "entropy": 1.953327089548111,
      "epoch": 0.25366584028165806,
      "grad_norm": 9.160079956054688,
      "learning_rate": 5.023013432745823e-06,
      "loss": 0.532,
      "mean_token_accuracy": 0.8363123252987862,
      "num_tokens": 98381007.0,
      "step": 81830
    },
    {
      "entropy": 1.9075039952993393,
      "epoch": 0.25369683940670773,
      "grad_norm": 8.570027351379395,
      "learning_rate": 5.022706539523518e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8331753626465798,
      "num_tokens": 98393129.0,
      "step": 81840
    },
    {
      "entropy": 1.9241829916834832,
      "epoch": 0.25372783853175745,
      "grad_norm": 3.9442126750946045,
      "learning_rate": 5.022399702545504e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8449306398630142,
      "num_tokens": 98405137.0,
      "step": 81850
    },
    {
      "entropy": 1.9484668985009193,
      "epoch": 0.2537588376568071,
      "grad_norm": 9.32621955871582,
      "learning_rate": 5.022092921794602e-06,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8393065080046653,
      "num_tokens": 98416316.0,
      "step": 81860
    },
    {
      "entropy": 1.8186310246586799,
      "epoch": 0.25378983678185685,
      "grad_norm": 8.844595909118652,
      "learning_rate": 5.021786197253644e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.84593296200037,
      "num_tokens": 98429654.0,
      "step": 81870
    },
    {
      "entropy": 1.8217352986335755,
      "epoch": 0.2538208359069065,
      "grad_norm": 3.006070137023926,
      "learning_rate": 5.021479528905465e-06,
      "loss": 0.4014,
      "mean_token_accuracy": 0.8578614339232444,
      "num_tokens": 98442831.0,
      "step": 81880
    },
    {
      "entropy": 1.9646976083517074,
      "epoch": 0.25385183503195624,
      "grad_norm": 8.035730361938477,
      "learning_rate": 5.02117291673291e-06,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8446223840117455,
      "num_tokens": 98454107.0,
      "step": 81890
    },
    {
      "entropy": 1.8987499296665191,
      "epoch": 0.2538828341570059,
      "grad_norm": 7.47988748550415,
      "learning_rate": 5.02086636071883e-06,
      "loss": 0.493,
      "mean_token_accuracy": 0.8420893490314484,
      "num_tokens": 98466334.0,
      "step": 81900
    },
    {
      "entropy": 1.9272898733615875,
      "epoch": 0.25391383328205563,
      "grad_norm": 9.14543342590332,
      "learning_rate": 5.020559860846086e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.841991500556469,
      "num_tokens": 98477479.0,
      "step": 81910
    },
    {
      "entropy": 1.929322722554207,
      "epoch": 0.2539448324071053,
      "grad_norm": 10.226633071899414,
      "learning_rate": 5.020253417097542e-06,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8391476318240165,
      "num_tokens": 98489083.0,
      "step": 81920
    },
    {
      "entropy": 1.9254152104258537,
      "epoch": 0.253975831532155,
      "grad_norm": 4.747035026550293,
      "learning_rate": 5.019947029456072e-06,
      "loss": 0.5488,
      "mean_token_accuracy": 0.841267442703247,
      "num_tokens": 98500874.0,
      "step": 81930
    },
    {
      "entropy": 1.8812214568257333,
      "epoch": 0.2540068306572047,
      "grad_norm": 8.595871925354004,
      "learning_rate": 5.019640697904557e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8475135773420334,
      "num_tokens": 98513127.0,
      "step": 81940
    },
    {
      "entropy": 1.8378457948565483,
      "epoch": 0.2540378297822544,
      "grad_norm": 8.983741760253906,
      "learning_rate": 5.019334422425887e-06,
      "loss": 0.4303,
      "mean_token_accuracy": 0.8528070077300072,
      "num_tokens": 98526234.0,
      "step": 81950
    },
    {
      "entropy": 1.953259851038456,
      "epoch": 0.2540688289073041,
      "grad_norm": 8.572074890136719,
      "learning_rate": 5.019028203002956e-06,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8342129945755005,
      "num_tokens": 98537899.0,
      "step": 81960
    },
    {
      "entropy": 1.9012918874621392,
      "epoch": 0.2540998280323538,
      "grad_norm": 7.741192817687988,
      "learning_rate": 5.018722039618667e-06,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8414129927754402,
      "num_tokens": 98549927.0,
      "step": 81970
    },
    {
      "entropy": 1.890120567381382,
      "epoch": 0.2541308271574035,
      "grad_norm": 9.252531051635742,
      "learning_rate": 5.018415932255929e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8448996976017952,
      "num_tokens": 98562215.0,
      "step": 81980
    },
    {
      "entropy": 1.8428632244467735,
      "epoch": 0.25416182628245315,
      "grad_norm": 3.551053762435913,
      "learning_rate": 5.0181098808976615e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.843362420797348,
      "num_tokens": 98574911.0,
      "step": 81990
    },
    {
      "entropy": 1.9683369904756547,
      "epoch": 0.2541928254075029,
      "grad_norm": 7.981224060058594,
      "learning_rate": 5.0178038855267885e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.841914513707161,
      "num_tokens": 98585178.0,
      "step": 82000
    },
    {
      "entropy": 1.8377211585640907,
      "epoch": 0.25422382453255254,
      "grad_norm": 3.9185397624969482,
      "learning_rate": 5.017497946126241e-06,
      "loss": 0.3844,
      "mean_token_accuracy": 0.8608157888054848,
      "num_tokens": 98598026.0,
      "step": 82010
    },
    {
      "entropy": 1.882440300285816,
      "epoch": 0.25425482365760227,
      "grad_norm": 10.865336418151855,
      "learning_rate": 5.01719206267896e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8479091212153435,
      "num_tokens": 98609992.0,
      "step": 82020
    },
    {
      "entropy": 1.8768093585968018,
      "epoch": 0.25428582278265194,
      "grad_norm": 9.238245010375977,
      "learning_rate": 5.016886235167892e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8360858023166656,
      "num_tokens": 98622049.0,
      "step": 82030
    },
    {
      "entropy": 1.9196750313043593,
      "epoch": 0.25431682190770166,
      "grad_norm": 10.01637077331543,
      "learning_rate": 5.016580463575987e-06,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8404786124825477,
      "num_tokens": 98633923.0,
      "step": 82040
    },
    {
      "entropy": 1.9348924160003662,
      "epoch": 0.25434782103275133,
      "grad_norm": 8.29347038269043,
      "learning_rate": 5.016274747886213e-06,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8385466530919075,
      "num_tokens": 98646426.0,
      "step": 82050
    },
    {
      "entropy": 1.8654908925294875,
      "epoch": 0.25437882015780106,
      "grad_norm": 7.821435451507568,
      "learning_rate": 5.0159690880815324e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8516849011182785,
      "num_tokens": 98659280.0,
      "step": 82060
    },
    {
      "entropy": 1.9111699253320693,
      "epoch": 0.2544098192828507,
      "grad_norm": 7.601536273956299,
      "learning_rate": 5.015663484144925e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.83255735039711,
      "num_tokens": 98671363.0,
      "step": 82070
    },
    {
      "entropy": 1.9060380011796951,
      "epoch": 0.25444081840790045,
      "grad_norm": 7.3169965744018555,
      "learning_rate": 5.0153579360593696e-06,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8374595761299133,
      "num_tokens": 98683802.0,
      "step": 82080
    },
    {
      "entropy": 1.93545441031456,
      "epoch": 0.2544718175329501,
      "grad_norm": 4.195225238800049,
      "learning_rate": 5.015052443807861e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.846612386405468,
      "num_tokens": 98695666.0,
      "step": 82090
    },
    {
      "entropy": 1.8932590246200562,
      "epoch": 0.25450281665799984,
      "grad_norm": 4.190147399902344,
      "learning_rate": 5.014747007373392e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8346954330801963,
      "num_tokens": 98707862.0,
      "step": 82100
    },
    {
      "entropy": 1.8419014051556588,
      "epoch": 0.2545338157830495,
      "grad_norm": 9.007070541381836,
      "learning_rate": 5.014441626738971e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8538463622331619,
      "num_tokens": 98719963.0,
      "step": 82110
    },
    {
      "entropy": 1.9747213378548623,
      "epoch": 0.25456481490809924,
      "grad_norm": 7.8509297370910645,
      "learning_rate": 5.014136301887608e-06,
      "loss": 0.5592,
      "mean_token_accuracy": 0.8374496519565582,
      "num_tokens": 98731159.0,
      "step": 82120
    },
    {
      "entropy": 1.950367696583271,
      "epoch": 0.2545958140331489,
      "grad_norm": 9.41358470916748,
      "learning_rate": 5.013831032802323e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.847783799469471,
      "num_tokens": 98742604.0,
      "step": 82130
    },
    {
      "entropy": 1.967973504960537,
      "epoch": 0.25462681315819863,
      "grad_norm": 8.383604049682617,
      "learning_rate": 5.013525819466142e-06,
      "loss": 0.51,
      "mean_token_accuracy": 0.8421026647090912,
      "num_tokens": 98754125.0,
      "step": 82140
    },
    {
      "entropy": 1.984065267443657,
      "epoch": 0.2546578122832483,
      "grad_norm": 8.877246856689453,
      "learning_rate": 5.013220661862098e-06,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8417667001485825,
      "num_tokens": 98764605.0,
      "step": 82150
    },
    {
      "entropy": 1.878348208218813,
      "epoch": 0.254688811408298,
      "grad_norm": 3.9120850563049316,
      "learning_rate": 5.012915559973233e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8392788946628571,
      "num_tokens": 98777263.0,
      "step": 82160
    },
    {
      "entropy": 2.001929074525833,
      "epoch": 0.2547198105333477,
      "grad_norm": 8.413508415222168,
      "learning_rate": 5.012610513782595e-06,
      "loss": 0.5951,
      "mean_token_accuracy": 0.8199850216507911,
      "num_tokens": 98788877.0,
      "step": 82170
    },
    {
      "entropy": 1.9609450832009316,
      "epoch": 0.2547508096583974,
      "grad_norm": 9.111734390258789,
      "learning_rate": 5.01230552327324e-06,
      "loss": 0.4973,
      "mean_token_accuracy": 0.844335462152958,
      "num_tokens": 98800367.0,
      "step": 82180
    },
    {
      "entropy": 1.9912359118461609,
      "epoch": 0.2547818087834471,
      "grad_norm": 9.116207122802734,
      "learning_rate": 5.012000588428227e-06,
      "loss": 0.5586,
      "mean_token_accuracy": 0.8303110048174858,
      "num_tokens": 98811373.0,
      "step": 82190
    },
    {
      "entropy": 1.8973192408680917,
      "epoch": 0.2548128079084968,
      "grad_norm": 8.382734298706055,
      "learning_rate": 5.01169570923063e-06,
      "loss": 0.441,
      "mean_token_accuracy": 0.8509496867656707,
      "num_tokens": 98823915.0,
      "step": 82200
    },
    {
      "entropy": 1.888469786942005,
      "epoch": 0.2548438070335465,
      "grad_norm": 4.078553676605225,
      "learning_rate": 5.011390885663524e-06,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8516004219651222,
      "num_tokens": 98835873.0,
      "step": 82210
    },
    {
      "entropy": 1.9228701919317246,
      "epoch": 0.2548748061585962,
      "grad_norm": 9.359138488769531,
      "learning_rate": 5.011086117709992e-06,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8420409306883812,
      "num_tokens": 98846828.0,
      "step": 82220
    },
    {
      "entropy": 1.9370484218001365,
      "epoch": 0.25490580528364587,
      "grad_norm": 9.458020210266113,
      "learning_rate": 5.010781405353129e-06,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8457821294665336,
      "num_tokens": 98858261.0,
      "step": 82230
    },
    {
      "entropy": 1.815031287074089,
      "epoch": 0.25493680440869554,
      "grad_norm": 7.948138236999512,
      "learning_rate": 5.010476748576029e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8458429276943207,
      "num_tokens": 98870539.0,
      "step": 82240
    },
    {
      "entropy": 1.9159539863467216,
      "epoch": 0.25496780353374526,
      "grad_norm": 4.097768306732178,
      "learning_rate": 5.010172147361801e-06,
      "loss": 0.542,
      "mean_token_accuracy": 0.8348632201552391,
      "num_tokens": 98882616.0,
      "step": 82250
    },
    {
      "entropy": 1.8219385787844657,
      "epoch": 0.25499880265879493,
      "grad_norm": 4.100628852844238,
      "learning_rate": 5.009867601693556e-06,
      "loss": 0.4851,
      "mean_token_accuracy": 0.8387814164161682,
      "num_tokens": 98895560.0,
      "step": 82260
    },
    {
      "entropy": 1.8934132039546967,
      "epoch": 0.25502980178384466,
      "grad_norm": 9.86856746673584,
      "learning_rate": 5.009563111554415e-06,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8422671303153038,
      "num_tokens": 98906966.0,
      "step": 82270
    },
    {
      "entropy": 1.9088118925690651,
      "epoch": 0.2550608009088943,
      "grad_norm": 4.092050075531006,
      "learning_rate": 5.009258676927506e-06,
      "loss": 0.4658,
      "mean_token_accuracy": 0.845549589395523,
      "num_tokens": 98918825.0,
      "step": 82280
    },
    {
      "entropy": 1.9305264785885812,
      "epoch": 0.25509180003394405,
      "grad_norm": 9.015393257141113,
      "learning_rate": 5.008954297795962e-06,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8419323205947876,
      "num_tokens": 98930506.0,
      "step": 82290
    },
    {
      "entropy": 1.9115617856383325,
      "epoch": 0.2551227991589937,
      "grad_norm": 9.485702514648438,
      "learning_rate": 5.008649974142926e-06,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8377510070800781,
      "num_tokens": 98941230.0,
      "step": 82300
    },
    {
      "entropy": 1.8585439771413803,
      "epoch": 0.25515379828404344,
      "grad_norm": 8.97275447845459,
      "learning_rate": 5.0083457059515476e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8494518890976905,
      "num_tokens": 98953271.0,
      "step": 82310
    },
    {
      "entropy": 1.781735722720623,
      "epoch": 0.2551847974090931,
      "grad_norm": 3.6371991634368896,
      "learning_rate": 5.008041493204978e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8600785121321678,
      "num_tokens": 98966568.0,
      "step": 82320
    },
    {
      "entropy": 1.9445871606469154,
      "epoch": 0.25521579653414284,
      "grad_norm": 7.9240899085998535,
      "learning_rate": 5.007737335886387e-06,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8479147583246232,
      "num_tokens": 98977867.0,
      "step": 82330
    },
    {
      "entropy": 1.9001877725124359,
      "epoch": 0.2552467956591925,
      "grad_norm": 9.06245231628418,
      "learning_rate": 5.007433233978939e-06,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8466285720467568,
      "num_tokens": 98990050.0,
      "step": 82340
    },
    {
      "entropy": 1.96756109893322,
      "epoch": 0.25527779478424223,
      "grad_norm": 7.101724624633789,
      "learning_rate": 5.007129187465815e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8483997210860252,
      "num_tokens": 99001143.0,
      "step": 82350
    },
    {
      "entropy": 1.9655298113822937,
      "epoch": 0.2553087939092919,
      "grad_norm": 11.712712287902832,
      "learning_rate": 5.006825196330199e-06,
      "loss": 0.5492,
      "mean_token_accuracy": 0.8361737057566643,
      "num_tokens": 99012612.0,
      "step": 82360
    },
    {
      "entropy": 1.8864617615938186,
      "epoch": 0.2553397930343416,
      "grad_norm": 9.521370887756348,
      "learning_rate": 5.006521260555282e-06,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8489149019122124,
      "num_tokens": 99024254.0,
      "step": 82370
    },
    {
      "entropy": 1.9336427330970765,
      "epoch": 0.2553707921593913,
      "grad_norm": 8.10055923461914,
      "learning_rate": 5.006217380124263e-06,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8430611327290535,
      "num_tokens": 99035921.0,
      "step": 82380
    },
    {
      "entropy": 1.8473542481660843,
      "epoch": 0.255401791284441,
      "grad_norm": 3.776082992553711,
      "learning_rate": 5.005913555020348e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8381215184926987,
      "num_tokens": 99048817.0,
      "step": 82390
    },
    {
      "entropy": 1.9615378364920617,
      "epoch": 0.2554327904094907,
      "grad_norm": 4.9302215576171875,
      "learning_rate": 5.00560978522675e-06,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8320355072617531,
      "num_tokens": 99060522.0,
      "step": 82400
    },
    {
      "entropy": 1.8505050733685493,
      "epoch": 0.2554637895345404,
      "grad_norm": 10.341068267822266,
      "learning_rate": 5.0053060707266894e-06,
      "loss": 0.4186,
      "mean_token_accuracy": 0.852343961596489,
      "num_tokens": 99073004.0,
      "step": 82410
    },
    {
      "entropy": 1.918088473379612,
      "epoch": 0.2554947886595901,
      "grad_norm": 7.087062835693359,
      "learning_rate": 5.005002411503397e-06,
      "loss": 0.4884,
      "mean_token_accuracy": 0.836826567351818,
      "num_tokens": 99084688.0,
      "step": 82420
    },
    {
      "entropy": 1.8801950231194495,
      "epoch": 0.2555257877846398,
      "grad_norm": 8.337331771850586,
      "learning_rate": 5.004698807540101e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8451901748776436,
      "num_tokens": 99097086.0,
      "step": 82430
    },
    {
      "entropy": 1.9792321264743804,
      "epoch": 0.25555678690968947,
      "grad_norm": 9.403802871704102,
      "learning_rate": 5.004395258820048e-06,
      "loss": 0.5651,
      "mean_token_accuracy": 0.8320185244083405,
      "num_tokens": 99108547.0,
      "step": 82440
    },
    {
      "entropy": 1.88391984552145,
      "epoch": 0.2555877860347392,
      "grad_norm": 7.921189308166504,
      "learning_rate": 5.004091765326484e-06,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8392195641994477,
      "num_tokens": 99121168.0,
      "step": 82450
    },
    {
      "entropy": 1.936862115561962,
      "epoch": 0.25561878515978886,
      "grad_norm": 8.814833641052246,
      "learning_rate": 5.003788327042666e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.842677928507328,
      "num_tokens": 99133107.0,
      "step": 82460
    },
    {
      "entropy": 1.900815899670124,
      "epoch": 0.25564978428483853,
      "grad_norm": 8.44135570526123,
      "learning_rate": 5.003484943951856e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8460040614008904,
      "num_tokens": 99145316.0,
      "step": 82470
    },
    {
      "entropy": 1.9658162996172905,
      "epoch": 0.25568078340988826,
      "grad_norm": 9.297836303710938,
      "learning_rate": 5.003181616037326e-06,
      "loss": 0.5277,
      "mean_token_accuracy": 0.842269380390644,
      "num_tokens": 99157003.0,
      "step": 82480
    },
    {
      "entropy": 1.8745719879865645,
      "epoch": 0.2557117825349379,
      "grad_norm": 6.9280805587768555,
      "learning_rate": 5.002878343282352e-06,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8584736675024033,
      "num_tokens": 99168906.0,
      "step": 82490
    },
    {
      "entropy": 1.8955281019210815,
      "epoch": 0.25574278165998765,
      "grad_norm": 4.146218299865723,
      "learning_rate": 5.0025751256702195e-06,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8466937229037285,
      "num_tokens": 99181020.0,
      "step": 82500
    },
    {
      "entropy": 1.9218576028943062,
      "epoch": 0.2557737807850373,
      "grad_norm": 7.757521629333496,
      "learning_rate": 5.0022719631842165e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8493802517652511,
      "num_tokens": 99192352.0,
      "step": 82510
    },
    {
      "entropy": 1.9906231686472893,
      "epoch": 0.25580477991008704,
      "grad_norm": 8.75942611694336,
      "learning_rate": 5.001968855807645e-06,
      "loss": 0.5567,
      "mean_token_accuracy": 0.8316405296325684,
      "num_tokens": 99203591.0,
      "step": 82520
    },
    {
      "entropy": 1.9479998797178268,
      "epoch": 0.2558357790351367,
      "grad_norm": 8.903733253479004,
      "learning_rate": 5.001665803523808e-06,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8445803746581078,
      "num_tokens": 99214411.0,
      "step": 82530
    },
    {
      "entropy": 1.920227263867855,
      "epoch": 0.25586677816018644,
      "grad_norm": 8.966920852661133,
      "learning_rate": 5.001362806316021e-06,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8462096214294433,
      "num_tokens": 99226223.0,
      "step": 82540
    },
    {
      "entropy": 1.914782066643238,
      "epoch": 0.2558977772852361,
      "grad_norm": 9.925724983215332,
      "learning_rate": 5.001059864167602e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8446723312139511,
      "num_tokens": 99238079.0,
      "step": 82550
    },
    {
      "entropy": 1.9496339425444602,
      "epoch": 0.25592877641028583,
      "grad_norm": 7.7270612716674805,
      "learning_rate": 5.000756977061877e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8507271915674209,
      "num_tokens": 99249265.0,
      "step": 82560
    },
    {
      "entropy": 1.9208646342158318,
      "epoch": 0.2559597755353355,
      "grad_norm": 7.940464019775391,
      "learning_rate": 5.000454144982181e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8384231016039848,
      "num_tokens": 99261128.0,
      "step": 82570
    },
    {
      "entropy": 1.9947969660162925,
      "epoch": 0.2559907746603852,
      "grad_norm": 8.898752212524414,
      "learning_rate": 5.000151367911854e-06,
      "loss": 0.5323,
      "mean_token_accuracy": 0.8397686898708343,
      "num_tokens": 99273022.0,
      "step": 82580
    },
    {
      "entropy": 2.0049765020608903,
      "epoch": 0.2560217737854349,
      "grad_norm": 8.215011596679688,
      "learning_rate": 4.999848645834245e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8358174741268158,
      "num_tokens": 99283993.0,
      "step": 82590
    },
    {
      "entropy": 1.9047711238265037,
      "epoch": 0.2560527729104846,
      "grad_norm": 7.696551322937012,
      "learning_rate": 4.999545978732709e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8554992079734802,
      "num_tokens": 99295756.0,
      "step": 82600
    },
    {
      "entropy": 1.7249569281935693,
      "epoch": 0.2560837720355343,
      "grad_norm": 3.8466684818267822,
      "learning_rate": 4.999243366590607e-06,
      "loss": 0.3453,
      "mean_token_accuracy": 0.8684180244803429,
      "num_tokens": 99310148.0,
      "step": 82610
    },
    {
      "entropy": 1.8993670761585235,
      "epoch": 0.256114771160584,
      "grad_norm": 4.150681972503662,
      "learning_rate": 4.998940809391308e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8339141398668289,
      "num_tokens": 99322579.0,
      "step": 82620
    },
    {
      "entropy": 1.8982570886611938,
      "epoch": 0.2561457702856337,
      "grad_norm": 7.722603797912598,
      "learning_rate": 4.998638307118189e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8633927822113037,
      "num_tokens": 99334404.0,
      "step": 82630
    },
    {
      "entropy": 1.8842903196811676,
      "epoch": 0.2561767694106834,
      "grad_norm": 8.307390213012695,
      "learning_rate": 4.998335859754634e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8525113880634307,
      "num_tokens": 99347056.0,
      "step": 82640
    },
    {
      "entropy": 1.9822999000549317,
      "epoch": 0.25620776853573307,
      "grad_norm": 3.5337090492248535,
      "learning_rate": 4.998033467284031e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8361595958471298,
      "num_tokens": 99359039.0,
      "step": 82650
    },
    {
      "entropy": 1.8225185006856919,
      "epoch": 0.2562387676607828,
      "grad_norm": 8.926518440246582,
      "learning_rate": 4.997731129689778e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8412432596087456,
      "num_tokens": 99373001.0,
      "step": 82660
    },
    {
      "entropy": 1.9116671606898308,
      "epoch": 0.25626976678583246,
      "grad_norm": 9.220154762268066,
      "learning_rate": 4.997428846955282e-06,
      "loss": 0.6017,
      "mean_token_accuracy": 0.8289077386260033,
      "num_tokens": 99385753.0,
      "step": 82670
    },
    {
      "entropy": 1.981253059208393,
      "epoch": 0.2563007659108822,
      "grad_norm": 9.764948844909668,
      "learning_rate": 4.99712661906395e-06,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8350262567400932,
      "num_tokens": 99396793.0,
      "step": 82680
    },
    {
      "entropy": 1.922914719581604,
      "epoch": 0.25633176503593186,
      "grad_norm": 9.412257194519043,
      "learning_rate": 4.996824445999205e-06,
      "loss": 0.493,
      "mean_token_accuracy": 0.8538893148303032,
      "num_tokens": 99408399.0,
      "step": 82690
    },
    {
      "entropy": 1.9097683414816857,
      "epoch": 0.2563627641609816,
      "grad_norm": 3.221794366836548,
      "learning_rate": 4.996522327744468e-06,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8507839366793633,
      "num_tokens": 99420256.0,
      "step": 82700
    },
    {
      "entropy": 1.8777303710579871,
      "epoch": 0.25639376328603125,
      "grad_norm": 8.001666069030762,
      "learning_rate": 4.996220264283173e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8601910769939423,
      "num_tokens": 99433400.0,
      "step": 82710
    },
    {
      "entropy": 1.9291005671024322,
      "epoch": 0.2564247624110809,
      "grad_norm": 10.794572830200195,
      "learning_rate": 4.99591825559876e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8337120041251183,
      "num_tokens": 99445875.0,
      "step": 82720
    },
    {
      "entropy": 1.8854432210326195,
      "epoch": 0.25645576153613064,
      "grad_norm": 3.7990822792053223,
      "learning_rate": 4.995616301674676e-06,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8595590248703957,
      "num_tokens": 99458264.0,
      "step": 82730
    },
    {
      "entropy": 1.9592019245028496,
      "epoch": 0.2564867606611803,
      "grad_norm": 8.916022300720215,
      "learning_rate": 4.995314402494373e-06,
      "loss": 0.5263,
      "mean_token_accuracy": 0.8408660888671875,
      "num_tokens": 99469951.0,
      "step": 82740
    },
    {
      "entropy": 1.9431997925043105,
      "epoch": 0.25651775978623004,
      "grad_norm": 6.906679153442383,
      "learning_rate": 4.99501255804131e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8524601250886917,
      "num_tokens": 99482201.0,
      "step": 82750
    },
    {
      "entropy": 1.91104666441679,
      "epoch": 0.2565487589112797,
      "grad_norm": 7.423837184906006,
      "learning_rate": 4.9947107682989585e-06,
      "loss": 0.4479,
      "mean_token_accuracy": 0.8510598838329315,
      "num_tokens": 99494427.0,
      "step": 82760
    },
    {
      "entropy": 1.9561400279402732,
      "epoch": 0.25657975803632943,
      "grad_norm": 9.361892700195312,
      "learning_rate": 4.994409033250789e-06,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8367945715785027,
      "num_tokens": 99506187.0,
      "step": 82770
    },
    {
      "entropy": 1.9289663940668107,
      "epoch": 0.2566107571613791,
      "grad_norm": 8.348969459533691,
      "learning_rate": 4.9941073528802855e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8491498351097106,
      "num_tokens": 99518185.0,
      "step": 82780
    },
    {
      "entropy": 1.9189318656921386,
      "epoch": 0.2566417562864288,
      "grad_norm": 10.50049877166748,
      "learning_rate": 4.993805727170934e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8476282939314842,
      "num_tokens": 99531191.0,
      "step": 82790
    },
    {
      "entropy": 1.9094542279839515,
      "epoch": 0.2566727554114785,
      "grad_norm": 8.101419448852539,
      "learning_rate": 4.993504156106232e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8485571637749671,
      "num_tokens": 99542937.0,
      "step": 82800
    },
    {
      "entropy": 1.880283573269844,
      "epoch": 0.2567037545365282,
      "grad_norm": 7.880818843841553,
      "learning_rate": 4.99320263966968e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8481839671730995,
      "num_tokens": 99555383.0,
      "step": 82810
    },
    {
      "entropy": 1.8960697516798972,
      "epoch": 0.2567347536615779,
      "grad_norm": Infinity,
      "learning_rate": 4.992901177844789e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.857147465646267,
      "num_tokens": 99567485.0,
      "step": 82820
    },
    {
      "entropy": 1.992298111319542,
      "epoch": 0.2567657527866276,
      "grad_norm": 7.422923564910889,
      "learning_rate": 4.992599770615074e-06,
      "loss": 0.492,
      "mean_token_accuracy": 0.8521854832768441,
      "num_tokens": 99578298.0,
      "step": 82830
    },
    {
      "entropy": 1.9098379969596864,
      "epoch": 0.2567967519116773,
      "grad_norm": 7.3638739585876465,
      "learning_rate": 4.992298417964059e-06,
      "loss": 0.4517,
      "mean_token_accuracy": 0.8458767265081406,
      "num_tokens": 99590734.0,
      "step": 82840
    },
    {
      "entropy": 1.8104907512664794,
      "epoch": 0.256827751036727,
      "grad_norm": 4.290040493011475,
      "learning_rate": 4.991997119875274e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8509851738810539,
      "num_tokens": 99603863.0,
      "step": 82850
    },
    {
      "entropy": 1.9496165543794632,
      "epoch": 0.25685875016177667,
      "grad_norm": 8.491985321044922,
      "learning_rate": 4.991695876332256e-06,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8299289435148239,
      "num_tokens": 99614829.0,
      "step": 82860
    },
    {
      "entropy": 1.7824837416410446,
      "epoch": 0.2568897492868264,
      "grad_norm": 10.624833106994629,
      "learning_rate": 4.991394687318549e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8593706652522087,
      "num_tokens": 99627639.0,
      "step": 82870
    },
    {
      "entropy": 1.8935058623552323,
      "epoch": 0.25692074841187607,
      "grad_norm": 3.4996988773345947,
      "learning_rate": 4.9910935528177034e-06,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8467808067798615,
      "num_tokens": 99639515.0,
      "step": 82880
    },
    {
      "entropy": 1.9544168338179588,
      "epoch": 0.2569517475369258,
      "grad_norm": 8.53878402709961,
      "learning_rate": 4.990792472813278e-06,
      "loss": 0.4862,
      "mean_token_accuracy": 0.846103484928608,
      "num_tokens": 99651187.0,
      "step": 82890
    },
    {
      "entropy": 1.9207098990678788,
      "epoch": 0.25698274666197546,
      "grad_norm": 9.49964427947998,
      "learning_rate": 4.990491447288838e-06,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8317215725779533,
      "num_tokens": 99662564.0,
      "step": 82900
    },
    {
      "entropy": 1.876941241323948,
      "epoch": 0.2570137457870252,
      "grad_norm": 7.434007167816162,
      "learning_rate": 4.990190476227954e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8501386985182762,
      "num_tokens": 99674973.0,
      "step": 82910
    },
    {
      "entropy": 1.9188931852579116,
      "epoch": 0.25704474491207485,
      "grad_norm": 7.2484331130981445,
      "learning_rate": 4.989889559614206e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8477975860238075,
      "num_tokens": 99686447.0,
      "step": 82920
    },
    {
      "entropy": 1.8817449413239955,
      "epoch": 0.2570757440371246,
      "grad_norm": 4.548129081726074,
      "learning_rate": 4.98958869743118e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8410625770688057,
      "num_tokens": 99698742.0,
      "step": 82930
    },
    {
      "entropy": 1.8077333301305771,
      "epoch": 0.25710674316217424,
      "grad_norm": 8.342117309570312,
      "learning_rate": 4.989287889662468e-06,
      "loss": 0.3919,
      "mean_token_accuracy": 0.8545293301343918,
      "num_tokens": 99712362.0,
      "step": 82940
    },
    {
      "entropy": 1.9139455810189248,
      "epoch": 0.25713774228722397,
      "grad_norm": 5.589305400848389,
      "learning_rate": 4.988987136291668e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8507546544075012,
      "num_tokens": 99724409.0,
      "step": 82950
    },
    {
      "entropy": 1.8849102705717087,
      "epoch": 0.25716874141227364,
      "grad_norm": 8.964681625366211,
      "learning_rate": 4.98868643730239e-06,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8342619195580483,
      "num_tokens": 99736094.0,
      "step": 82960
    },
    {
      "entropy": 1.9025137901306153,
      "epoch": 0.2571997405373233,
      "grad_norm": 4.793386936187744,
      "learning_rate": 4.9883857926782435e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8469575837254524,
      "num_tokens": 99748673.0,
      "step": 82970
    },
    {
      "entropy": 1.9453041523694992,
      "epoch": 0.25723073966237303,
      "grad_norm": 6.930491924285889,
      "learning_rate": 4.988085202402852e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8533410027623176,
      "num_tokens": 99760599.0,
      "step": 82980
    },
    {
      "entropy": 1.9440397009253503,
      "epoch": 0.2572617387874227,
      "grad_norm": 10.45510196685791,
      "learning_rate": 4.987784666459842e-06,
      "loss": 0.5538,
      "mean_token_accuracy": 0.8280631914734841,
      "num_tokens": 99771974.0,
      "step": 82990
    },
    {
      "entropy": 1.9070080369710922,
      "epoch": 0.2572927379124724,
      "grad_norm": 10.232966423034668,
      "learning_rate": 4.987484184832846e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8449847355484963,
      "num_tokens": 99784534.0,
      "step": 83000
    },
    {
      "entropy": 1.9375251710414887,
      "epoch": 0.2573237370375221,
      "grad_norm": 3.9058806896209717,
      "learning_rate": 4.9871837575055064e-06,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8462153524160385,
      "num_tokens": 99796298.0,
      "step": 83010
    },
    {
      "entropy": 1.9144327610731124,
      "epoch": 0.2573547361625718,
      "grad_norm": 10.11092472076416,
      "learning_rate": 4.9868833844614715e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8435170888900757,
      "num_tokens": 99807740.0,
      "step": 83020
    },
    {
      "entropy": 1.8731989122927188,
      "epoch": 0.2573857352876215,
      "grad_norm": 7.537945747375488,
      "learning_rate": 4.986583065684396e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8457079976797104,
      "num_tokens": 99819888.0,
      "step": 83030
    },
    {
      "entropy": 1.9202357351779937,
      "epoch": 0.2574167344126712,
      "grad_norm": 8.85262393951416,
      "learning_rate": 4.986282801157941e-06,
      "loss": 0.4819,
      "mean_token_accuracy": 0.844878327846527,
      "num_tokens": 99831444.0,
      "step": 83040
    },
    {
      "entropy": 1.8694255024194717,
      "epoch": 0.2574477335377209,
      "grad_norm": 5.111952304840088,
      "learning_rate": 4.9859825908657735e-06,
      "loss": 0.512,
      "mean_token_accuracy": 0.835330268740654,
      "num_tokens": 99844184.0,
      "step": 83050
    },
    {
      "entropy": 1.9209684014320374,
      "epoch": 0.2574787326627706,
      "grad_norm": 6.903238773345947,
      "learning_rate": 4.985682434791573e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.845658715069294,
      "num_tokens": 99856287.0,
      "step": 83060
    },
    {
      "entropy": 1.8327378258109093,
      "epoch": 0.2575097317878203,
      "grad_norm": 9.568243026733398,
      "learning_rate": 4.9853823329190185e-06,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8382641449570656,
      "num_tokens": 99869468.0,
      "step": 83070
    },
    {
      "entropy": 1.9063044503331183,
      "epoch": 0.25754073091287,
      "grad_norm": 3.617058515548706,
      "learning_rate": 4.985082285231801e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8415888145565986,
      "num_tokens": 99881146.0,
      "step": 83080
    },
    {
      "entropy": 1.7789159893989563,
      "epoch": 0.25757173003791967,
      "grad_norm": 7.718278884887695,
      "learning_rate": 4.9847822917136154e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8528930768370628,
      "num_tokens": 99894684.0,
      "step": 83090
    },
    {
      "entropy": 1.946078921854496,
      "epoch": 0.2576027291629694,
      "grad_norm": 8.70456600189209,
      "learning_rate": 4.984482352348164e-06,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8418860971927643,
      "num_tokens": 99905660.0,
      "step": 83100
    },
    {
      "entropy": 1.9768987134099008,
      "epoch": 0.25763372828801906,
      "grad_norm": 9.16089153289795,
      "learning_rate": 4.9841824671191594e-06,
      "loss": 0.5825,
      "mean_token_accuracy": 0.8343336895108223,
      "num_tokens": 99917120.0,
      "step": 83110
    },
    {
      "entropy": 1.889198412001133,
      "epoch": 0.2576647274130688,
      "grad_norm": 7.599856853485107,
      "learning_rate": 4.983882636010317e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8526261046528816,
      "num_tokens": 99928830.0,
      "step": 83120
    },
    {
      "entropy": 1.9562062606215478,
      "epoch": 0.25769572653811845,
      "grad_norm": 8.07655143737793,
      "learning_rate": 4.983582859005359e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.831449817121029,
      "num_tokens": 99940719.0,
      "step": 83130
    },
    {
      "entropy": 1.896943534910679,
      "epoch": 0.2577267256631682,
      "grad_norm": 8.602059364318848,
      "learning_rate": 4.983283136088018e-06,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8445844024419784,
      "num_tokens": 99953117.0,
      "step": 83140
    },
    {
      "entropy": 1.9443399429321289,
      "epoch": 0.25775772478821785,
      "grad_norm": 7.0254693031311035,
      "learning_rate": 4.982983467242029e-06,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8422896787524223,
      "num_tokens": 99964325.0,
      "step": 83150
    },
    {
      "entropy": 1.9491709470748901,
      "epoch": 0.25778872391326757,
      "grad_norm": 11.82594108581543,
      "learning_rate": 4.982683852451138e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8369784370064736,
      "num_tokens": 99975197.0,
      "step": 83160
    },
    {
      "entropy": 1.8532451003789903,
      "epoch": 0.25781972303831724,
      "grad_norm": 3.9568448066711426,
      "learning_rate": 4.982384291699096e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8396353736519814,
      "num_tokens": 99987783.0,
      "step": 83170
    },
    {
      "entropy": 1.9464505165815353,
      "epoch": 0.25785072216336696,
      "grad_norm": 10.583782196044922,
      "learning_rate": 4.982084784969659e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.8377991825342178,
      "num_tokens": 100000020.0,
      "step": 83180
    },
    {
      "entropy": 1.9166669502854348,
      "epoch": 0.25788172128841663,
      "grad_norm": 9.20700740814209,
      "learning_rate": 4.981785332246592e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8458949193358422,
      "num_tokens": 100011762.0,
      "step": 83190
    },
    {
      "entropy": 1.9713429033756256,
      "epoch": 0.25791272041346636,
      "grad_norm": 9.420412063598633,
      "learning_rate": 4.981485933513668e-06,
      "loss": 0.4776,
      "mean_token_accuracy": 0.8544860139489174,
      "num_tokens": 100023022.0,
      "step": 83200
    },
    {
      "entropy": 1.9141755923628807,
      "epoch": 0.257943719538516,
      "grad_norm": 9.200664520263672,
      "learning_rate": 4.981186588754664e-06,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8375491261482239,
      "num_tokens": 100034493.0,
      "step": 83210
    },
    {
      "entropy": 1.8373278394341468,
      "epoch": 0.2579747186635657,
      "grad_norm": 3.6070780754089355,
      "learning_rate": 4.980887297953366e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.853793577849865,
      "num_tokens": 100047907.0,
      "step": 83220
    },
    {
      "entropy": 1.905658522248268,
      "epoch": 0.2580057177886154,
      "grad_norm": 33.18852996826172,
      "learning_rate": 4.980588061093565e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.844824655354023,
      "num_tokens": 100059902.0,
      "step": 83230
    },
    {
      "entropy": 1.9518269658088685,
      "epoch": 0.2580367169136651,
      "grad_norm": 7.96235990524292,
      "learning_rate": 4.980288878159059e-06,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8474616572260857,
      "num_tokens": 100070453.0,
      "step": 83240
    },
    {
      "entropy": 1.971319329738617,
      "epoch": 0.2580677160387148,
      "grad_norm": 8.494059562683105,
      "learning_rate": 4.9799897491336564e-06,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8315307438373566,
      "num_tokens": 100082070.0,
      "step": 83250
    },
    {
      "entropy": 1.9625032275915146,
      "epoch": 0.2580987151637645,
      "grad_norm": 7.6323466300964355,
      "learning_rate": 4.979690674001167e-06,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8371849060058594,
      "num_tokens": 100093397.0,
      "step": 83260
    },
    {
      "entropy": 1.8805483281612396,
      "epoch": 0.2581297142888142,
      "grad_norm": 8.261248588562012,
      "learning_rate": 4.979391652745411e-06,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8561758771538734,
      "num_tokens": 100105997.0,
      "step": 83270
    },
    {
      "entropy": 1.8767956405878068,
      "epoch": 0.2581607134138639,
      "grad_norm": 3.6510441303253174,
      "learning_rate": 4.9790926853502125e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8472672030329704,
      "num_tokens": 100117793.0,
      "step": 83280
    },
    {
      "entropy": 1.7951404377818108,
      "epoch": 0.2581917125389136,
      "grad_norm": 4.375608921051025,
      "learning_rate": 4.978793771799407e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.8583429366350174,
      "num_tokens": 100130835.0,
      "step": 83290
    },
    {
      "entropy": 1.971182020008564,
      "epoch": 0.25822271166396327,
      "grad_norm": 10.078766822814941,
      "learning_rate": 4.978494912076833e-06,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8344864457845688,
      "num_tokens": 100141909.0,
      "step": 83300
    },
    {
      "entropy": 1.9942039638757705,
      "epoch": 0.258253710789013,
      "grad_norm": 8.207865715026855,
      "learning_rate": 4.978196106166336e-06,
      "loss": 0.574,
      "mean_token_accuracy": 0.828905712068081,
      "num_tokens": 100152944.0,
      "step": 83310
    },
    {
      "entropy": 1.8606795385479926,
      "epoch": 0.25828470991406266,
      "grad_norm": 4.853648662567139,
      "learning_rate": 4.97789735405177e-06,
      "loss": 0.4582,
      "mean_token_accuracy": 0.849544158577919,
      "num_tokens": 100166124.0,
      "step": 83320
    },
    {
      "entropy": 1.9509618058800697,
      "epoch": 0.2583157090391124,
      "grad_norm": 6.577378273010254,
      "learning_rate": 4.977598655716994e-06,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8365138009190559,
      "num_tokens": 100177997.0,
      "step": 83330
    },
    {
      "entropy": 1.9413925766944886,
      "epoch": 0.25834670816416205,
      "grad_norm": 7.664201736450195,
      "learning_rate": 4.977300011145877e-06,
      "loss": 0.497,
      "mean_token_accuracy": 0.8374240696430206,
      "num_tokens": 100189432.0,
      "step": 83340
    },
    {
      "entropy": 1.9520146340131759,
      "epoch": 0.2583777072892118,
      "grad_norm": 3.7987093925476074,
      "learning_rate": 4.97700142032229e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8407413721084595,
      "num_tokens": 100201276.0,
      "step": 83350
    },
    {
      "entropy": 1.9646510049700736,
      "epoch": 0.25840870641426145,
      "grad_norm": 8.533371925354004,
      "learning_rate": 4.976702883230112e-06,
      "loss": 0.5696,
      "mean_token_accuracy": 0.8276545464992523,
      "num_tokens": 100212137.0,
      "step": 83360
    },
    {
      "entropy": 1.8566021844744682,
      "epoch": 0.25843970553931117,
      "grad_norm": 9.078433990478516,
      "learning_rate": 4.976404399853234e-06,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8363482743501663,
      "num_tokens": 100224796.0,
      "step": 83370
    },
    {
      "entropy": 1.9266142144799232,
      "epoch": 0.25847070466436084,
      "grad_norm": 8.397682189941406,
      "learning_rate": 4.9761059701755475e-06,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8483969137072563,
      "num_tokens": 100236575.0,
      "step": 83380
    },
    {
      "entropy": 1.9420603141188622,
      "epoch": 0.25850170378941056,
      "grad_norm": 8.181031227111816,
      "learning_rate": 4.975807594180953e-06,
      "loss": 0.5213,
      "mean_token_accuracy": 0.844786761701107,
      "num_tokens": 100248133.0,
      "step": 83390
    },
    {
      "entropy": 1.7950069807469844,
      "epoch": 0.25853270291446023,
      "grad_norm": 7.972817420959473,
      "learning_rate": 4.975509271853358e-06,
      "loss": 0.3808,
      "mean_token_accuracy": 0.8578445136547088,
      "num_tokens": 100261552.0,
      "step": 83400
    },
    {
      "entropy": 1.9505249321460725,
      "epoch": 0.25856370203950996,
      "grad_norm": 3.8303351402282715,
      "learning_rate": 4.975211003176676e-06,
      "loss": 0.5553,
      "mean_token_accuracy": 0.8300097927451133,
      "num_tokens": 100272407.0,
      "step": 83410
    },
    {
      "entropy": 1.9518933594226837,
      "epoch": 0.2585947011645596,
      "grad_norm": 8.778966903686523,
      "learning_rate": 4.97491278813483e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8443615302443505,
      "num_tokens": 100284483.0,
      "step": 83420
    },
    {
      "entropy": 1.825792530924082,
      "epoch": 0.25862570028960935,
      "grad_norm": 8.191307067871094,
      "learning_rate": 4.974614626711745e-06,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8553913697600365,
      "num_tokens": 100298331.0,
      "step": 83430
    },
    {
      "entropy": 1.9290891095995903,
      "epoch": 0.258656699414659,
      "grad_norm": 8.639996528625488,
      "learning_rate": 4.9743165188913564e-06,
      "loss": 0.5654,
      "mean_token_accuracy": 0.8340361103415489,
      "num_tokens": 100309563.0,
      "step": 83440
    },
    {
      "entropy": 1.8901295900344848,
      "epoch": 0.25868769853970874,
      "grad_norm": 4.191493988037109,
      "learning_rate": 4.9740184646576036e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8442709073424339,
      "num_tokens": 100321894.0,
      "step": 83450
    },
    {
      "entropy": 1.8269127234816551,
      "epoch": 0.2587186976647584,
      "grad_norm": 4.574268817901611,
      "learning_rate": 4.9737204639944376e-06,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8548773661255836,
      "num_tokens": 100334501.0,
      "step": 83460
    },
    {
      "entropy": 1.8532571867108345,
      "epoch": 0.2587496967898081,
      "grad_norm": 3.918750762939453,
      "learning_rate": 4.973422516885809e-06,
      "loss": 0.4792,
      "mean_token_accuracy": 0.846004131436348,
      "num_tokens": 100346656.0,
      "step": 83470
    },
    {
      "entropy": 1.8245261393487453,
      "epoch": 0.2587806959148578,
      "grad_norm": 3.927574872970581,
      "learning_rate": 4.973124623315682e-06,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8500414654612541,
      "num_tokens": 100359699.0,
      "step": 83480
    },
    {
      "entropy": 1.77280533015728,
      "epoch": 0.2588116950399075,
      "grad_norm": 4.003964424133301,
      "learning_rate": 4.972826783268022e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8541926577687263,
      "num_tokens": 100373537.0,
      "step": 83490
    },
    {
      "entropy": 1.894656127691269,
      "epoch": 0.2588426941649572,
      "grad_norm": 8.42563247680664,
      "learning_rate": 4.972528996726807e-06,
      "loss": 0.494,
      "mean_token_accuracy": 0.8406502559781075,
      "num_tokens": 100385609.0,
      "step": 83500
    },
    {
      "entropy": 1.9357464522123338,
      "epoch": 0.25887369329000687,
      "grad_norm": 9.4671049118042,
      "learning_rate": 4.972231263676015e-06,
      "loss": 0.496,
      "mean_token_accuracy": 0.8405605420470238,
      "num_tokens": 100397032.0,
      "step": 83510
    },
    {
      "entropy": 1.9452675521373748,
      "epoch": 0.2589046924150566,
      "grad_norm": 12.05567455291748,
      "learning_rate": 4.971933584099637e-06,
      "loss": 0.5118,
      "mean_token_accuracy": 0.832793453335762,
      "num_tokens": 100408610.0,
      "step": 83520
    },
    {
      "entropy": 1.9751826629042626,
      "epoch": 0.25893569154010626,
      "grad_norm": 8.384647369384766,
      "learning_rate": 4.971635957981665e-06,
      "loss": 0.513,
      "mean_token_accuracy": 0.8382089316844941,
      "num_tokens": 100420251.0,
      "step": 83530
    },
    {
      "entropy": 1.911265040934086,
      "epoch": 0.258966690665156,
      "grad_norm": 8.947986602783203,
      "learning_rate": 4.971338385306102e-06,
      "loss": 0.4612,
      "mean_token_accuracy": 0.853412701189518,
      "num_tokens": 100432233.0,
      "step": 83540
    },
    {
      "entropy": 1.9375739723443985,
      "epoch": 0.25899768979020565,
      "grad_norm": 7.377249240875244,
      "learning_rate": 4.9710408660569555e-06,
      "loss": 0.4935,
      "mean_token_accuracy": 0.8441775634884834,
      "num_tokens": 100443781.0,
      "step": 83550
    },
    {
      "entropy": 1.9460734203457832,
      "epoch": 0.2590286889152554,
      "grad_norm": 8.327649116516113,
      "learning_rate": 4.970743400218241e-06,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8340301439166069,
      "num_tokens": 100454515.0,
      "step": 83560
    },
    {
      "entropy": 1.9006448999047278,
      "epoch": 0.25905968804030505,
      "grad_norm": 4.181915760040283,
      "learning_rate": 4.9704459877739805e-06,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8439620837569237,
      "num_tokens": 100466532.0,
      "step": 83570
    },
    {
      "entropy": 2.0110969811677935,
      "epoch": 0.25909068716535477,
      "grad_norm": 7.133867263793945,
      "learning_rate": 4.9701486287082e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.84031240940094,
      "num_tokens": 100477071.0,
      "step": 83580
    },
    {
      "entropy": 1.8151050120592118,
      "epoch": 0.25912168629040444,
      "grad_norm": 4.122489929199219,
      "learning_rate": 4.9698513230049375e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8427058979868889,
      "num_tokens": 100490782.0,
      "step": 83590
    },
    {
      "entropy": 1.8606194391846658,
      "epoch": 0.25915268541545416,
      "grad_norm": 3.6981208324432373,
      "learning_rate": 4.969554070648234e-06,
      "loss": 0.456,
      "mean_token_accuracy": 0.8553460776805878,
      "num_tokens": 100503176.0,
      "step": 83600
    },
    {
      "entropy": 1.9358109071850778,
      "epoch": 0.25918368454050383,
      "grad_norm": 4.142949104309082,
      "learning_rate": 4.9692568716221355e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8340194925665856,
      "num_tokens": 100514798.0,
      "step": 83610
    },
    {
      "entropy": 1.8827938050031663,
      "epoch": 0.25921468366555356,
      "grad_norm": 8.898327827453613,
      "learning_rate": 4.968959725910699e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8450506746768951,
      "num_tokens": 100526985.0,
      "step": 83620
    },
    {
      "entropy": 1.9060676455497743,
      "epoch": 0.2592456827906032,
      "grad_norm": 9.307757377624512,
      "learning_rate": 4.968662633497986e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8441753014922142,
      "num_tokens": 100539529.0,
      "step": 83630
    },
    {
      "entropy": 1.8917318254709243,
      "epoch": 0.25927668191565295,
      "grad_norm": 7.30048942565918,
      "learning_rate": 4.968365594368065e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.849764208495617,
      "num_tokens": 100551756.0,
      "step": 83640
    },
    {
      "entropy": 1.8924952074885368,
      "epoch": 0.2593076810407026,
      "grad_norm": 4.036615371704102,
      "learning_rate": 4.968068608505009e-06,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8556442365050316,
      "num_tokens": 100564203.0,
      "step": 83650
    },
    {
      "entropy": 1.8785409778356552,
      "epoch": 0.25933868016575234,
      "grad_norm": 4.065882682800293,
      "learning_rate": 4.967771675892903e-06,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8419948130846023,
      "num_tokens": 100575948.0,
      "step": 83660
    },
    {
      "entropy": 1.92302625477314,
      "epoch": 0.259369679290802,
      "grad_norm": 5.085048675537109,
      "learning_rate": 4.967474796515832e-06,
      "loss": 0.54,
      "mean_token_accuracy": 0.8281594708561897,
      "num_tokens": 100588049.0,
      "step": 83670
    },
    {
      "entropy": 1.8481286302208901,
      "epoch": 0.25940067841585174,
      "grad_norm": 10.417789459228516,
      "learning_rate": 4.9671779703578934e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.848712483048439,
      "num_tokens": 100600632.0,
      "step": 83680
    },
    {
      "entropy": 1.8013351663947106,
      "epoch": 0.2594316775409014,
      "grad_norm": 8.973665237426758,
      "learning_rate": 4.966881197403189e-06,
      "loss": 0.4899,
      "mean_token_accuracy": 0.83788081407547,
      "num_tokens": 100613842.0,
      "step": 83690
    },
    {
      "entropy": 1.8483421131968498,
      "epoch": 0.25946267666595113,
      "grad_norm": 7.858611583709717,
      "learning_rate": 4.966584477635825e-06,
      "loss": 0.4966,
      "mean_token_accuracy": 0.8433775931596756,
      "num_tokens": 100626204.0,
      "step": 83700
    },
    {
      "entropy": 1.8308247201144696,
      "epoch": 0.2594936757910008,
      "grad_norm": 8.385130882263184,
      "learning_rate": 4.966287811039917e-06,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8558708354830742,
      "num_tokens": 100639013.0,
      "step": 83710
    },
    {
      "entropy": 1.9155432254076004,
      "epoch": 0.25952467491605047,
      "grad_norm": 9.558300971984863,
      "learning_rate": 4.965991197599587e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8412322223186492,
      "num_tokens": 100649836.0,
      "step": 83720
    },
    {
      "entropy": 1.8477103784680367,
      "epoch": 0.2595556740411002,
      "grad_norm": 3.852271556854248,
      "learning_rate": 4.965694637298964e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.848893304169178,
      "num_tokens": 100662993.0,
      "step": 83730
    },
    {
      "entropy": 1.9299596205353737,
      "epoch": 0.25958667316614986,
      "grad_norm": 8.642110824584961,
      "learning_rate": 4.9653981301221825e-06,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8376413077116013,
      "num_tokens": 100674572.0,
      "step": 83740
    },
    {
      "entropy": 1.8460013434290885,
      "epoch": 0.2596176722911996,
      "grad_norm": 10.720597267150879,
      "learning_rate": 4.9651016760533816e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8525181710720062,
      "num_tokens": 100687154.0,
      "step": 83750
    },
    {
      "entropy": 1.8346559152007103,
      "epoch": 0.25964867141624925,
      "grad_norm": 7.259960651397705,
      "learning_rate": 4.964805275076713e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8434622973203659,
      "num_tokens": 100699665.0,
      "step": 83760
    },
    {
      "entropy": 1.9489260196685791,
      "epoch": 0.259679670541299,
      "grad_norm": 8.808201789855957,
      "learning_rate": 4.96450892717633e-06,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8384017795324326,
      "num_tokens": 100710924.0,
      "step": 83770
    },
    {
      "entropy": 1.771421131491661,
      "epoch": 0.25971066966634865,
      "grad_norm": 7.782841682434082,
      "learning_rate": 4.964212632336392e-06,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8582013815641403,
      "num_tokens": 100724256.0,
      "step": 83780
    },
    {
      "entropy": 1.8205000385642052,
      "epoch": 0.2597416687913984,
      "grad_norm": 7.884888648986816,
      "learning_rate": 4.963916390541071e-06,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8426259756088257,
      "num_tokens": 100737066.0,
      "step": 83790
    },
    {
      "entropy": 1.9249374747276307,
      "epoch": 0.25977266791644804,
      "grad_norm": 8.629660606384277,
      "learning_rate": 4.963620201774537e-06,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8426753237843514,
      "num_tokens": 100748371.0,
      "step": 83800
    },
    {
      "entropy": 1.869993396103382,
      "epoch": 0.25980366704149777,
      "grad_norm": 7.948463439941406,
      "learning_rate": 4.963324066020974e-06,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8404388725757599,
      "num_tokens": 100760062.0,
      "step": 83810
    },
    {
      "entropy": 1.8810945719480514,
      "epoch": 0.25983466616654743,
      "grad_norm": 8.10258960723877,
      "learning_rate": 4.963027983264569e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.85190070271492,
      "num_tokens": 100771799.0,
      "step": 83820
    },
    {
      "entropy": 1.932461032271385,
      "epoch": 0.25986566529159716,
      "grad_norm": 10.165177345275879,
      "learning_rate": 4.9627319534895166e-06,
      "loss": 0.5158,
      "mean_token_accuracy": 0.8437979429960251,
      "num_tokens": 100783029.0,
      "step": 83830
    },
    {
      "entropy": 1.8803840085864068,
      "epoch": 0.2598966644166468,
      "grad_norm": 3.2739675045013428,
      "learning_rate": 4.962435976680019e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.853711499273777,
      "num_tokens": 100794075.0,
      "step": 83840
    },
    {
      "entropy": 1.879416285455227,
      "epoch": 0.25992766354169655,
      "grad_norm": 8.004817008972168,
      "learning_rate": 4.9621400528202814e-06,
      "loss": 0.5403,
      "mean_token_accuracy": 0.829861244559288,
      "num_tokens": 100805286.0,
      "step": 83850
    },
    {
      "entropy": 1.8929785311222076,
      "epoch": 0.2599586626667462,
      "grad_norm": 8.447615623474121,
      "learning_rate": 4.96184418189452e-06,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8426691144704819,
      "num_tokens": 100816771.0,
      "step": 83860
    },
    {
      "entropy": 1.8302429109811782,
      "epoch": 0.25998966179179595,
      "grad_norm": 6.880725860595703,
      "learning_rate": 4.9615483638869545e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8440744116902351,
      "num_tokens": 100829800.0,
      "step": 83870
    },
    {
      "entropy": 1.947381141781807,
      "epoch": 0.2600206609168456,
      "grad_norm": 8.487883567810059,
      "learning_rate": 4.961252598781812e-06,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8282511562108994,
      "num_tokens": 100840370.0,
      "step": 83880
    },
    {
      "entropy": 1.8844257071614265,
      "epoch": 0.26005166004189534,
      "grad_norm": 5.970224380493164,
      "learning_rate": 4.960956886563326e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8489205956459045,
      "num_tokens": 100852053.0,
      "step": 83890
    },
    {
      "entropy": 1.783471368253231,
      "epoch": 0.260082659166945,
      "grad_norm": 9.093459129333496,
      "learning_rate": 4.960661227215739e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.858287438750267,
      "num_tokens": 100865934.0,
      "step": 83900
    },
    {
      "entropy": 1.9448846295475959,
      "epoch": 0.26011365829199473,
      "grad_norm": 4.641866683959961,
      "learning_rate": 4.960365620723297e-06,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8277464538812638,
      "num_tokens": 100876883.0,
      "step": 83910
    },
    {
      "entropy": 1.8289760783314706,
      "epoch": 0.2601446574170444,
      "grad_norm": 4.197445392608643,
      "learning_rate": 4.960070067070255e-06,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8453979283571244,
      "num_tokens": 100889433.0,
      "step": 83920
    },
    {
      "entropy": 1.7806072607636452,
      "epoch": 0.2601756565420941,
      "grad_norm": 3.8781611919403076,
      "learning_rate": 4.95977456624087e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.85598254352808,
      "num_tokens": 100902238.0,
      "step": 83930
    },
    {
      "entropy": 1.8489570021629333,
      "epoch": 0.2602066556671438,
      "grad_norm": 4.011428356170654,
      "learning_rate": 4.95947911821941e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8497495472431182,
      "num_tokens": 100914487.0,
      "step": 83940
    },
    {
      "entropy": 1.8972808972001076,
      "epoch": 0.2602376547921935,
      "grad_norm": 7.408459663391113,
      "learning_rate": 4.959183722990151e-06,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8526468485593796,
      "num_tokens": 100925789.0,
      "step": 83950
    },
    {
      "entropy": 1.8424657888710498,
      "epoch": 0.2602686539172432,
      "grad_norm": 8.171313285827637,
      "learning_rate": 4.9588883805373686e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.8596995994448662,
      "num_tokens": 100937994.0,
      "step": 83960
    },
    {
      "entropy": 1.8914167776703834,
      "epoch": 0.26029965304229286,
      "grad_norm": 7.714399814605713,
      "learning_rate": 4.958593090845352e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8574990868568421,
      "num_tokens": 100949553.0,
      "step": 83970
    },
    {
      "entropy": 1.8375475853681564,
      "epoch": 0.2603306521673426,
      "grad_norm": 9.123080253601074,
      "learning_rate": 4.958297853898395e-06,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8516534611582756,
      "num_tokens": 100962264.0,
      "step": 83980
    },
    {
      "entropy": 1.9303360790014268,
      "epoch": 0.26036165129239225,
      "grad_norm": 9.462535858154297,
      "learning_rate": 4.958002669680794e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8448026522994041,
      "num_tokens": 100972790.0,
      "step": 83990
    },
    {
      "entropy": 1.9033367425203322,
      "epoch": 0.260392650417442,
      "grad_norm": 6.338131427764893,
      "learning_rate": 4.957707538176859e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.837484510242939,
      "num_tokens": 100984286.0,
      "step": 84000
    },
    {
      "entropy": 1.9744462817907333,
      "epoch": 0.26042364954249164,
      "grad_norm": 8.549060821533203,
      "learning_rate": 4.9574124593708985e-06,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8344295799732209,
      "num_tokens": 100995371.0,
      "step": 84010
    },
    {
      "entropy": 1.7369410261511802,
      "epoch": 0.26045464866754137,
      "grad_norm": 8.24365520477295,
      "learning_rate": 4.957117433247236e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8557371169328689,
      "num_tokens": 101009845.0,
      "step": 84020
    },
    {
      "entropy": 1.9079995438456536,
      "epoch": 0.26048564779259104,
      "grad_norm": 9.164064407348633,
      "learning_rate": 4.956822459790193e-06,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8357024490833282,
      "num_tokens": 101021686.0,
      "step": 84030
    },
    {
      "entropy": 1.8953372284770011,
      "epoch": 0.26051664691764076,
      "grad_norm": 9.431413650512695,
      "learning_rate": 4.956527538984104e-06,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8483018189668655,
      "num_tokens": 101033454.0,
      "step": 84040
    },
    {
      "entropy": 1.8954351365566253,
      "epoch": 0.26054764604269043,
      "grad_norm": 3.8915398120880127,
      "learning_rate": 4.956232670813308e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8407752260565757,
      "num_tokens": 101044798.0,
      "step": 84050
    },
    {
      "entropy": 1.8448843389749527,
      "epoch": 0.26057864516774015,
      "grad_norm": 4.189465045928955,
      "learning_rate": 4.955937855262149e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.85302524715662,
      "num_tokens": 101057618.0,
      "step": 84060
    },
    {
      "entropy": 1.832655143737793,
      "epoch": 0.2606096442927898,
      "grad_norm": 7.561342239379883,
      "learning_rate": 4.955643092314979e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8396886929869651,
      "num_tokens": 101069666.0,
      "step": 84070
    },
    {
      "entropy": 1.898691761493683,
      "epoch": 0.26064064341783955,
      "grad_norm": 8.714390754699707,
      "learning_rate": 4.955348381956157e-06,
      "loss": 0.5377,
      "mean_token_accuracy": 0.827970777451992,
      "num_tokens": 101081285.0,
      "step": 84080
    },
    {
      "entropy": 1.7665072850883008,
      "epoch": 0.2606716425428892,
      "grad_norm": 8.24121379852295,
      "learning_rate": 4.9550537241700474e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8553405046463013,
      "num_tokens": 101094537.0,
      "step": 84090
    },
    {
      "entropy": 1.8713910579681396,
      "epoch": 0.26070264166793894,
      "grad_norm": 9.43482780456543,
      "learning_rate": 4.954759118941024e-06,
      "loss": 0.4749,
      "mean_token_accuracy": 0.845010656118393,
      "num_tokens": 101106520.0,
      "step": 84100
    },
    {
      "entropy": 1.8151496931910516,
      "epoch": 0.2607336407929886,
      "grad_norm": 9.085859298706055,
      "learning_rate": 4.954464566253459e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8474389865994454,
      "num_tokens": 101119130.0,
      "step": 84110
    },
    {
      "entropy": 1.9014802396297454,
      "epoch": 0.26076463991803833,
      "grad_norm": 3.9436869621276855,
      "learning_rate": 4.95417006609174e-06,
      "loss": 0.5645,
      "mean_token_accuracy": 0.8273421004414558,
      "num_tokens": 101130816.0,
      "step": 84120
    },
    {
      "entropy": 1.9717437848448753,
      "epoch": 0.260795639043088,
      "grad_norm": 8.339731216430664,
      "learning_rate": 4.953875618440259e-06,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8363761708140374,
      "num_tokens": 101142058.0,
      "step": 84130
    },
    {
      "entropy": 1.8874025538563728,
      "epoch": 0.2608266381681377,
      "grad_norm": 5.998051643371582,
      "learning_rate": 4.953581223283413e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8477589651942253,
      "num_tokens": 101153159.0,
      "step": 84140
    },
    {
      "entropy": 1.7967896267771721,
      "epoch": 0.2608576372931874,
      "grad_norm": 8.227131843566895,
      "learning_rate": 4.953286880605603e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8455402344465256,
      "num_tokens": 101165998.0,
      "step": 84150
    },
    {
      "entropy": 1.8968286886811256,
      "epoch": 0.2608886364182371,
      "grad_norm": 7.774280071258545,
      "learning_rate": 4.952992590391242e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8559000626206398,
      "num_tokens": 101177800.0,
      "step": 84160
    },
    {
      "entropy": 1.7744996875524521,
      "epoch": 0.2609196355432868,
      "grad_norm": 8.268373489379883,
      "learning_rate": 4.952698352624746e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8595765858888627,
      "num_tokens": 101190665.0,
      "step": 84170
    },
    {
      "entropy": 1.8480232998728752,
      "epoch": 0.2609506346683365,
      "grad_norm": 7.481622219085693,
      "learning_rate": 4.952404167290538e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.849323433637619,
      "num_tokens": 101203178.0,
      "step": 84180
    },
    {
      "entropy": 1.8875873282551765,
      "epoch": 0.2609816337933862,
      "grad_norm": 4.315346717834473,
      "learning_rate": 4.952110034373047e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.8455449759960174,
      "num_tokens": 101214641.0,
      "step": 84190
    },
    {
      "entropy": 1.7977977305650712,
      "epoch": 0.26101263291843585,
      "grad_norm": 7.563590049743652,
      "learning_rate": 4.9518159538567115e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8515077039599419,
      "num_tokens": 101227190.0,
      "step": 84200
    },
    {
      "entropy": 1.8671275533735752,
      "epoch": 0.2610436320434856,
      "grad_norm": 8.546191215515137,
      "learning_rate": 4.951521925725971e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8378291219472885,
      "num_tokens": 101239940.0,
      "step": 84210
    },
    {
      "entropy": 1.9210672780871392,
      "epoch": 0.26107463116853524,
      "grad_norm": 8.786494255065918,
      "learning_rate": 4.951227949965277e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8403745800256729,
      "num_tokens": 101251411.0,
      "step": 84220
    },
    {
      "entropy": 1.870493806898594,
      "epoch": 0.26110563029358497,
      "grad_norm": 7.429269313812256,
      "learning_rate": 4.9509340265590865e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8402852207422257,
      "num_tokens": 101263421.0,
      "step": 84230
    },
    {
      "entropy": 1.8900667324662208,
      "epoch": 0.26113662941863464,
      "grad_norm": 9.321823120117188,
      "learning_rate": 4.950640155491857e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8420423567295074,
      "num_tokens": 101275836.0,
      "step": 84240
    },
    {
      "entropy": 1.6973878130316735,
      "epoch": 0.26116762854368436,
      "grad_norm": 7.488030910491943,
      "learning_rate": 4.95034633674806e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.868793734908104,
      "num_tokens": 101289722.0,
      "step": 84250
    },
    {
      "entropy": 1.8956893503665924,
      "epoch": 0.26119862766873403,
      "grad_norm": 8.01175594329834,
      "learning_rate": 4.950052570312171e-06,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8392576932907104,
      "num_tokens": 101301555.0,
      "step": 84260
    },
    {
      "entropy": 1.9166808992624282,
      "epoch": 0.26122962679378375,
      "grad_norm": 6.5919928550720215,
      "learning_rate": 4.94975885616867e-06,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8426938772201538,
      "num_tokens": 101312392.0,
      "step": 84270
    },
    {
      "entropy": 1.9476675420999527,
      "epoch": 0.2612606259188334,
      "grad_norm": 9.677026748657227,
      "learning_rate": 4.9494651943020455e-06,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8333592966198922,
      "num_tokens": 101323006.0,
      "step": 84280
    },
    {
      "entropy": 1.818124982714653,
      "epoch": 0.26129162504388315,
      "grad_norm": 4.290999889373779,
      "learning_rate": 4.94917158469679e-06,
      "loss": 0.4158,
      "mean_token_accuracy": 0.8480872571468353,
      "num_tokens": 101335938.0,
      "step": 84290
    },
    {
      "entropy": 1.8131288312375546,
      "epoch": 0.2613226241689328,
      "grad_norm": 8.009013175964355,
      "learning_rate": 4.948878027337407e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.839995177090168,
      "num_tokens": 101349763.0,
      "step": 84300
    },
    {
      "entropy": 1.8972848400473594,
      "epoch": 0.26135362329398254,
      "grad_norm": 8.380783081054688,
      "learning_rate": 4.948584522208402e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8484173402190208,
      "num_tokens": 101361738.0,
      "step": 84310
    },
    {
      "entropy": 1.91866305321455,
      "epoch": 0.2613846224190322,
      "grad_norm": 5.500089645385742,
      "learning_rate": 4.948291069294289e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8474255546927452,
      "num_tokens": 101373866.0,
      "step": 84320
    },
    {
      "entropy": 1.777488799393177,
      "epoch": 0.26141562154408193,
      "grad_norm": 8.77873706817627,
      "learning_rate": 4.947997668579589e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8580469012260437,
      "num_tokens": 101387715.0,
      "step": 84330
    },
    {
      "entropy": 1.9092938348650932,
      "epoch": 0.2614466206691316,
      "grad_norm": 8.913098335266113,
      "learning_rate": 4.947704320048827e-06,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8450995326042176,
      "num_tokens": 101399203.0,
      "step": 84340
    },
    {
      "entropy": 1.8070560455322267,
      "epoch": 0.2614776197941813,
      "grad_norm": 8.269847869873047,
      "learning_rate": 4.947411023686535e-06,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8473980888724327,
      "num_tokens": 101411534.0,
      "step": 84350
    },
    {
      "entropy": 1.8068130150437356,
      "epoch": 0.261508618919231,
      "grad_norm": 4.175093173980713,
      "learning_rate": 4.9471177794772555e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.8587175875902175,
      "num_tokens": 101425216.0,
      "step": 84360
    },
    {
      "entropy": 1.906698650121689,
      "epoch": 0.2615396180442807,
      "grad_norm": 9.216361999511719,
      "learning_rate": 4.946824587405532e-06,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8331103935837746,
      "num_tokens": 101436502.0,
      "step": 84370
    },
    {
      "entropy": 1.7797369614243508,
      "epoch": 0.2615706171693304,
      "grad_norm": 2.6891257762908936,
      "learning_rate": 4.946531447455915e-06,
      "loss": 0.3698,
      "mean_token_accuracy": 0.8665702223777771,
      "num_tokens": 101449912.0,
      "step": 84380
    },
    {
      "entropy": 1.8645511791110039,
      "epoch": 0.2616016162943801,
      "grad_norm": 4.295123100280762,
      "learning_rate": 4.946238359612967e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8417528823018074,
      "num_tokens": 101463078.0,
      "step": 84390
    },
    {
      "entropy": 1.8012978717684747,
      "epoch": 0.2616326154194298,
      "grad_norm": 7.812502861022949,
      "learning_rate": 4.945945323861249e-06,
      "loss": 0.4191,
      "mean_token_accuracy": 0.8614039555191993,
      "num_tokens": 101475914.0,
      "step": 84400
    },
    {
      "entropy": 1.886396862566471,
      "epoch": 0.2616636145444795,
      "grad_norm": 7.294875144958496,
      "learning_rate": 4.945652340185336e-06,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8441071733832359,
      "num_tokens": 101487890.0,
      "step": 84410
    },
    {
      "entropy": 1.7817871391773223,
      "epoch": 0.2616946136695292,
      "grad_norm": 10.357475280761719,
      "learning_rate": 4.9453594085698036e-06,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8648449271917343,
      "num_tokens": 101501358.0,
      "step": 84420
    },
    {
      "entropy": 1.9272992476820945,
      "epoch": 0.2617256127945789,
      "grad_norm": 9.080680847167969,
      "learning_rate": 4.9450665289992355e-06,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8325537592172623,
      "num_tokens": 101513086.0,
      "step": 84430
    },
    {
      "entropy": 1.9224831491708756,
      "epoch": 0.26175661191962857,
      "grad_norm": 8.535758972167969,
      "learning_rate": 4.9447737014582235e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8473838672041893,
      "num_tokens": 101524876.0,
      "step": 84440
    },
    {
      "entropy": 1.8453692942857742,
      "epoch": 0.26178761104467824,
      "grad_norm": 4.549447059631348,
      "learning_rate": 4.9444809259313635e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8549572035670281,
      "num_tokens": 101537098.0,
      "step": 84450
    },
    {
      "entropy": 1.9371949970722198,
      "epoch": 0.26181861016972796,
      "grad_norm": 8.131068229675293,
      "learning_rate": 4.944188202403261e-06,
      "loss": 0.5529,
      "mean_token_accuracy": 0.8304179921746254,
      "num_tokens": 101548246.0,
      "step": 84460
    },
    {
      "entropy": 1.8725028142333031,
      "epoch": 0.26184960929477763,
      "grad_norm": 7.878946781158447,
      "learning_rate": 4.943895530858521e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.840125747025013,
      "num_tokens": 101560349.0,
      "step": 84470
    },
    {
      "entropy": 1.86340638846159,
      "epoch": 0.26188060841982735,
      "grad_norm": 7.6368818283081055,
      "learning_rate": 4.943602911281764e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8389788880944252,
      "num_tokens": 101572940.0,
      "step": 84480
    },
    {
      "entropy": 1.9475991562008859,
      "epoch": 0.261911607544877,
      "grad_norm": 8.88925552368164,
      "learning_rate": 4.943310343657611e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8385830983519554,
      "num_tokens": 101584643.0,
      "step": 84490
    },
    {
      "entropy": 1.9739305421710014,
      "epoch": 0.26194260666992675,
      "grad_norm": 6.695905685424805,
      "learning_rate": 4.943017827970689e-06,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8335414931178093,
      "num_tokens": 101595908.0,
      "step": 84500
    },
    {
      "entropy": 1.9505710929632187,
      "epoch": 0.2619736057949764,
      "grad_norm": 8.216191291809082,
      "learning_rate": 4.942725364205635e-06,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8390222921967506,
      "num_tokens": 101606676.0,
      "step": 84510
    },
    {
      "entropy": 1.932286986708641,
      "epoch": 0.26200460492002614,
      "grad_norm": 7.449299335479736,
      "learning_rate": 4.942432952347092e-06,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8347490966320038,
      "num_tokens": 101618109.0,
      "step": 84520
    },
    {
      "entropy": 1.9106614410877227,
      "epoch": 0.2620356040450758,
      "grad_norm": 9.041545867919922,
      "learning_rate": 4.942140592379704e-06,
      "loss": 0.5185,
      "mean_token_accuracy": 0.8461478322744369,
      "num_tokens": 101629907.0,
      "step": 84530
    },
    {
      "entropy": 1.8061072051525116,
      "epoch": 0.26206660317012553,
      "grad_norm": 9.392251968383789,
      "learning_rate": 4.941848284288129e-06,
      "loss": 0.4095,
      "mean_token_accuracy": 0.8683505341410637,
      "num_tokens": 101642763.0,
      "step": 84540
    },
    {
      "entropy": 1.8147373288869857,
      "epoch": 0.2620976022951752,
      "grad_norm": 3.8025999069213867,
      "learning_rate": 4.941556028057024e-06,
      "loss": 0.3993,
      "mean_token_accuracy": 0.8555989354848862,
      "num_tokens": 101655606.0,
      "step": 84550
    },
    {
      "entropy": 1.9344567239284516,
      "epoch": 0.2621286014202249,
      "grad_norm": 9.05136489868164,
      "learning_rate": 4.9412638236710595e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8368930295109749,
      "num_tokens": 101667711.0,
      "step": 84560
    },
    {
      "entropy": 1.882543933391571,
      "epoch": 0.2621596005452746,
      "grad_norm": 7.798995018005371,
      "learning_rate": 4.940971671114905e-06,
      "loss": 0.4799,
      "mean_token_accuracy": 0.852934005856514,
      "num_tokens": 101678953.0,
      "step": 84570
    },
    {
      "entropy": 1.9046371474862098,
      "epoch": 0.2621905996703243,
      "grad_norm": 8.572905540466309,
      "learning_rate": 4.9406795703732436e-06,
      "loss": 0.4999,
      "mean_token_accuracy": 0.8400033757090568,
      "num_tokens": 101690994.0,
      "step": 84580
    },
    {
      "entropy": 1.955969288945198,
      "epoch": 0.262221598795374,
      "grad_norm": 6.911218166351318,
      "learning_rate": 4.94038752143076e-06,
      "loss": 0.5409,
      "mean_token_accuracy": 0.8287007570266723,
      "num_tokens": 101702443.0,
      "step": 84590
    },
    {
      "entropy": 1.8909879684448243,
      "epoch": 0.2622525979204237,
      "grad_norm": 9.11408519744873,
      "learning_rate": 4.940095524272145e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8456226110458374,
      "num_tokens": 101714218.0,
      "step": 84600
    },
    {
      "entropy": 1.9254190102219582,
      "epoch": 0.2622835970454734,
      "grad_norm": 6.650926113128662,
      "learning_rate": 4.939803578882099e-06,
      "loss": 0.446,
      "mean_token_accuracy": 0.8574258908629417,
      "num_tokens": 101725211.0,
      "step": 84610
    },
    {
      "entropy": 1.9131127685308456,
      "epoch": 0.2623145961705231,
      "grad_norm": 8.146158218383789,
      "learning_rate": 4.939511685245327e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8508924171328545,
      "num_tokens": 101736251.0,
      "step": 84620
    },
    {
      "entropy": 1.9253426373004914,
      "epoch": 0.2623455952955728,
      "grad_norm": 6.606905937194824,
      "learning_rate": 4.939219843346538e-06,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8473613321781158,
      "num_tokens": 101748009.0,
      "step": 84630
    },
    {
      "entropy": 1.9077803820371628,
      "epoch": 0.2623765944206225,
      "grad_norm": 9.684782981872559,
      "learning_rate": 4.938928053170453e-06,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8346043735742569,
      "num_tokens": 101759254.0,
      "step": 84640
    },
    {
      "entropy": 1.9003472611308099,
      "epoch": 0.26240759354567217,
      "grad_norm": 4.827385425567627,
      "learning_rate": 4.938636314701793e-06,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8391556084156037,
      "num_tokens": 101770387.0,
      "step": 84650
    },
    {
      "entropy": 1.911148864030838,
      "epoch": 0.2624385926707219,
      "grad_norm": 8.909619331359863,
      "learning_rate": 4.9383446279252895e-06,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8415949329733848,
      "num_tokens": 101781627.0,
      "step": 84660
    },
    {
      "entropy": 1.8494424894452095,
      "epoch": 0.26246959179577156,
      "grad_norm": 4.314811706542969,
      "learning_rate": 4.93805299282568e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8588860586285592,
      "num_tokens": 101794354.0,
      "step": 84670
    },
    {
      "entropy": 1.9854395300149918,
      "epoch": 0.2625005909208213,
      "grad_norm": 4.768241882324219,
      "learning_rate": 4.937761409387704e-06,
      "loss": 0.5498,
      "mean_token_accuracy": 0.8315106898546218,
      "num_tokens": 101805875.0,
      "step": 84680
    },
    {
      "entropy": 1.8694578632712364,
      "epoch": 0.26253159004587096,
      "grad_norm": 11.459022521972656,
      "learning_rate": 4.937469877596115e-06,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8435337409377098,
      "num_tokens": 101818131.0,
      "step": 84690
    },
    {
      "entropy": 1.9157974675297738,
      "epoch": 0.2625625891709206,
      "grad_norm": 10.549239158630371,
      "learning_rate": 4.9371783974356665e-06,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8447314709424972,
      "num_tokens": 101830440.0,
      "step": 84700
    },
    {
      "entropy": 1.9070345923304557,
      "epoch": 0.26259358829597035,
      "grad_norm": 7.496677398681641,
      "learning_rate": 4.93688696889112e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8451457783579827,
      "num_tokens": 101841964.0,
      "step": 84710
    },
    {
      "entropy": 1.8439010679721832,
      "epoch": 0.26262458742102,
      "grad_norm": 3.86055850982666,
      "learning_rate": 4.936595591947242e-06,
      "loss": 0.4291,
      "mean_token_accuracy": 0.8469921678304673,
      "num_tokens": 101854984.0,
      "step": 84720
    },
    {
      "entropy": 1.8208919912576675,
      "epoch": 0.26265558654606974,
      "grad_norm": 8.848628044128418,
      "learning_rate": 4.93630426658881e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8524587988853455,
      "num_tokens": 101868205.0,
      "step": 84730
    },
    {
      "entropy": 1.9341197982430458,
      "epoch": 0.2626865856711194,
      "grad_norm": 9.021467208862305,
      "learning_rate": 4.936012992800602e-06,
      "loss": 0.5118,
      "mean_token_accuracy": 0.8401493951678276,
      "num_tokens": 101879993.0,
      "step": 84740
    },
    {
      "entropy": 1.900348238646984,
      "epoch": 0.26271758479616913,
      "grad_norm": 9.046464920043945,
      "learning_rate": 4.935721770567406e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8420239523053169,
      "num_tokens": 101891699.0,
      "step": 84750
    },
    {
      "entropy": 2.011109836399555,
      "epoch": 0.2627485839212188,
      "grad_norm": 9.504515647888184,
      "learning_rate": 4.935430599874017e-06,
      "loss": 0.5426,
      "mean_token_accuracy": 0.8297847151756287,
      "num_tokens": 101902570.0,
      "step": 84760
    },
    {
      "entropy": 1.8008052319288255,
      "epoch": 0.26277958304626853,
      "grad_norm": 7.699909687042236,
      "learning_rate": 4.9351394807052325e-06,
      "loss": 0.3834,
      "mean_token_accuracy": 0.8641871735453606,
      "num_tokens": 101915635.0,
      "step": 84770
    },
    {
      "entropy": 1.8245060086250304,
      "epoch": 0.2628105821713182,
      "grad_norm": 8.3301362991333,
      "learning_rate": 4.934848413045857e-06,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8514155328273774,
      "num_tokens": 101928385.0,
      "step": 84780
    },
    {
      "entropy": 1.8403597161173821,
      "epoch": 0.2628415812963679,
      "grad_norm": 3.8224892616271973,
      "learning_rate": 4.934557396880704e-06,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8499826923012733,
      "num_tokens": 101941790.0,
      "step": 84790
    },
    {
      "entropy": 1.950501237809658,
      "epoch": 0.2628725804214176,
      "grad_norm": 6.607431411743164,
      "learning_rate": 4.934266432194593e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8484028398990631,
      "num_tokens": 101953387.0,
      "step": 84800
    },
    {
      "entropy": 1.913737191259861,
      "epoch": 0.2629035795464673,
      "grad_norm": 7.8601508140563965,
      "learning_rate": 4.933975518972347e-06,
      "loss": 0.485,
      "mean_token_accuracy": 0.850081168115139,
      "num_tokens": 101964983.0,
      "step": 84810
    },
    {
      "entropy": 1.9299700886011124,
      "epoch": 0.262934578671517,
      "grad_norm": 8.866097450256348,
      "learning_rate": 4.9336846571987965e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8501459851861,
      "num_tokens": 101976619.0,
      "step": 84820
    },
    {
      "entropy": 1.9162129878997802,
      "epoch": 0.2629655777965667,
      "grad_norm": 8.77447509765625,
      "learning_rate": 4.93339384685878e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8534630805253982,
      "num_tokens": 101988887.0,
      "step": 84830
    },
    {
      "entropy": 1.9051405146718026,
      "epoch": 0.2629965769216164,
      "grad_norm": 7.113283157348633,
      "learning_rate": 4.93310308793714e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8431653559207917,
      "num_tokens": 102000822.0,
      "step": 84840
    },
    {
      "entropy": 1.8918147072196008,
      "epoch": 0.2630275760466661,
      "grad_norm": 7.229731559753418,
      "learning_rate": 4.9328123804187265e-06,
      "loss": 0.42,
      "mean_token_accuracy": 0.8591841533780098,
      "num_tokens": 102012280.0,
      "step": 84850
    },
    {
      "entropy": 1.84690400660038,
      "epoch": 0.26305857517171577,
      "grad_norm": 4.94510555267334,
      "learning_rate": 4.932521724288395e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8478389665484428,
      "num_tokens": 102025734.0,
      "step": 84860
    },
    {
      "entropy": 1.8911878764629364,
      "epoch": 0.2630895742967655,
      "grad_norm": 4.1434783935546875,
      "learning_rate": 4.932231119531007e-06,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8523555234074592,
      "num_tokens": 102038473.0,
      "step": 84870
    },
    {
      "entropy": 1.9033065840601922,
      "epoch": 0.26312057342181516,
      "grad_norm": 3.400432586669922,
      "learning_rate": 4.9319405661314326e-06,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8344430983066559,
      "num_tokens": 102050229.0,
      "step": 84880
    },
    {
      "entropy": 1.9568065509200097,
      "epoch": 0.2631515725468649,
      "grad_norm": 10.253649711608887,
      "learning_rate": 4.931650064074543e-06,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8353859558701515,
      "num_tokens": 102062301.0,
      "step": 84890
    },
    {
      "entropy": 1.8090133965015411,
      "epoch": 0.26318257167191456,
      "grad_norm": 9.690705299377441,
      "learning_rate": 4.931359613345223e-06,
      "loss": 0.409,
      "mean_token_accuracy": 0.8670252472162246,
      "num_tokens": 102076084.0,
      "step": 84900
    },
    {
      "entropy": 1.881265440583229,
      "epoch": 0.2632135707969643,
      "grad_norm": 8.040104866027832,
      "learning_rate": 4.9310692139283576e-06,
      "loss": 0.4399,
      "mean_token_accuracy": 0.848574922978878,
      "num_tokens": 102088089.0,
      "step": 84910
    },
    {
      "entropy": 1.823999959230423,
      "epoch": 0.26324456992201395,
      "grad_norm": 4.09398078918457,
      "learning_rate": 4.9307788658088396e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8424740388989449,
      "num_tokens": 102101088.0,
      "step": 84920
    },
    {
      "entropy": 1.8733570471405983,
      "epoch": 0.2632755690470637,
      "grad_norm": 7.2603230476379395,
      "learning_rate": 4.93048856897157e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8479010179638863,
      "num_tokens": 102113485.0,
      "step": 84930
    },
    {
      "entropy": 1.8576104506850242,
      "epoch": 0.26330656817211334,
      "grad_norm": 8.178943634033203,
      "learning_rate": 4.930198323401454e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8650277197360993,
      "num_tokens": 102126971.0,
      "step": 84940
    },
    {
      "entropy": 1.819041645526886,
      "epoch": 0.263337567297163,
      "grad_norm": 10.852204322814941,
      "learning_rate": 4.929908129083402e-06,
      "loss": 0.4542,
      "mean_token_accuracy": 0.84122084826231,
      "num_tokens": 102140395.0,
      "step": 84950
    },
    {
      "entropy": 1.8517862141132355,
      "epoch": 0.26336856642221274,
      "grad_norm": 7.924666404724121,
      "learning_rate": 4.929617986002334e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.841532975435257,
      "num_tokens": 102153419.0,
      "step": 84960
    },
    {
      "entropy": 1.9836448311805726,
      "epoch": 0.2633995655472624,
      "grad_norm": 7.1847004890441895,
      "learning_rate": 4.9293278941431724e-06,
      "loss": 0.5787,
      "mean_token_accuracy": 0.830291448533535,
      "num_tokens": 102164560.0,
      "step": 84970
    },
    {
      "entropy": 1.9570752799510955,
      "epoch": 0.26343056467231213,
      "grad_norm": 7.852522850036621,
      "learning_rate": 4.929037853490851e-06,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8373702332377434,
      "num_tokens": 102176029.0,
      "step": 84980
    },
    {
      "entropy": 1.9419500917196273,
      "epoch": 0.2634615637973618,
      "grad_norm": 8.110418319702148,
      "learning_rate": 4.928747864030304e-06,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8400467559695244,
      "num_tokens": 102188021.0,
      "step": 84990
    },
    {
      "entropy": 1.9458017632365228,
      "epoch": 0.2634925629224115,
      "grad_norm": 10.08768081665039,
      "learning_rate": 4.928457925746475e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8457490563392639,
      "num_tokens": 102199697.0,
      "step": 85000
    },
    {
      "entropy": 1.9600411295890807,
      "epoch": 0.2635235620474612,
      "grad_norm": 8.410282135009766,
      "learning_rate": 4.928168038624313e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8549587532877923,
      "num_tokens": 102210184.0,
      "step": 85010
    },
    {
      "entropy": 1.7980479046702385,
      "epoch": 0.2635545611725109,
      "grad_norm": 7.882757663726807,
      "learning_rate": 4.927878202648774e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8440237864851952,
      "num_tokens": 102223273.0,
      "step": 85020
    },
    {
      "entropy": 1.9615017265081405,
      "epoch": 0.2635855602975606,
      "grad_norm": 9.58923053741455,
      "learning_rate": 4.927588417804819e-06,
      "loss": 0.5578,
      "mean_token_accuracy": 0.8358608677983284,
      "num_tokens": 102234501.0,
      "step": 85030
    },
    {
      "entropy": 1.864892715215683,
      "epoch": 0.2636165594226103,
      "grad_norm": 7.730620384216309,
      "learning_rate": 4.9272986840774155e-06,
      "loss": 0.456,
      "mean_token_accuracy": 0.8360424354672432,
      "num_tokens": 102247359.0,
      "step": 85040
    },
    {
      "entropy": 1.8440838590264321,
      "epoch": 0.26364755854766,
      "grad_norm": 3.6796610355377197,
      "learning_rate": 4.927009001451538e-06,
      "loss": 0.4102,
      "mean_token_accuracy": 0.8569944024085998,
      "num_tokens": 102260021.0,
      "step": 85050
    },
    {
      "entropy": 1.905400250852108,
      "epoch": 0.2636785576727097,
      "grad_norm": 3.788116931915283,
      "learning_rate": 4.926719369912167e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8591458886861801,
      "num_tokens": 102271674.0,
      "step": 85060
    },
    {
      "entropy": 1.8443732798099517,
      "epoch": 0.26370955679775937,
      "grad_norm": 7.661864757537842,
      "learning_rate": 4.926429789444288e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8470209822058677,
      "num_tokens": 102284578.0,
      "step": 85070
    },
    {
      "entropy": 1.8414702624082566,
      "epoch": 0.2637405559228091,
      "grad_norm": 7.228357791900635,
      "learning_rate": 4.926140260032895e-06,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8367054253816605,
      "num_tokens": 102296979.0,
      "step": 85080
    },
    {
      "entropy": 1.9466857939958573,
      "epoch": 0.26377155504785876,
      "grad_norm": 11.023140907287598,
      "learning_rate": 4.925850781662983e-06,
      "loss": 0.5557,
      "mean_token_accuracy": 0.8282995998859406,
      "num_tokens": 102308493.0,
      "step": 85090
    },
    {
      "entropy": 1.963821244239807,
      "epoch": 0.2638025541729085,
      "grad_norm": 11.524614334106445,
      "learning_rate": 4.92556135431956e-06,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8317586749792099,
      "num_tokens": 102320319.0,
      "step": 85100
    },
    {
      "entropy": 1.9346794202923774,
      "epoch": 0.26383355329795816,
      "grad_norm": 7.5674214363098145,
      "learning_rate": 4.9252719779876374e-06,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8399398401379585,
      "num_tokens": 102331510.0,
      "step": 85110
    },
    {
      "entropy": 1.8758284986019134,
      "epoch": 0.2638645524230079,
      "grad_norm": 3.6614625453948975,
      "learning_rate": 4.92498265265223e-06,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8341593369841576,
      "num_tokens": 102344007.0,
      "step": 85120
    },
    {
      "entropy": 1.9138728231191635,
      "epoch": 0.26389555154805755,
      "grad_norm": 4.82459831237793,
      "learning_rate": 4.924693378298362e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8473570510745049,
      "num_tokens": 102355863.0,
      "step": 85130
    },
    {
      "entropy": 1.9841463685035705,
      "epoch": 0.2639265506731073,
      "grad_norm": 7.99963903427124,
      "learning_rate": 4.924404154911063e-06,
      "loss": 0.5413,
      "mean_token_accuracy": 0.8389569863677024,
      "num_tokens": 102366684.0,
      "step": 85140
    },
    {
      "entropy": 1.8069503650069236,
      "epoch": 0.26395754979815694,
      "grad_norm": 7.8926568031311035,
      "learning_rate": 4.92411498247537e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8335721820592881,
      "num_tokens": 102380841.0,
      "step": 85150
    },
    {
      "entropy": 1.9148933947086335,
      "epoch": 0.26398854892320667,
      "grad_norm": 7.529796600341797,
      "learning_rate": 4.923825860976324e-06,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8480879321694375,
      "num_tokens": 102392714.0,
      "step": 85160
    },
    {
      "entropy": 1.9118664294481278,
      "epoch": 0.26401954804825634,
      "grad_norm": 9.968255996704102,
      "learning_rate": 4.9235367903989705e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8447333693504333,
      "num_tokens": 102404766.0,
      "step": 85170
    },
    {
      "entropy": 1.9002967342734336,
      "epoch": 0.26405054717330606,
      "grad_norm": 9.95118236541748,
      "learning_rate": 4.923247770728366e-06,
      "loss": 0.4568,
      "mean_token_accuracy": 0.855803194642067,
      "num_tokens": 102416668.0,
      "step": 85180
    },
    {
      "entropy": 1.8763260886073112,
      "epoch": 0.26408154629835573,
      "grad_norm": 7.586075305938721,
      "learning_rate": 4.9229588019495714e-06,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8425167426466942,
      "num_tokens": 102429365.0,
      "step": 85190
    },
    {
      "entropy": 2.0041319727897644,
      "epoch": 0.2641125454234054,
      "grad_norm": 9.366758346557617,
      "learning_rate": 4.922669884047651e-06,
      "loss": 0.5717,
      "mean_token_accuracy": 0.8297568678855896,
      "num_tokens": 102440584.0,
      "step": 85200
    },
    {
      "entropy": 1.9379279300570489,
      "epoch": 0.2641435445484551,
      "grad_norm": 8.975931167602539,
      "learning_rate": 4.922381017007679e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8403313905000687,
      "num_tokens": 102452027.0,
      "step": 85210
    },
    {
      "entropy": 1.9602735459804534,
      "epoch": 0.2641745436735048,
      "grad_norm": 8.235448837280273,
      "learning_rate": 4.9220922008147325e-06,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8311275467276573,
      "num_tokens": 102463662.0,
      "step": 85220
    },
    {
      "entropy": 1.8260437846183777,
      "epoch": 0.2642055427985545,
      "grad_norm": 9.317481994628906,
      "learning_rate": 4.921803435453896e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8505254268646241,
      "num_tokens": 102476893.0,
      "step": 85230
    },
    {
      "entropy": 1.8744411259889602,
      "epoch": 0.2642365419236042,
      "grad_norm": 7.509456634521484,
      "learning_rate": 4.921514720910262e-06,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8349858596920967,
      "num_tokens": 102489557.0,
      "step": 85240
    },
    {
      "entropy": 1.8670647412538528,
      "epoch": 0.2642675410486539,
      "grad_norm": 3.518993377685547,
      "learning_rate": 4.921226057168927e-06,
      "loss": 0.4244,
      "mean_token_accuracy": 0.8494436904788017,
      "num_tokens": 102502419.0,
      "step": 85250
    },
    {
      "entropy": 1.9050493866205216,
      "epoch": 0.2642985401737036,
      "grad_norm": 8.357398986816406,
      "learning_rate": 4.920937444214995e-06,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8412919878959656,
      "num_tokens": 102514450.0,
      "step": 85260
    },
    {
      "entropy": 1.79098000228405,
      "epoch": 0.2643295392987533,
      "grad_norm": 4.370672702789307,
      "learning_rate": 4.920648882033572e-06,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8489180520176888,
      "num_tokens": 102527984.0,
      "step": 85270
    },
    {
      "entropy": 1.9773102968931198,
      "epoch": 0.26436053842380297,
      "grad_norm": 7.729239463806152,
      "learning_rate": 4.920360370609777e-06,
      "loss": 0.4939,
      "mean_token_accuracy": 0.8391767382621765,
      "num_tokens": 102538749.0,
      "step": 85280
    },
    {
      "entropy": 1.935066755115986,
      "epoch": 0.2643915375488527,
      "grad_norm": 8.969949722290039,
      "learning_rate": 4.920071909928729e-06,
      "loss": 0.5233,
      "mean_token_accuracy": 0.842116117477417,
      "num_tokens": 102550262.0,
      "step": 85290
    },
    {
      "entropy": 1.8114179536700248,
      "epoch": 0.26442253667390236,
      "grad_norm": 7.595683574676514,
      "learning_rate": 4.919783499975556e-06,
      "loss": 0.3997,
      "mean_token_accuracy": 0.8623112186789512,
      "num_tokens": 102563124.0,
      "step": 85300
    },
    {
      "entropy": 1.8905057892203332,
      "epoch": 0.2644535357989521,
      "grad_norm": 7.8862199783325195,
      "learning_rate": 4.919495140735392e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8516170725226402,
      "num_tokens": 102575911.0,
      "step": 85310
    },
    {
      "entropy": 1.7555623829364777,
      "epoch": 0.26448453492400176,
      "grad_norm": 7.201220989227295,
      "learning_rate": 4.919206832193378e-06,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8540243923664093,
      "num_tokens": 102589847.0,
      "step": 85320
    },
    {
      "entropy": 1.9273912906646729,
      "epoch": 0.2645155340490515,
      "grad_norm": 8.060796737670898,
      "learning_rate": 4.918918574334659e-06,
      "loss": 0.4865,
      "mean_token_accuracy": 0.848893666267395,
      "num_tokens": 102601212.0,
      "step": 85330
    },
    {
      "entropy": 1.8354052126407623,
      "epoch": 0.26454653317410115,
      "grad_norm": 8.087235450744629,
      "learning_rate": 4.918630367144384e-06,
      "loss": 0.543,
      "mean_token_accuracy": 0.8323055505752563,
      "num_tokens": 102614529.0,
      "step": 85340
    },
    {
      "entropy": 1.9022644311189651,
      "epoch": 0.2645775322991509,
      "grad_norm": 9.103813171386719,
      "learning_rate": 4.918342210607715e-06,
      "loss": 0.5306,
      "mean_token_accuracy": 0.828366307914257,
      "num_tokens": 102627163.0,
      "step": 85350
    },
    {
      "entropy": 1.8333250015974045,
      "epoch": 0.26460853142420054,
      "grad_norm": 9.28365707397461,
      "learning_rate": 4.918054104709815e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8535311058163643,
      "num_tokens": 102639926.0,
      "step": 85360
    },
    {
      "entropy": 1.8946420654654503,
      "epoch": 0.26463953054925027,
      "grad_norm": 7.266073703765869,
      "learning_rate": 4.917766049435854e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8455268830060959,
      "num_tokens": 102652224.0,
      "step": 85370
    },
    {
      "entropy": 1.9376175180077553,
      "epoch": 0.26467052967429994,
      "grad_norm": 7.632312297821045,
      "learning_rate": 4.917478044771007e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8479833841323853,
      "num_tokens": 102664350.0,
      "step": 85380
    },
    {
      "entropy": 1.9229725405573845,
      "epoch": 0.26470152879934966,
      "grad_norm": 7.888539791107178,
      "learning_rate": 4.9171900907004585e-06,
      "loss": 0.5669,
      "mean_token_accuracy": 0.8408885851502419,
      "num_tokens": 102675841.0,
      "step": 85390
    },
    {
      "entropy": 1.8650186344981194,
      "epoch": 0.26473252792439933,
      "grad_norm": 7.847538471221924,
      "learning_rate": 4.916902187209395e-06,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8481592908501625,
      "num_tokens": 102687826.0,
      "step": 85400
    },
    {
      "entropy": 1.8229721501469611,
      "epoch": 0.26476352704944905,
      "grad_norm": 7.471333026885986,
      "learning_rate": 4.916614334283012e-06,
      "loss": 0.406,
      "mean_token_accuracy": 0.8582251816987991,
      "num_tokens": 102700892.0,
      "step": 85410
    },
    {
      "entropy": 1.934105758368969,
      "epoch": 0.2647945261744987,
      "grad_norm": 3.659540891647339,
      "learning_rate": 4.91632653190651e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8533608466386795,
      "num_tokens": 102712704.0,
      "step": 85420
    },
    {
      "entropy": 1.9130233809351922,
      "epoch": 0.26482552529954845,
      "grad_norm": 9.22014331817627,
      "learning_rate": 4.916038780065096e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8448394879698753,
      "num_tokens": 102724684.0,
      "step": 85430
    },
    {
      "entropy": 1.9155945912003518,
      "epoch": 0.2648565244245981,
      "grad_norm": 8.854886054992676,
      "learning_rate": 4.9157510787439814e-06,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8422875538468361,
      "num_tokens": 102736595.0,
      "step": 85440
    },
    {
      "entropy": 1.8779947102069854,
      "epoch": 0.2648875235496478,
      "grad_norm": 7.41634464263916,
      "learning_rate": 4.9154634279283864e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.8430365577340126,
      "num_tokens": 102749126.0,
      "step": 85450
    },
    {
      "entropy": 1.8960552558302879,
      "epoch": 0.2649185226746975,
      "grad_norm": 10.775057792663574,
      "learning_rate": 4.915175827603535e-06,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8464867159724235,
      "num_tokens": 102761000.0,
      "step": 85460
    },
    {
      "entropy": 1.8219962686300277,
      "epoch": 0.2649495217997472,
      "grad_norm": 4.283982753753662,
      "learning_rate": 4.914888277754658e-06,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8421214401721955,
      "num_tokens": 102774313.0,
      "step": 85470
    },
    {
      "entropy": 1.9417266234755517,
      "epoch": 0.2649805209247969,
      "grad_norm": 3.7757527828216553,
      "learning_rate": 4.914600778366993e-06,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8345933735370636,
      "num_tokens": 102785961.0,
      "step": 85480
    },
    {
      "entropy": 1.9103323504328729,
      "epoch": 0.26501152004984657,
      "grad_norm": 3.362926721572876,
      "learning_rate": 4.9143133294257815e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8477786153554916,
      "num_tokens": 102798060.0,
      "step": 85490
    },
    {
      "entropy": 1.943329544365406,
      "epoch": 0.2650425191748963,
      "grad_norm": 9.327890396118164,
      "learning_rate": 4.914025930916273e-06,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8373201444745064,
      "num_tokens": 102809320.0,
      "step": 85500
    },
    {
      "entropy": 1.881453277170658,
      "epoch": 0.26507351829994596,
      "grad_norm": 3.5227839946746826,
      "learning_rate": 4.913738582823723e-06,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8572242721915245,
      "num_tokens": 102820907.0,
      "step": 85510
    },
    {
      "entropy": 1.8902042105793952,
      "epoch": 0.2651045174249957,
      "grad_norm": 7.871299743652344,
      "learning_rate": 4.913451285133394e-06,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8486515626311302,
      "num_tokens": 102832171.0,
      "step": 85520
    },
    {
      "entropy": 1.9584900826215743,
      "epoch": 0.26513551655004536,
      "grad_norm": 3.7320501804351807,
      "learning_rate": 4.91316403783055e-06,
      "loss": 0.5608,
      "mean_token_accuracy": 0.8346157044172287,
      "num_tokens": 102843429.0,
      "step": 85530
    },
    {
      "entropy": 1.8431198254227639,
      "epoch": 0.2651665156750951,
      "grad_norm": 3.781135320663452,
      "learning_rate": 4.912876840900466e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8510622009634972,
      "num_tokens": 102855860.0,
      "step": 85540
    },
    {
      "entropy": 1.9608791798353196,
      "epoch": 0.26519751480014475,
      "grad_norm": 9.239676475524902,
      "learning_rate": 4.91258969432842e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8367750391364097,
      "num_tokens": 102867302.0,
      "step": 85550
    },
    {
      "entropy": 1.8789279222488404,
      "epoch": 0.2652285139251945,
      "grad_norm": 3.7677621841430664,
      "learning_rate": 4.912302598099698e-06,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8507544234395027,
      "num_tokens": 102879643.0,
      "step": 85560
    },
    {
      "entropy": 1.9066937759518623,
      "epoch": 0.26525951305024414,
      "grad_norm": 10.09460163116455,
      "learning_rate": 4.9120155521995925e-06,
      "loss": 0.4615,
      "mean_token_accuracy": 0.8424730479717255,
      "num_tokens": 102892044.0,
      "step": 85570
    },
    {
      "entropy": 1.8171610802412033,
      "epoch": 0.26529051217529387,
      "grad_norm": 7.337011337280273,
      "learning_rate": 4.911728556613397e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8553290620446206,
      "num_tokens": 102904614.0,
      "step": 85580
    },
    {
      "entropy": 1.8738331109285356,
      "epoch": 0.26532151130034354,
      "grad_norm": 9.422406196594238,
      "learning_rate": 4.911441611326418e-06,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8550050288438797,
      "num_tokens": 102916052.0,
      "step": 85590
    },
    {
      "entropy": 1.776879619061947,
      "epoch": 0.26535251042539326,
      "grad_norm": 7.145488262176514,
      "learning_rate": 4.911154716323966e-06,
      "loss": 0.3784,
      "mean_token_accuracy": 0.8612538367509842,
      "num_tokens": 102929553.0,
      "step": 85600
    },
    {
      "entropy": 1.9105664394795894,
      "epoch": 0.26538350955044293,
      "grad_norm": 12.720219612121582,
      "learning_rate": 4.91086787159135e-06,
      "loss": 0.5435,
      "mean_token_accuracy": 0.8275511726737023,
      "num_tokens": 102940839.0,
      "step": 85610
    },
    {
      "entropy": 1.8572928413748742,
      "epoch": 0.26541450867549266,
      "grad_norm": 2.5268802642822266,
      "learning_rate": 4.910581077113897e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8346114948391914,
      "num_tokens": 102953161.0,
      "step": 85620
    },
    {
      "entropy": 1.8211661458015442,
      "epoch": 0.2654455078005423,
      "grad_norm": 4.548563480377197,
      "learning_rate": 4.910294332876931e-06,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8422662571072579,
      "num_tokens": 102966676.0,
      "step": 85630
    },
    {
      "entropy": 1.874080342054367,
      "epoch": 0.26547650692559205,
      "grad_norm": 9.284010887145996,
      "learning_rate": 4.910007638865787e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8418890178203583,
      "num_tokens": 102978834.0,
      "step": 85640
    },
    {
      "entropy": 1.9538467079401016,
      "epoch": 0.2655075060506417,
      "grad_norm": 7.954508304595947,
      "learning_rate": 4.909720995065805e-06,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8483369365334511,
      "num_tokens": 102989479.0,
      "step": 85650
    },
    {
      "entropy": 1.9834106057882308,
      "epoch": 0.26553850517569144,
      "grad_norm": 9.530325889587402,
      "learning_rate": 4.909434401462327e-06,
      "loss": 0.5648,
      "mean_token_accuracy": 0.8310058429837227,
      "num_tokens": 103000295.0,
      "step": 85660
    },
    {
      "entropy": 1.8795513778924942,
      "epoch": 0.2655695043007411,
      "grad_norm": 9.067574501037598,
      "learning_rate": 4.9091478580407075e-06,
      "loss": 0.4291,
      "mean_token_accuracy": 0.8638965725898743,
      "num_tokens": 103012052.0,
      "step": 85670
    },
    {
      "entropy": 1.8716339603066445,
      "epoch": 0.2656005034257908,
      "grad_norm": 11.293997764587402,
      "learning_rate": 4.908861364786301e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8458243295550346,
      "num_tokens": 103023681.0,
      "step": 85680
    },
    {
      "entropy": 1.8574876859784126,
      "epoch": 0.2656315025508405,
      "grad_norm": 8.510407447814941,
      "learning_rate": 4.908574921684474e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8516736879944802,
      "num_tokens": 103035416.0,
      "step": 85690
    },
    {
      "entropy": 1.8569199055433274,
      "epoch": 0.2656625016758902,
      "grad_norm": 4.086944103240967,
      "learning_rate": 4.908288528720592e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8491792261600495,
      "num_tokens": 103046882.0,
      "step": 85700
    },
    {
      "entropy": 1.9134871244430542,
      "epoch": 0.2656935008009399,
      "grad_norm": 7.807006359100342,
      "learning_rate": 4.908002185880031e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.847848904132843,
      "num_tokens": 103058705.0,
      "step": 85710
    },
    {
      "entropy": 1.911807608604431,
      "epoch": 0.26572449992598957,
      "grad_norm": 9.756507873535156,
      "learning_rate": 4.907715893148174e-06,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8321962848305702,
      "num_tokens": 103071199.0,
      "step": 85720
    },
    {
      "entropy": 1.883806975185871,
      "epoch": 0.2657554990510393,
      "grad_norm": 6.826350212097168,
      "learning_rate": 4.9074296505104055e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8427292719483376,
      "num_tokens": 103082635.0,
      "step": 85730
    },
    {
      "entropy": 1.9495513945817948,
      "epoch": 0.26578649817608896,
      "grad_norm": 10.174969673156738,
      "learning_rate": 4.9071434579521205e-06,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8444296821951867,
      "num_tokens": 103093525.0,
      "step": 85740
    },
    {
      "entropy": 1.934284047782421,
      "epoch": 0.2658174973011387,
      "grad_norm": 9.920119285583496,
      "learning_rate": 4.9068573154587165e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8450621247291565,
      "num_tokens": 103105373.0,
      "step": 85750
    },
    {
      "entropy": 1.7569790095090867,
      "epoch": 0.26584849642618835,
      "grad_norm": 7.43773078918457,
      "learning_rate": 4.9065712230156e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8663360059261322,
      "num_tokens": 103118573.0,
      "step": 85760
    },
    {
      "entropy": 1.8341956838965416,
      "epoch": 0.2658794955512381,
      "grad_norm": 7.597226619720459,
      "learning_rate": 4.906285180608181e-06,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8481891751289368,
      "num_tokens": 103131304.0,
      "step": 85770
    },
    {
      "entropy": 1.9248315215110778,
      "epoch": 0.26591049467628775,
      "grad_norm": 8.323822021484375,
      "learning_rate": 4.905999188221875e-06,
      "loss": 0.519,
      "mean_token_accuracy": 0.8360444948077201,
      "num_tokens": 103142765.0,
      "step": 85780
    },
    {
      "entropy": 1.892122246325016,
      "epoch": 0.26594149380133747,
      "grad_norm": 8.809585571289062,
      "learning_rate": 4.905713245842107e-06,
      "loss": 0.5083,
      "mean_token_accuracy": 0.834871856868267,
      "num_tokens": 103154444.0,
      "step": 85790
    },
    {
      "entropy": 1.9127579972147941,
      "epoch": 0.26597249292638714,
      "grad_norm": 8.541977882385254,
      "learning_rate": 4.905427353454305e-06,
      "loss": 0.5018,
      "mean_token_accuracy": 0.839243420958519,
      "num_tokens": 103166217.0,
      "step": 85800
    },
    {
      "entropy": 1.8976189360022544,
      "epoch": 0.26600349205143686,
      "grad_norm": 8.723061561584473,
      "learning_rate": 4.905141511043905e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8352738916873932,
      "num_tokens": 103178533.0,
      "step": 85810
    },
    {
      "entropy": 1.9483844459056854,
      "epoch": 0.26603449117648653,
      "grad_norm": 8.283819198608398,
      "learning_rate": 4.904855718596345e-06,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8419344946742058,
      "num_tokens": 103188784.0,
      "step": 85820
    },
    {
      "entropy": 1.8302007086575032,
      "epoch": 0.26606549030153626,
      "grad_norm": 7.834589004516602,
      "learning_rate": 4.9045699760970725e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8487353563308716,
      "num_tokens": 103201469.0,
      "step": 85830
    },
    {
      "entropy": 1.9522307723760606,
      "epoch": 0.2660964894265859,
      "grad_norm": 11.1272554397583,
      "learning_rate": 4.904284283531541e-06,
      "loss": 0.4868,
      "mean_token_accuracy": 0.8433923795819283,
      "num_tokens": 103212554.0,
      "step": 85840
    },
    {
      "entropy": 1.9034996896982193,
      "epoch": 0.26612748855163565,
      "grad_norm": 8.404441833496094,
      "learning_rate": 4.903998640885207e-06,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8423371136188507,
      "num_tokens": 103223756.0,
      "step": 85850
    },
    {
      "entropy": 1.8009837806224822,
      "epoch": 0.2661584876766853,
      "grad_norm": 9.704791069030762,
      "learning_rate": 4.903713048143537e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.851232835650444,
      "num_tokens": 103237242.0,
      "step": 85860
    },
    {
      "entropy": 1.8711876735091209,
      "epoch": 0.26618948680173504,
      "grad_norm": 8.211481094360352,
      "learning_rate": 4.903427505292001e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.842264424264431,
      "num_tokens": 103249121.0,
      "step": 85870
    },
    {
      "entropy": 1.7938581064343453,
      "epoch": 0.2662204859267847,
      "grad_norm": 3.8567850589752197,
      "learning_rate": 4.903142012316073e-06,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8489397838711739,
      "num_tokens": 103261025.0,
      "step": 85880
    },
    {
      "entropy": 1.9541205585002899,
      "epoch": 0.26625148505183444,
      "grad_norm": 10.45516300201416,
      "learning_rate": 4.902856569201237e-06,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8429744258522988,
      "num_tokens": 103271985.0,
      "step": 85890
    },
    {
      "entropy": 1.7870109647512435,
      "epoch": 0.2662824841768841,
      "grad_norm": 2.5378193855285645,
      "learning_rate": 4.90257117593298e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8521895721554756,
      "num_tokens": 103285897.0,
      "step": 85900
    },
    {
      "entropy": 1.769122688472271,
      "epoch": 0.26631348330193383,
      "grad_norm": 8.693305969238281,
      "learning_rate": 4.902285832496798e-06,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8523093193769455,
      "num_tokens": 103299434.0,
      "step": 85910
    },
    {
      "entropy": 1.8774210557341575,
      "epoch": 0.2663444824269835,
      "grad_norm": 7.8302507400512695,
      "learning_rate": 4.902000538878188e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8508623972535133,
      "num_tokens": 103312561.0,
      "step": 85920
    },
    {
      "entropy": 1.8622436970472336,
      "epoch": 0.26637548155203317,
      "grad_norm": 8.453393936157227,
      "learning_rate": 4.9017152950626585e-06,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8456780537962914,
      "num_tokens": 103324471.0,
      "step": 85930
    },
    {
      "entropy": 1.9303288459777832,
      "epoch": 0.2664064806770829,
      "grad_norm": 9.193410873413086,
      "learning_rate": 4.901430101035719e-06,
      "loss": 0.494,
      "mean_token_accuracy": 0.8492061391472816,
      "num_tokens": 103336441.0,
      "step": 85940
    },
    {
      "entropy": 1.9568013072013855,
      "epoch": 0.26643747980213256,
      "grad_norm": 8.285205841064453,
      "learning_rate": 4.901144956782889e-06,
      "loss": 0.5306,
      "mean_token_accuracy": 0.8406900480389595,
      "num_tokens": 103347523.0,
      "step": 85950
    },
    {
      "entropy": 1.9028612434864045,
      "epoch": 0.2664684789271823,
      "grad_norm": 8.355502128601074,
      "learning_rate": 4.900859862289691e-06,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8526147872209549,
      "num_tokens": 103358369.0,
      "step": 85960
    },
    {
      "entropy": 1.929163258522749,
      "epoch": 0.26649947805223195,
      "grad_norm": 8.101810455322266,
      "learning_rate": 4.900574817541653e-06,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8357960835099221,
      "num_tokens": 103369659.0,
      "step": 85970
    },
    {
      "entropy": 1.910251635313034,
      "epoch": 0.2665304771772817,
      "grad_norm": 8.729976654052734,
      "learning_rate": 4.900289822524311e-06,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8469779431819916,
      "num_tokens": 103381688.0,
      "step": 85980
    },
    {
      "entropy": 1.9128791213035583,
      "epoch": 0.26656147630233135,
      "grad_norm": 4.791084289550781,
      "learning_rate": 4.900004877223208e-06,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8399335280060768,
      "num_tokens": 103393309.0,
      "step": 85990
    },
    {
      "entropy": 1.8070705935359002,
      "epoch": 0.26659247542738107,
      "grad_norm": 4.363813877105713,
      "learning_rate": 4.899719981623888e-06,
      "loss": 0.3931,
      "mean_token_accuracy": 0.8653892949223518,
      "num_tokens": 103406198.0,
      "step": 86000
    },
    {
      "entropy": 1.858493185043335,
      "epoch": 0.26662347455243074,
      "grad_norm": 8.77523422241211,
      "learning_rate": 4.899435135711908e-06,
      "loss": 0.4392,
      "mean_token_accuracy": 0.853272658586502,
      "num_tokens": 103418547.0,
      "step": 86010
    },
    {
      "entropy": 1.8741918861865998,
      "epoch": 0.26665447367748046,
      "grad_norm": 8.620582580566406,
      "learning_rate": 4.899150339472823e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8357633009552956,
      "num_tokens": 103430557.0,
      "step": 86020
    },
    {
      "entropy": 1.7336426332592965,
      "epoch": 0.26668547280253013,
      "grad_norm": 3.6474039554595947,
      "learning_rate": 4.898865592892199e-06,
      "loss": 0.4074,
      "mean_token_accuracy": 0.8620437294244766,
      "num_tokens": 103444908.0,
      "step": 86030
    },
    {
      "entropy": 1.8742715016007423,
      "epoch": 0.26671647192757986,
      "grad_norm": 9.908348083496094,
      "learning_rate": 4.8985808959556055e-06,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8436573952436447,
      "num_tokens": 103457439.0,
      "step": 86040
    },
    {
      "entropy": 1.8521887600421905,
      "epoch": 0.2667474710526295,
      "grad_norm": 9.044089317321777,
      "learning_rate": 4.8982962486486215e-06,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8412206932902336,
      "num_tokens": 103469444.0,
      "step": 86050
    },
    {
      "entropy": 1.892989605665207,
      "epoch": 0.26677847017767925,
      "grad_norm": 9.907114028930664,
      "learning_rate": 4.898011650956826e-06,
      "loss": 0.5118,
      "mean_token_accuracy": 0.8311768040060997,
      "num_tokens": 103480820.0,
      "step": 86060
    },
    {
      "entropy": 1.9290689766407012,
      "epoch": 0.2668094693027289,
      "grad_norm": 9.498516082763672,
      "learning_rate": 4.897727102865811e-06,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8411960631608963,
      "num_tokens": 103491988.0,
      "step": 86070
    },
    {
      "entropy": 1.857162345945835,
      "epoch": 0.26684046842777864,
      "grad_norm": 9.090229988098145,
      "learning_rate": 4.897442604361166e-06,
      "loss": 0.4517,
      "mean_token_accuracy": 0.8425956487655639,
      "num_tokens": 103504063.0,
      "step": 86080
    },
    {
      "entropy": 1.8890363931655885,
      "epoch": 0.2668714675528283,
      "grad_norm": 7.7313947677612305,
      "learning_rate": 4.8971581554284956e-06,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8431884482502937,
      "num_tokens": 103515289.0,
      "step": 86090
    },
    {
      "entropy": 1.7703155070543288,
      "epoch": 0.26690246667787804,
      "grad_norm": 9.580257415771484,
      "learning_rate": 4.896873756053401e-06,
      "loss": 0.4185,
      "mean_token_accuracy": 0.8482003748416901,
      "num_tokens": 103528699.0,
      "step": 86100
    },
    {
      "entropy": 1.8586272314190864,
      "epoch": 0.2669334658029277,
      "grad_norm": 8.929677963256836,
      "learning_rate": 4.8965894062214955e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8518487945199013,
      "num_tokens": 103540418.0,
      "step": 86110
    },
    {
      "entropy": 1.7248114220798016,
      "epoch": 0.26696446492797743,
      "grad_norm": 9.42658805847168,
      "learning_rate": 4.896305105918398e-06,
      "loss": 0.3427,
      "mean_token_accuracy": 0.8690110579133034,
      "num_tokens": 103554470.0,
      "step": 86120
    },
    {
      "entropy": 1.8832508057355881,
      "epoch": 0.2669954640530271,
      "grad_norm": 8.39008903503418,
      "learning_rate": 4.89602085512973e-06,
      "loss": 0.5325,
      "mean_token_accuracy": 0.8384171515703202,
      "num_tokens": 103566005.0,
      "step": 86130
    },
    {
      "entropy": 1.88730780929327,
      "epoch": 0.2670264631780768,
      "grad_norm": 7.347865104675293,
      "learning_rate": 4.895736653841122e-06,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8440589159727097,
      "num_tokens": 103577436.0,
      "step": 86140
    },
    {
      "entropy": 1.87120311409235,
      "epoch": 0.2670574623031265,
      "grad_norm": 10.774239540100098,
      "learning_rate": 4.895452502038206e-06,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8394436612725258,
      "num_tokens": 103589660.0,
      "step": 86150
    },
    {
      "entropy": 1.8455160409212112,
      "epoch": 0.2670884614281762,
      "grad_norm": 4.862356662750244,
      "learning_rate": 4.895168399706626e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8532706961035729,
      "num_tokens": 103601065.0,
      "step": 86160
    },
    {
      "entropy": 1.8480889692902565,
      "epoch": 0.2671194605532259,
      "grad_norm": 7.0500688552856445,
      "learning_rate": 4.894884346832027e-06,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8402373760938644,
      "num_tokens": 103613385.0,
      "step": 86170
    },
    {
      "entropy": 1.8862942337989808,
      "epoch": 0.26715045967827555,
      "grad_norm": 8.611431121826172,
      "learning_rate": 4.894600343400061e-06,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8492077127099037,
      "num_tokens": 103625382.0,
      "step": 86180
    },
    {
      "entropy": 1.8930604338645936,
      "epoch": 0.2671814588033253,
      "grad_norm": 7.554574966430664,
      "learning_rate": 4.894316389396388e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8508794024586678,
      "num_tokens": 103637191.0,
      "step": 86190
    },
    {
      "entropy": 1.921440924704075,
      "epoch": 0.26721245792837495,
      "grad_norm": 8.372522354125977,
      "learning_rate": 4.894032484806671e-06,
      "loss": 0.485,
      "mean_token_accuracy": 0.8466569393873214,
      "num_tokens": 103648987.0,
      "step": 86200
    },
    {
      "entropy": 1.9750257551670074,
      "epoch": 0.26724345705342467,
      "grad_norm": 6.741246700286865,
      "learning_rate": 4.893748629616579e-06,
      "loss": 0.5252,
      "mean_token_accuracy": 0.83692237585783,
      "num_tokens": 103659952.0,
      "step": 86210
    },
    {
      "entropy": 1.8079271107912063,
      "epoch": 0.26727445617847434,
      "grad_norm": 4.520543098449707,
      "learning_rate": 4.89346482381179e-06,
      "loss": 0.4497,
      "mean_token_accuracy": 0.851072619855404,
      "num_tokens": 103672555.0,
      "step": 86220
    },
    {
      "entropy": 1.9372185349464417,
      "epoch": 0.26730545530352406,
      "grad_norm": 8.379465103149414,
      "learning_rate": 4.8931810673779826e-06,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8367386668920517,
      "num_tokens": 103683329.0,
      "step": 86230
    },
    {
      "entropy": 1.8972872629761697,
      "epoch": 0.26733645442857373,
      "grad_norm": 8.481284141540527,
      "learning_rate": 4.8928973603008466e-06,
      "loss": 0.5154,
      "mean_token_accuracy": 0.836910292506218,
      "num_tokens": 103694708.0,
      "step": 86240
    },
    {
      "entropy": 1.8982973158359528,
      "epoch": 0.26736745355362346,
      "grad_norm": 9.208887100219727,
      "learning_rate": 4.892613702566074e-06,
      "loss": 0.528,
      "mean_token_accuracy": 0.8313050597906113,
      "num_tokens": 103707234.0,
      "step": 86250
    },
    {
      "entropy": 1.9922932714223862,
      "epoch": 0.2673984526786731,
      "grad_norm": 8.469325065612793,
      "learning_rate": 4.892330094159364e-06,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8502902209758758,
      "num_tokens": 103717727.0,
      "step": 86260
    },
    {
      "entropy": 1.8680451080203055,
      "epoch": 0.26742945180372285,
      "grad_norm": 9.319469451904297,
      "learning_rate": 4.892046535066422e-06,
      "loss": 0.4582,
      "mean_token_accuracy": 0.844627107679844,
      "num_tokens": 103729854.0,
      "step": 86270
    },
    {
      "entropy": 1.9178496971726418,
      "epoch": 0.2674604509287725,
      "grad_norm": 3.9577114582061768,
      "learning_rate": 4.891763025272957e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8466630190610885,
      "num_tokens": 103742121.0,
      "step": 86280
    },
    {
      "entropy": 1.9498088628053665,
      "epoch": 0.26749145005382224,
      "grad_norm": 7.915012359619141,
      "learning_rate": 4.891479564764686e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8389876633882523,
      "num_tokens": 103753077.0,
      "step": 86290
    },
    {
      "entropy": 1.8598709747195243,
      "epoch": 0.2675224491788719,
      "grad_norm": 5.166788101196289,
      "learning_rate": 4.891196153527332e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8542357131838798,
      "num_tokens": 103765572.0,
      "step": 86300
    },
    {
      "entropy": 1.87834425419569,
      "epoch": 0.26755344830392164,
      "grad_norm": 5.004603862762451,
      "learning_rate": 4.890912791546621e-06,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8426024809479713,
      "num_tokens": 103777759.0,
      "step": 86310
    },
    {
      "entropy": 1.9219666391611099,
      "epoch": 0.2675844474289713,
      "grad_norm": 7.858938694000244,
      "learning_rate": 4.8906294788082895e-06,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8506090447306633,
      "num_tokens": 103790518.0,
      "step": 86320
    },
    {
      "entropy": 1.9264658272266388,
      "epoch": 0.26761544655402103,
      "grad_norm": 8.828166007995605,
      "learning_rate": 4.890346215298074e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8476721942424774,
      "num_tokens": 103801617.0,
      "step": 86330
    },
    {
      "entropy": 1.8663738921284676,
      "epoch": 0.2676464456790707,
      "grad_norm": 8.534181594848633,
      "learning_rate": 4.890063001001723e-06,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8441844269633293,
      "num_tokens": 103814001.0,
      "step": 86340
    },
    {
      "entropy": 1.9115709066390991,
      "epoch": 0.2676774448041204,
      "grad_norm": 3.800837278366089,
      "learning_rate": 4.889779835904984e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8440373882651329,
      "num_tokens": 103826239.0,
      "step": 86350
    },
    {
      "entropy": 1.9562127739191055,
      "epoch": 0.2677084439291701,
      "grad_norm": 9.03986930847168,
      "learning_rate": 4.889496719993616e-06,
      "loss": 0.5581,
      "mean_token_accuracy": 0.8341035321354866,
      "num_tokens": 103837216.0,
      "step": 86360
    },
    {
      "entropy": 1.8933446779847145,
      "epoch": 0.2677394430542198,
      "grad_norm": 8.497458457946777,
      "learning_rate": 4.889213653253382e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8447739273309708,
      "num_tokens": 103849512.0,
      "step": 86370
    },
    {
      "entropy": 1.940016995370388,
      "epoch": 0.2677704421792695,
      "grad_norm": 9.415840148925781,
      "learning_rate": 4.88893063567005e-06,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8498956128954888,
      "num_tokens": 103860985.0,
      "step": 86380
    },
    {
      "entropy": 1.9313148841261865,
      "epoch": 0.2678014413043192,
      "grad_norm": 8.689431190490723,
      "learning_rate": 4.888647667229392e-06,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8414283052086831,
      "num_tokens": 103872780.0,
      "step": 86390
    },
    {
      "entropy": 1.9360795393586159,
      "epoch": 0.2678324404293689,
      "grad_norm": 9.888842582702637,
      "learning_rate": 4.888364747917191e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8488156452775002,
      "num_tokens": 103884056.0,
      "step": 86400
    },
    {
      "entropy": 1.8884512677788734,
      "epoch": 0.2678634395544186,
      "grad_norm": 9.365774154663086,
      "learning_rate": 4.888081877719231e-06,
      "loss": 0.489,
      "mean_token_accuracy": 0.8477799639105796,
      "num_tokens": 103895788.0,
      "step": 86410
    },
    {
      "entropy": 1.8175974920392037,
      "epoch": 0.26789443867946827,
      "grad_norm": 8.061196327209473,
      "learning_rate": 4.887799056621303e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8709731310606003,
      "num_tokens": 103908055.0,
      "step": 86420
    },
    {
      "entropy": 1.815473848581314,
      "epoch": 0.26792543780451794,
      "grad_norm": 8.550416946411133,
      "learning_rate": 4.887516284609206e-06,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8482015028595924,
      "num_tokens": 103921019.0,
      "step": 86430
    },
    {
      "entropy": 1.8964615538716316,
      "epoch": 0.26795643692956767,
      "grad_norm": 7.224480152130127,
      "learning_rate": 4.887233561668741e-06,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8473669067025185,
      "num_tokens": 103932376.0,
      "step": 86440
    },
    {
      "entropy": 1.8842689141631126,
      "epoch": 0.26798743605461733,
      "grad_norm": 3.8930001258850098,
      "learning_rate": 4.886950887785717e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8496074840426445,
      "num_tokens": 103944476.0,
      "step": 86450
    },
    {
      "entropy": 1.946980032324791,
      "epoch": 0.26801843517966706,
      "grad_norm": 9.502989768981934,
      "learning_rate": 4.886668262945951e-06,
      "loss": 0.5331,
      "mean_token_accuracy": 0.8362909764051437,
      "num_tokens": 103955653.0,
      "step": 86460
    },
    {
      "entropy": 1.9322603926062585,
      "epoch": 0.2680494343047167,
      "grad_norm": 8.219178199768066,
      "learning_rate": 4.886385687135257e-06,
      "loss": 0.5317,
      "mean_token_accuracy": 0.8409213215112686,
      "num_tokens": 103967278.0,
      "step": 86470
    },
    {
      "entropy": 1.9176155909895898,
      "epoch": 0.26808043342976645,
      "grad_norm": 6.983169078826904,
      "learning_rate": 4.886103160339469e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8338638916611671,
      "num_tokens": 103979747.0,
      "step": 86480
    },
    {
      "entropy": 1.89949888586998,
      "epoch": 0.2681114325548161,
      "grad_norm": 8.026739120483398,
      "learning_rate": 4.885820682544414e-06,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8553324237465858,
      "num_tokens": 103991452.0,
      "step": 86490
    },
    {
      "entropy": 1.9148534148931504,
      "epoch": 0.26814243167986584,
      "grad_norm": 8.056224822998047,
      "learning_rate": 4.885538253735928e-06,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8458040565252304,
      "num_tokens": 104003229.0,
      "step": 86500
    },
    {
      "entropy": 1.9580560460686685,
      "epoch": 0.2681734308049155,
      "grad_norm": 9.577115058898926,
      "learning_rate": 4.885255873899857e-06,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8413164436817169,
      "num_tokens": 104014372.0,
      "step": 86510
    },
    {
      "entropy": 1.8256202667951584,
      "epoch": 0.26820442992996524,
      "grad_norm": 4.1934733390808105,
      "learning_rate": 4.884973543022048e-06,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8329330936074257,
      "num_tokens": 104027131.0,
      "step": 86520
    },
    {
      "entropy": 1.9278512462973594,
      "epoch": 0.2682354290550149,
      "grad_norm": 6.770676612854004,
      "learning_rate": 4.884691261088359e-06,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8434366106986999,
      "num_tokens": 104038489.0,
      "step": 86530
    },
    {
      "entropy": 1.7697638273239136,
      "epoch": 0.26826642818006463,
      "grad_norm": 9.314817428588867,
      "learning_rate": 4.884409028084645e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8589093953371048,
      "num_tokens": 104051697.0,
      "step": 86540
    },
    {
      "entropy": 1.8212177708745003,
      "epoch": 0.2682974273051143,
      "grad_norm": 8.558985710144043,
      "learning_rate": 4.8841268439967744e-06,
      "loss": 0.3656,
      "mean_token_accuracy": 0.8659472689032555,
      "num_tokens": 104064685.0,
      "step": 86550
    },
    {
      "entropy": 1.8339843198657035,
      "epoch": 0.268328426430164,
      "grad_norm": 5.08605432510376,
      "learning_rate": 4.883844708810621e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8505037307739258,
      "num_tokens": 104077583.0,
      "step": 86560
    },
    {
      "entropy": 1.9253122627735137,
      "epoch": 0.2683594255552137,
      "grad_norm": 7.690542697906494,
      "learning_rate": 4.883562622512059e-06,
      "loss": 0.522,
      "mean_token_accuracy": 0.8420386493206025,
      "num_tokens": 104088267.0,
      "step": 86570
    },
    {
      "entropy": 1.8858140379190445,
      "epoch": 0.2683904246802634,
      "grad_norm": 8.237641334533691,
      "learning_rate": 4.883280585086974e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8476707622408867,
      "num_tokens": 104099930.0,
      "step": 86580
    },
    {
      "entropy": 1.858066162467003,
      "epoch": 0.2684214238053131,
      "grad_norm": 8.798260688781738,
      "learning_rate": 4.882998596521253e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8484656035900116,
      "num_tokens": 104111901.0,
      "step": 86590
    },
    {
      "entropy": 1.92180934548378,
      "epoch": 0.2684524229303628,
      "grad_norm": 9.742270469665527,
      "learning_rate": 4.882716656800792e-06,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8452163085341453,
      "num_tokens": 104122977.0,
      "step": 86600
    },
    {
      "entropy": 1.867686577141285,
      "epoch": 0.2684834220554125,
      "grad_norm": 10.891302108764648,
      "learning_rate": 4.882434765911489e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8573295056819916,
      "num_tokens": 104135141.0,
      "step": 86610
    },
    {
      "entropy": 1.9070853009819984,
      "epoch": 0.2685144211804622,
      "grad_norm": 6.908443450927734,
      "learning_rate": 4.882152923839252e-06,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8457126155495643,
      "num_tokens": 104146361.0,
      "step": 86620
    },
    {
      "entropy": 1.9306767612695694,
      "epoch": 0.2685454203055119,
      "grad_norm": 8.809544563293457,
      "learning_rate": 4.881871130569993e-06,
      "loss": 0.4938,
      "mean_token_accuracy": 0.831186157464981,
      "num_tokens": 104158055.0,
      "step": 86630
    },
    {
      "entropy": 1.7939895704388618,
      "epoch": 0.2685764194305616,
      "grad_norm": 6.0357184410095215,
      "learning_rate": 4.8815893860896265e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.8565925478935241,
      "num_tokens": 104171215.0,
      "step": 86640
    },
    {
      "entropy": 1.8932900041341783,
      "epoch": 0.26860741855561127,
      "grad_norm": 9.490089416503906,
      "learning_rate": 4.881307690384079e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.851920661330223,
      "num_tokens": 104183346.0,
      "step": 86650
    },
    {
      "entropy": 1.7981626734137535,
      "epoch": 0.268638417680661,
      "grad_norm": 4.065007209777832,
      "learning_rate": 4.881026043439277e-06,
      "loss": 0.4316,
      "mean_token_accuracy": 0.850606782734394,
      "num_tokens": 104197135.0,
      "step": 86660
    },
    {
      "entropy": 1.9110074192285538,
      "epoch": 0.26866941680571066,
      "grad_norm": 8.759778022766113,
      "learning_rate": 4.880744445241155e-06,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8564579889178277,
      "num_tokens": 104207822.0,
      "step": 86670
    },
    {
      "entropy": 1.980569313466549,
      "epoch": 0.26870041593076033,
      "grad_norm": 8.87386703491211,
      "learning_rate": 4.880462895775654e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8343950539827347,
      "num_tokens": 104219081.0,
      "step": 86680
    },
    {
      "entropy": 1.8610002383589745,
      "epoch": 0.26873141505581005,
      "grad_norm": 8.677059173583984,
      "learning_rate": 4.880181395028719e-06,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8492666438221932,
      "num_tokens": 104231630.0,
      "step": 86690
    },
    {
      "entropy": 1.899726065993309,
      "epoch": 0.2687624141808597,
      "grad_norm": 8.745160102844238,
      "learning_rate": 4.879899942986303e-06,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8413066044449806,
      "num_tokens": 104242889.0,
      "step": 86700
    },
    {
      "entropy": 1.8926833271980286,
      "epoch": 0.26879341330590945,
      "grad_norm": 3.9941442012786865,
      "learning_rate": 4.87961853963436e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8299556851387024,
      "num_tokens": 104254389.0,
      "step": 86710
    },
    {
      "entropy": 1.875252665579319,
      "epoch": 0.2688244124309591,
      "grad_norm": 8.266586303710938,
      "learning_rate": 4.879337184958854e-06,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8451120510697365,
      "num_tokens": 104266592.0,
      "step": 86720
    },
    {
      "entropy": 1.8953094065189362,
      "epoch": 0.26885541155600884,
      "grad_norm": 8.299905776977539,
      "learning_rate": 4.8790558789457545e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8386603578925133,
      "num_tokens": 104278272.0,
      "step": 86730
    },
    {
      "entropy": 1.8657784268260003,
      "epoch": 0.2688864106810585,
      "grad_norm": 8.039691925048828,
      "learning_rate": 4.878774621581035e-06,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8498806357383728,
      "num_tokens": 104290052.0,
      "step": 86740
    },
    {
      "entropy": 1.924768103659153,
      "epoch": 0.26891740980610823,
      "grad_norm": 8.882854461669922,
      "learning_rate": 4.878493412850675e-06,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8329790845513344,
      "num_tokens": 104301303.0,
      "step": 86750
    },
    {
      "entropy": 1.8597337052226066,
      "epoch": 0.2689484089311579,
      "grad_norm": 8.194012641906738,
      "learning_rate": 4.878212252740661e-06,
      "loss": 0.4851,
      "mean_token_accuracy": 0.8411440491676331,
      "num_tokens": 104313725.0,
      "step": 86760
    },
    {
      "entropy": 1.955818921327591,
      "epoch": 0.2689794080562076,
      "grad_norm": 8.100213050842285,
      "learning_rate": 4.877931141236982e-06,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8334660112857819,
      "num_tokens": 104325362.0,
      "step": 86770
    },
    {
      "entropy": 1.8957918226718902,
      "epoch": 0.2690104071812573,
      "grad_norm": 5.112435340881348,
      "learning_rate": 4.877650078325635e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8381680518388748,
      "num_tokens": 104337583.0,
      "step": 86780
    },
    {
      "entropy": 1.9326987490057945,
      "epoch": 0.269041406306307,
      "grad_norm": 8.216394424438477,
      "learning_rate": 4.8773690639926246e-06,
      "loss": 0.4806,
      "mean_token_accuracy": 0.8478187531232834,
      "num_tokens": 104349160.0,
      "step": 86790
    },
    {
      "entropy": 1.8107408866286279,
      "epoch": 0.2690724054313567,
      "grad_norm": 8.589932441711426,
      "learning_rate": 4.8770880982239565e-06,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8569507896900177,
      "num_tokens": 104362032.0,
      "step": 86800
    },
    {
      "entropy": 1.9010547578334809,
      "epoch": 0.2691034045564064,
      "grad_norm": 7.747185707092285,
      "learning_rate": 4.876807181005645e-06,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8419987246394157,
      "num_tokens": 104374183.0,
      "step": 86810
    },
    {
      "entropy": 1.9327505372464657,
      "epoch": 0.2691344036814561,
      "grad_norm": 7.296648025512695,
      "learning_rate": 4.87652631232371e-06,
      "loss": 0.5457,
      "mean_token_accuracy": 0.8373480141162872,
      "num_tokens": 104385704.0,
      "step": 86820
    },
    {
      "entropy": 1.9187294945120812,
      "epoch": 0.2691654028065058,
      "grad_norm": 9.021369934082031,
      "learning_rate": 4.876245492164175e-06,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8350897148251534,
      "num_tokens": 104397570.0,
      "step": 86830
    },
    {
      "entropy": 1.8674516141414643,
      "epoch": 0.2691964019315555,
      "grad_norm": 8.63692855834961,
      "learning_rate": 4.875964720513072e-06,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8315334841609001,
      "num_tokens": 104409275.0,
      "step": 86840
    },
    {
      "entropy": 1.8427454948425293,
      "epoch": 0.2692274010566052,
      "grad_norm": 3.408676862716675,
      "learning_rate": 4.875683997356437e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.85171507447958,
      "num_tokens": 104421885.0,
      "step": 86850
    },
    {
      "entropy": 1.9858239054679871,
      "epoch": 0.26925840018165487,
      "grad_norm": 8.626346588134766,
      "learning_rate": 4.87540332268031e-06,
      "loss": 0.6016,
      "mean_token_accuracy": 0.8261340126395226,
      "num_tokens": 104433208.0,
      "step": 86860
    },
    {
      "entropy": 1.8474789157509803,
      "epoch": 0.2692893993067046,
      "grad_norm": 3.6940155029296875,
      "learning_rate": 4.87512269647074e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8468289896845818,
      "num_tokens": 104445950.0,
      "step": 86870
    },
    {
      "entropy": 1.7919086948037148,
      "epoch": 0.26932039843175426,
      "grad_norm": 7.589000225067139,
      "learning_rate": 4.8748421187137786e-06,
      "loss": 0.4006,
      "mean_token_accuracy": 0.8585388869047165,
      "num_tokens": 104459408.0,
      "step": 86880
    },
    {
      "entropy": 1.817798225581646,
      "epoch": 0.269351397556804,
      "grad_norm": 2.4566121101379395,
      "learning_rate": 4.8745615893954875e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.856153316795826,
      "num_tokens": 104472083.0,
      "step": 86890
    },
    {
      "entropy": 1.829006166756153,
      "epoch": 0.26938239668185365,
      "grad_norm": 3.47737455368042,
      "learning_rate": 4.8742811085019294e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8479893788695335,
      "num_tokens": 104485313.0,
      "step": 86900
    },
    {
      "entropy": 1.824818679690361,
      "epoch": 0.2694133958069034,
      "grad_norm": 10.421551704406738,
      "learning_rate": 4.8740006760191715e-06,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8599545955657959,
      "num_tokens": 104498290.0,
      "step": 86910
    },
    {
      "entropy": 1.8766320884227752,
      "epoch": 0.26944439493195305,
      "grad_norm": 8.567687034606934,
      "learning_rate": 4.873720291933294e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8494315207004547,
      "num_tokens": 104510307.0,
      "step": 86920
    },
    {
      "entropy": 1.8442302539944648,
      "epoch": 0.2694753940570027,
      "grad_norm": 3.7030580043792725,
      "learning_rate": 4.873439956230375e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.848169319331646,
      "num_tokens": 104522429.0,
      "step": 86930
    },
    {
      "entropy": 1.890008282661438,
      "epoch": 0.26950639318205244,
      "grad_norm": 8.97164249420166,
      "learning_rate": 4.873159668896501e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8344573676586151,
      "num_tokens": 104534058.0,
      "step": 86940
    },
    {
      "entropy": 1.818005445599556,
      "epoch": 0.2695373923071021,
      "grad_norm": 8.09427547454834,
      "learning_rate": 4.8728794299177655e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8522487103939056,
      "num_tokens": 104546415.0,
      "step": 86950
    },
    {
      "entropy": 1.838931292295456,
      "epoch": 0.26956839143215183,
      "grad_norm": 7.319267272949219,
      "learning_rate": 4.8725992392802655e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8448920026421547,
      "num_tokens": 104558731.0,
      "step": 86960
    },
    {
      "entropy": 1.8411687076091767,
      "epoch": 0.2695993905572015,
      "grad_norm": 3.573763847351074,
      "learning_rate": 4.872319096970106e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8461699530482292,
      "num_tokens": 104571213.0,
      "step": 86970
    },
    {
      "entropy": 1.8673091575503349,
      "epoch": 0.2696303896822512,
      "grad_norm": 4.311919689178467,
      "learning_rate": 4.872039002973394e-06,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8393310904502869,
      "num_tokens": 104583049.0,
      "step": 86980
    },
    {
      "entropy": 1.8796320587396622,
      "epoch": 0.2696613888073009,
      "grad_norm": 9.332188606262207,
      "learning_rate": 4.871758957276246e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8384488835930825,
      "num_tokens": 104595379.0,
      "step": 86990
    },
    {
      "entropy": 1.8900238052010536,
      "epoch": 0.2696923879323506,
      "grad_norm": 7.770622253417969,
      "learning_rate": 4.871478959864781e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8497345179319382,
      "num_tokens": 104606394.0,
      "step": 87000
    },
    {
      "entropy": 1.839431057870388,
      "epoch": 0.2697233870574003,
      "grad_norm": 8.380780220031738,
      "learning_rate": 4.871199010725126e-06,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8396735802292824,
      "num_tokens": 104619650.0,
      "step": 87010
    },
    {
      "entropy": 1.827296996116638,
      "epoch": 0.26975438618245,
      "grad_norm": 8.714934349060059,
      "learning_rate": 4.870919109843412e-06,
      "loss": 0.509,
      "mean_token_accuracy": 0.8432153165340424,
      "num_tokens": 104632127.0,
      "step": 87020
    },
    {
      "entropy": 1.8935890957713126,
      "epoch": 0.2697853853074997,
      "grad_norm": 9.051390647888184,
      "learning_rate": 4.870639257205774e-06,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8443873509764671,
      "num_tokens": 104643478.0,
      "step": 87030
    },
    {
      "entropy": 1.8204589426517486,
      "epoch": 0.2698163844325494,
      "grad_norm": 7.4907379150390625,
      "learning_rate": 4.870359452798357e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8611741289496422,
      "num_tokens": 104655632.0,
      "step": 87040
    },
    {
      "entropy": 1.8853262856602668,
      "epoch": 0.2698473835575991,
      "grad_norm": 8.072641372680664,
      "learning_rate": 4.8700796966073084e-06,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8454017639160156,
      "num_tokens": 104667018.0,
      "step": 87050
    },
    {
      "entropy": 1.8324878126382829,
      "epoch": 0.2698783826826488,
      "grad_norm": 8.132426261901855,
      "learning_rate": 4.869799988618784e-06,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8415536895394325,
      "num_tokens": 104679052.0,
      "step": 87060
    },
    {
      "entropy": 1.884521934390068,
      "epoch": 0.26990938180769847,
      "grad_norm": 9.089740753173828,
      "learning_rate": 4.869520328818938e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8491548255085946,
      "num_tokens": 104690472.0,
      "step": 87070
    },
    {
      "entropy": 1.7616024523973466,
      "epoch": 0.2699403809327482,
      "grad_norm": 5.119749546051025,
      "learning_rate": 4.86924071719394e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8421557962894439,
      "num_tokens": 104703935.0,
      "step": 87080
    },
    {
      "entropy": 1.8466105580329895,
      "epoch": 0.26997138005779786,
      "grad_norm": 6.891587734222412,
      "learning_rate": 4.868961153729958e-06,
      "loss": 0.5664,
      "mean_token_accuracy": 0.8311922624707222,
      "num_tokens": 104715762.0,
      "step": 87090
    },
    {
      "entropy": 1.830435362458229,
      "epoch": 0.2700023791828476,
      "grad_norm": 9.212553024291992,
      "learning_rate": 4.86868163841317e-06,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8380068346858025,
      "num_tokens": 104727912.0,
      "step": 87100
    },
    {
      "entropy": 1.8718600660562514,
      "epoch": 0.27003337830789725,
      "grad_norm": 8.763904571533203,
      "learning_rate": 4.8684021712297545e-06,
      "loss": 0.485,
      "mean_token_accuracy": 0.8420531466603279,
      "num_tokens": 104740144.0,
      "step": 87110
    },
    {
      "entropy": 1.8686859756708145,
      "epoch": 0.270064377432947,
      "grad_norm": 4.998740196228027,
      "learning_rate": 4.868122752165901e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8439597055315972,
      "num_tokens": 104751419.0,
      "step": 87120
    },
    {
      "entropy": 1.821141104400158,
      "epoch": 0.27009537655799665,
      "grad_norm": 8.166898727416992,
      "learning_rate": 4.867843381207802e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8448257312178612,
      "num_tokens": 104764014.0,
      "step": 87130
    },
    {
      "entropy": 1.8268358632922173,
      "epoch": 0.27012637568304637,
      "grad_norm": 9.196722984313965,
      "learning_rate": 4.867564058341654e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8561849161982537,
      "num_tokens": 104776029.0,
      "step": 87140
    },
    {
      "entropy": 1.8853852570056915,
      "epoch": 0.27015737480809604,
      "grad_norm": 7.830870151519775,
      "learning_rate": 4.867284783553663e-06,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8442418292164803,
      "num_tokens": 104787139.0,
      "step": 87150
    },
    {
      "entropy": 1.7887479558587074,
      "epoch": 0.27018837393314576,
      "grad_norm": 8.097960472106934,
      "learning_rate": 4.867005556830035e-06,
      "loss": 0.3924,
      "mean_token_accuracy": 0.8651418328285218,
      "num_tokens": 104799482.0,
      "step": 87160
    },
    {
      "entropy": 1.8204817980527879,
      "epoch": 0.27021937305819543,
      "grad_norm": 8.041654586791992,
      "learning_rate": 4.8667263781569875e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8555315598845482,
      "num_tokens": 104811350.0,
      "step": 87170
    },
    {
      "entropy": 1.7814991921186447,
      "epoch": 0.2702503721832451,
      "grad_norm": 7.963624477386475,
      "learning_rate": 4.86644724752074e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8552462846040726,
      "num_tokens": 104824216.0,
      "step": 87180
    },
    {
      "entropy": 1.9131552398204803,
      "epoch": 0.2702813713082948,
      "grad_norm": 9.447894096374512,
      "learning_rate": 4.86616816490752e-06,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8396668806672096,
      "num_tokens": 104835560.0,
      "step": 87190
    },
    {
      "entropy": 1.824287013709545,
      "epoch": 0.2703123704333445,
      "grad_norm": 9.329700469970703,
      "learning_rate": 4.865889130303556e-06,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8538261905312539,
      "num_tokens": 104847449.0,
      "step": 87200
    },
    {
      "entropy": 1.8483013391494751,
      "epoch": 0.2703433695583942,
      "grad_norm": 8.532864570617676,
      "learning_rate": 4.865610143695086e-06,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8494655951857567,
      "num_tokens": 104859220.0,
      "step": 87210
    },
    {
      "entropy": 1.8401647135615349,
      "epoch": 0.2703743686834439,
      "grad_norm": 7.677708148956299,
      "learning_rate": 4.8653312050683524e-06,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8437712222337723,
      "num_tokens": 104870632.0,
      "step": 87220
    },
    {
      "entropy": 1.8975766867399215,
      "epoch": 0.2704053678084936,
      "grad_norm": 7.6379780769348145,
      "learning_rate": 4.865052314409605e-06,
      "loss": 0.5677,
      "mean_token_accuracy": 0.8343770399689674,
      "num_tokens": 104881139.0,
      "step": 87230
    },
    {
      "entropy": 1.842423902451992,
      "epoch": 0.2704363669335433,
      "grad_norm": 8.821908950805664,
      "learning_rate": 4.864773471705094e-06,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8355495795607567,
      "num_tokens": 104893189.0,
      "step": 87240
    },
    {
      "entropy": 1.8596946865320205,
      "epoch": 0.270467366058593,
      "grad_norm": 8.274313926696777,
      "learning_rate": 4.86449467694108e-06,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8518381923437118,
      "num_tokens": 104904842.0,
      "step": 87250
    },
    {
      "entropy": 1.7499984815716743,
      "epoch": 0.2704983651836427,
      "grad_norm": 8.799256324768066,
      "learning_rate": 4.864215930103828e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8650006785988807,
      "num_tokens": 104918602.0,
      "step": 87260
    },
    {
      "entropy": 1.8299191161990165,
      "epoch": 0.2705293643086924,
      "grad_norm": 8.756028175354004,
      "learning_rate": 4.863937231179608e-06,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8430431827902793,
      "num_tokens": 104930274.0,
      "step": 87270
    },
    {
      "entropy": 1.8168650731444358,
      "epoch": 0.27056036343374207,
      "grad_norm": 8.896665573120117,
      "learning_rate": 4.863658580154694e-06,
      "loss": 0.4216,
      "mean_token_accuracy": 0.8473124504089355,
      "num_tokens": 104943125.0,
      "step": 87280
    },
    {
      "entropy": 1.8909821808338165,
      "epoch": 0.2705913625587918,
      "grad_norm": 11.013216018676758,
      "learning_rate": 4.863379977015369e-06,
      "loss": 0.5405,
      "mean_token_accuracy": 0.8345570892095566,
      "num_tokens": 104954178.0,
      "step": 87290
    },
    {
      "entropy": 1.8781416594982148,
      "epoch": 0.27062236168384146,
      "grad_norm": 8.503053665161133,
      "learning_rate": 4.863101421747918e-06,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8471224218606949,
      "num_tokens": 104965323.0,
      "step": 87300
    },
    {
      "entropy": 1.7377824038267136,
      "epoch": 0.2706533608088912,
      "grad_norm": 3.2067818641662598,
      "learning_rate": 4.862822914338635e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8619665712118149,
      "num_tokens": 104979529.0,
      "step": 87310
    },
    {
      "entropy": 1.8459437713027,
      "epoch": 0.27068435993394085,
      "grad_norm": 9.494819641113281,
      "learning_rate": 4.862544454773815e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8475456729531288,
      "num_tokens": 104991778.0,
      "step": 87320
    },
    {
      "entropy": 1.8079747796058654,
      "epoch": 0.2707153590589906,
      "grad_norm": 4.102128505706787,
      "learning_rate": 4.86226604303976e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.8489979475736618,
      "num_tokens": 105004887.0,
      "step": 87330
    },
    {
      "entropy": 1.8566383227705956,
      "epoch": 0.27074635818404025,
      "grad_norm": 9.459664344787598,
      "learning_rate": 4.8619876791227834e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.8472496911883354,
      "num_tokens": 105015787.0,
      "step": 87340
    },
    {
      "entropy": 1.8375443920493126,
      "epoch": 0.27077735730908997,
      "grad_norm": 8.758172988891602,
      "learning_rate": 4.861709363009195e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8570204228162766,
      "num_tokens": 105027235.0,
      "step": 87350
    },
    {
      "entropy": 1.8649962782859801,
      "epoch": 0.27080835643413964,
      "grad_norm": 9.725914001464844,
      "learning_rate": 4.861431094685316e-06,
      "loss": 0.5095,
      "mean_token_accuracy": 0.8399304628372193,
      "num_tokens": 105038434.0,
      "step": 87360
    },
    {
      "entropy": 1.7663211345672607,
      "epoch": 0.27083935555918937,
      "grad_norm": 7.783099174499512,
      "learning_rate": 4.86115287413747e-06,
      "loss": 0.4074,
      "mean_token_accuracy": 0.8482378587126732,
      "num_tokens": 105050998.0,
      "step": 87370
    },
    {
      "entropy": 1.8368576869368554,
      "epoch": 0.27087035468423903,
      "grad_norm": 9.434002876281738,
      "learning_rate": 4.8608747013519896e-06,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8487260073423386,
      "num_tokens": 105062505.0,
      "step": 87380
    },
    {
      "entropy": 1.7946744754910469,
      "epoch": 0.27090135380928876,
      "grad_norm": 8.645105361938477,
      "learning_rate": 4.860596576315209e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8437353923916817,
      "num_tokens": 105075561.0,
      "step": 87390
    },
    {
      "entropy": 1.8670944333076478,
      "epoch": 0.2709323529343384,
      "grad_norm": 8.151835441589355,
      "learning_rate": 4.860318499013468e-06,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8455032199621201,
      "num_tokens": 105087761.0,
      "step": 87400
    },
    {
      "entropy": 1.8563807129859924,
      "epoch": 0.2709633520593881,
      "grad_norm": 8.382278442382812,
      "learning_rate": 4.860040469433119e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8437006428837777,
      "num_tokens": 105099328.0,
      "step": 87410
    },
    {
      "entropy": 1.8392608642578125,
      "epoch": 0.2709943511844378,
      "grad_norm": 6.8433098793029785,
      "learning_rate": 4.8597624875605076e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8475878596305847,
      "num_tokens": 105111492.0,
      "step": 87420
    },
    {
      "entropy": 1.896904969215393,
      "epoch": 0.2710253503094875,
      "grad_norm": 9.722173690795898,
      "learning_rate": 4.859484553381996e-06,
      "loss": 0.5157,
      "mean_token_accuracy": 0.8329481184482574,
      "num_tokens": 105123552.0,
      "step": 87430
    },
    {
      "entropy": 1.8754973500967025,
      "epoch": 0.2710563494345372,
      "grad_norm": 3.6622695922851562,
      "learning_rate": 4.859206666883946e-06,
      "loss": 0.4939,
      "mean_token_accuracy": 0.8413586497306824,
      "num_tokens": 105135820.0,
      "step": 87440
    },
    {
      "entropy": 1.8094146370887756,
      "epoch": 0.2710873485595869,
      "grad_norm": 9.525456428527832,
      "learning_rate": 4.858928828052725e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8486891463398933,
      "num_tokens": 105147391.0,
      "step": 87450
    },
    {
      "entropy": 1.9204721599817276,
      "epoch": 0.2711183476846366,
      "grad_norm": 8.702652931213379,
      "learning_rate": 4.858651036874711e-06,
      "loss": 0.5346,
      "mean_token_accuracy": 0.837224043905735,
      "num_tokens": 105159245.0,
      "step": 87460
    },
    {
      "entropy": 1.9499018788337708,
      "epoch": 0.2711493468096863,
      "grad_norm": 9.13373851776123,
      "learning_rate": 4.858373293336278e-06,
      "loss": 0.5221,
      "mean_token_accuracy": 0.8495534911751748,
      "num_tokens": 105169906.0,
      "step": 87470
    },
    {
      "entropy": 1.865763219445944,
      "epoch": 0.271180345934736,
      "grad_norm": 8.000102043151855,
      "learning_rate": 4.858095597423816e-06,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8466974958777428,
      "num_tokens": 105182696.0,
      "step": 87480
    },
    {
      "entropy": 1.8530535832047463,
      "epoch": 0.27121134505978567,
      "grad_norm": 9.345272064208984,
      "learning_rate": 4.8578179491237135e-06,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8383201539516449,
      "num_tokens": 105194697.0,
      "step": 87490
    },
    {
      "entropy": 1.9272264629602431,
      "epoch": 0.2712423441848354,
      "grad_norm": 8.332130432128906,
      "learning_rate": 4.857540348422365e-06,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8568297758698463,
      "num_tokens": 105205400.0,
      "step": 87500
    },
    {
      "entropy": 1.8942699432373047,
      "epoch": 0.27127334330988506,
      "grad_norm": 8.65912914276123,
      "learning_rate": 4.857262795306176e-06,
      "loss": 0.4641,
      "mean_token_accuracy": 0.8520406141877175,
      "num_tokens": 105217304.0,
      "step": 87510
    },
    {
      "entropy": 1.7813436336815358,
      "epoch": 0.2713043424349348,
      "grad_norm": 4.422085762023926,
      "learning_rate": 4.8569852897615476e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8593872472643852,
      "num_tokens": 105230048.0,
      "step": 87520
    },
    {
      "entropy": 1.9077577859163284,
      "epoch": 0.27133534155998446,
      "grad_norm": 8.263802528381348,
      "learning_rate": 4.856707831774897e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8396870777010917,
      "num_tokens": 105241231.0,
      "step": 87530
    },
    {
      "entropy": 1.8345076471567154,
      "epoch": 0.2713663406850342,
      "grad_norm": 8.313085556030273,
      "learning_rate": 4.856430421332639e-06,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8465406790375709,
      "num_tokens": 105253356.0,
      "step": 87540
    },
    {
      "entropy": 1.8547306582331657,
      "epoch": 0.27139733981008385,
      "grad_norm": 7.723386287689209,
      "learning_rate": 4.856153058421199e-06,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8484765499830246,
      "num_tokens": 105265696.0,
      "step": 87550
    },
    {
      "entropy": 1.8216210514307023,
      "epoch": 0.2714283389351336,
      "grad_norm": 7.835054874420166,
      "learning_rate": 4.855875743027003e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8526194721460343,
      "num_tokens": 105277875.0,
      "step": 87560
    },
    {
      "entropy": 1.8399008169770241,
      "epoch": 0.27145933806018324,
      "grad_norm": 7.611944198608398,
      "learning_rate": 4.855598475136486e-06,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8431843638420105,
      "num_tokens": 105290058.0,
      "step": 87570
    },
    {
      "entropy": 1.8355013683438302,
      "epoch": 0.27149033718523297,
      "grad_norm": 8.615690231323242,
      "learning_rate": 4.855321254736087e-06,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8420334681868553,
      "num_tokens": 105302569.0,
      "step": 87580
    },
    {
      "entropy": 1.8711734786629677,
      "epoch": 0.27152133631028263,
      "grad_norm": 8.991096496582031,
      "learning_rate": 4.855044081812253e-06,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8353176578879357,
      "num_tokens": 105314367.0,
      "step": 87590
    },
    {
      "entropy": 1.8946691662073136,
      "epoch": 0.27155233543533236,
      "grad_norm": 7.5710577964782715,
      "learning_rate": 4.854766956351432e-06,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8443698287010193,
      "num_tokens": 105325617.0,
      "step": 87600
    },
    {
      "entropy": 1.8058865994215012,
      "epoch": 0.27158333456038203,
      "grad_norm": 2.2330517768859863,
      "learning_rate": 4.854489878340079e-06,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8525756880640983,
      "num_tokens": 105338254.0,
      "step": 87610
    },
    {
      "entropy": 1.9165008813142776,
      "epoch": 0.27161433368543175,
      "grad_norm": 8.385638236999512,
      "learning_rate": 4.854212847764657e-06,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8384706929326058,
      "num_tokens": 105349177.0,
      "step": 87620
    },
    {
      "entropy": 1.906524208188057,
      "epoch": 0.2716453328104814,
      "grad_norm": 8.768911361694336,
      "learning_rate": 4.853935864611632e-06,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8458487808704376,
      "num_tokens": 105360116.0,
      "step": 87630
    },
    {
      "entropy": 1.8195328041911125,
      "epoch": 0.27167633193553115,
      "grad_norm": 4.211945056915283,
      "learning_rate": 4.853658928867475e-06,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8428912833333015,
      "num_tokens": 105372704.0,
      "step": 87640
    },
    {
      "entropy": 1.8345939561724662,
      "epoch": 0.2717073310605808,
      "grad_norm": 10.063714981079102,
      "learning_rate": 4.853382040518665e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8555810198187828,
      "num_tokens": 105384756.0,
      "step": 87650
    },
    {
      "entropy": 1.8712797194719315,
      "epoch": 0.2717383301856305,
      "grad_norm": 7.05558967590332,
      "learning_rate": 4.853105199551681e-06,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8413851290941239,
      "num_tokens": 105396711.0,
      "step": 87660
    },
    {
      "entropy": 1.8855430006980896,
      "epoch": 0.2717693293106802,
      "grad_norm": 9.500574111938477,
      "learning_rate": 4.8528284059530145e-06,
      "loss": 0.507,
      "mean_token_accuracy": 0.841337351500988,
      "num_tokens": 105407862.0,
      "step": 87670
    },
    {
      "entropy": 1.7173856884241103,
      "epoch": 0.2718003284357299,
      "grad_norm": 4.654793739318848,
      "learning_rate": 4.852551659709158e-06,
      "loss": 0.3756,
      "mean_token_accuracy": 0.8598586186766625,
      "num_tokens": 105422391.0,
      "step": 87680
    },
    {
      "entropy": 1.9247530341148376,
      "epoch": 0.2718313275607796,
      "grad_norm": 9.287614822387695,
      "learning_rate": 4.85227496080661e-06,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8380204871296882,
      "num_tokens": 105433845.0,
      "step": 87690
    },
    {
      "entropy": 1.9517512962222099,
      "epoch": 0.27186232668582927,
      "grad_norm": 6.864774227142334,
      "learning_rate": 4.851998309231874e-06,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8360002964735032,
      "num_tokens": 105445129.0,
      "step": 87700
    },
    {
      "entropy": 1.9424969971179962,
      "epoch": 0.271893325810879,
      "grad_norm": 8.023564338684082,
      "learning_rate": 4.8517217049714625e-06,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8406699001789093,
      "num_tokens": 105457210.0,
      "step": 87710
    },
    {
      "entropy": 1.8285282671451568,
      "epoch": 0.27192432493592866,
      "grad_norm": 9.891963958740234,
      "learning_rate": 4.851445148011887e-06,
      "loss": 0.454,
      "mean_token_accuracy": 0.853777602314949,
      "num_tokens": 105469532.0,
      "step": 87720
    },
    {
      "entropy": 1.825080545246601,
      "epoch": 0.2719553240609784,
      "grad_norm": 6.351779460906982,
      "learning_rate": 4.8511686383396706e-06,
      "loss": 0.4105,
      "mean_token_accuracy": 0.858852119743824,
      "num_tokens": 105482237.0,
      "step": 87730
    },
    {
      "entropy": 1.9200027763843537,
      "epoch": 0.27198632318602806,
      "grad_norm": 8.065184593200684,
      "learning_rate": 4.850892175941337e-06,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8401974439620972,
      "num_tokens": 105493906.0,
      "step": 87740
    },
    {
      "entropy": 1.8792470768094063,
      "epoch": 0.2720173223110778,
      "grad_norm": 8.423270225524902,
      "learning_rate": 4.8506157608034186e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8373197540640831,
      "num_tokens": 105505414.0,
      "step": 87750
    },
    {
      "entropy": 1.900409395992756,
      "epoch": 0.27204832143612745,
      "grad_norm": 7.148952007293701,
      "learning_rate": 4.850339392912451e-06,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8442846789956093,
      "num_tokens": 105516469.0,
      "step": 87760
    },
    {
      "entropy": 1.8958803132176398,
      "epoch": 0.2720793205611772,
      "grad_norm": 8.72636604309082,
      "learning_rate": 4.850063072254976e-06,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8429311379790306,
      "num_tokens": 105528375.0,
      "step": 87770
    },
    {
      "entropy": 1.8233585372567176,
      "epoch": 0.27211031968622684,
      "grad_norm": 8.03514289855957,
      "learning_rate": 4.849786798817542e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.8540633916854858,
      "num_tokens": 105540482.0,
      "step": 87780
    },
    {
      "entropy": 1.8922078132629394,
      "epoch": 0.27214131881127657,
      "grad_norm": 8.679289817810059,
      "learning_rate": 4.8495105725867e-06,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8497767493128776,
      "num_tokens": 105551735.0,
      "step": 87790
    },
    {
      "entropy": 1.809168304502964,
      "epoch": 0.27217231793632624,
      "grad_norm": 9.962617874145508,
      "learning_rate": 4.84923439354901e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8472581446170807,
      "num_tokens": 105564314.0,
      "step": 87800
    },
    {
      "entropy": 1.843416164815426,
      "epoch": 0.27220331706137596,
      "grad_norm": 7.68436861038208,
      "learning_rate": 4.848958261691033e-06,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8343981161713601,
      "num_tokens": 105576827.0,
      "step": 87810
    },
    {
      "entropy": 1.8259743750095367,
      "epoch": 0.27223431618642563,
      "grad_norm": 7.573136329650879,
      "learning_rate": 4.84868217699934e-06,
      "loss": 0.4568,
      "mean_token_accuracy": 0.852935828268528,
      "num_tokens": 105589417.0,
      "step": 87820
    },
    {
      "entropy": 1.8123556807637216,
      "epoch": 0.27226531531147535,
      "grad_norm": 3.7434639930725098,
      "learning_rate": 4.848406139460503e-06,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8481271311640739,
      "num_tokens": 105602368.0,
      "step": 87830
    },
    {
      "entropy": 1.8485308945178986,
      "epoch": 0.272296314436525,
      "grad_norm": 7.756927490234375,
      "learning_rate": 4.848130149061103e-06,
      "loss": 0.448,
      "mean_token_accuracy": 0.8554665118455886,
      "num_tokens": 105613811.0,
      "step": 87840
    },
    {
      "entropy": 1.8490839540958404,
      "epoch": 0.27232731356157475,
      "grad_norm": 6.909817695617676,
      "learning_rate": 4.847854205787724e-06,
      "loss": 0.4426,
      "mean_token_accuracy": 0.8522930264472961,
      "num_tokens": 105625880.0,
      "step": 87850
    },
    {
      "entropy": 1.882150113582611,
      "epoch": 0.2723583126866244,
      "grad_norm": 9.245566368103027,
      "learning_rate": 4.847578309626954e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8421803861856461,
      "num_tokens": 105638135.0,
      "step": 87860
    },
    {
      "entropy": 1.8645110860466958,
      "epoch": 0.27238931181167414,
      "grad_norm": 8.556130409240723,
      "learning_rate": 4.847302460565392e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8380250200629235,
      "num_tokens": 105649997.0,
      "step": 87870
    },
    {
      "entropy": 1.9121182456612587,
      "epoch": 0.2724203109367238,
      "grad_norm": 6.9276862144470215,
      "learning_rate": 4.847026658589637e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8434757009148598,
      "num_tokens": 105661149.0,
      "step": 87880
    },
    {
      "entropy": 1.8295169189572333,
      "epoch": 0.27245131006177353,
      "grad_norm": 9.222490310668945,
      "learning_rate": 4.846750903686295e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8513430684804917,
      "num_tokens": 105673234.0,
      "step": 87890
    },
    {
      "entropy": 1.9055490285158156,
      "epoch": 0.2724823091868232,
      "grad_norm": 8.43136215209961,
      "learning_rate": 4.846475195841978e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.849004752933979,
      "num_tokens": 105684180.0,
      "step": 87900
    },
    {
      "entropy": 1.7300630405545234,
      "epoch": 0.27251330831187287,
      "grad_norm": 4.451493740081787,
      "learning_rate": 4.846199535043302e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.8522528484463692,
      "num_tokens": 105698528.0,
      "step": 87910
    },
    {
      "entropy": 1.855320343375206,
      "epoch": 0.2725443074369226,
      "grad_norm": 7.741222381591797,
      "learning_rate": 4.845923921276889e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.843564510345459,
      "num_tokens": 105710686.0,
      "step": 87920
    },
    {
      "entropy": 1.8979840464890003,
      "epoch": 0.27257530656197226,
      "grad_norm": Infinity,
      "learning_rate": 4.845648354529367e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8410883545875549,
      "num_tokens": 105722636.0,
      "step": 87930
    },
    {
      "entropy": 1.9001470863819123,
      "epoch": 0.272606305687022,
      "grad_norm": 7.072505474090576,
      "learning_rate": 4.845372834787369e-06,
      "loss": 0.5521,
      "mean_token_accuracy": 0.8388607785105705,
      "num_tokens": 105733839.0,
      "step": 87940
    },
    {
      "entropy": 1.8749860867857933,
      "epoch": 0.27263730481207166,
      "grad_norm": 9.139892578125,
      "learning_rate": 4.845097362037533e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8396959900856018,
      "num_tokens": 105746121.0,
      "step": 87950
    },
    {
      "entropy": 1.8686931714415551,
      "epoch": 0.2726683039371214,
      "grad_norm": 9.034399032592773,
      "learning_rate": 4.844821936266501e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8411045849323273,
      "num_tokens": 105758031.0,
      "step": 87960
    },
    {
      "entropy": 1.9092384189367295,
      "epoch": 0.27269930306217105,
      "grad_norm": 4.113952159881592,
      "learning_rate": 4.844546557460922e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8454341217875481,
      "num_tokens": 105769453.0,
      "step": 87970
    },
    {
      "entropy": 1.8871904879808425,
      "epoch": 0.2727303021872208,
      "grad_norm": 7.443089962005615,
      "learning_rate": 4.844271225607452e-06,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8477641880512238,
      "num_tokens": 105781468.0,
      "step": 87980
    },
    {
      "entropy": 1.887850184738636,
      "epoch": 0.27276130131227044,
      "grad_norm": 7.757199287414551,
      "learning_rate": 4.843995940692748e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8484135344624519,
      "num_tokens": 105793191.0,
      "step": 87990
    },
    {
      "entropy": 1.9092398703098297,
      "epoch": 0.27279230043732017,
      "grad_norm": 8.517298698425293,
      "learning_rate": 4.843720702703475e-06,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8369758918881416,
      "num_tokens": 105804532.0,
      "step": 88000
    },
    {
      "entropy": 1.7902554288506507,
      "epoch": 0.27282329956236984,
      "grad_norm": 4.456912517547607,
      "learning_rate": 4.843445511626304e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8474618881940842,
      "num_tokens": 105817217.0,
      "step": 88010
    },
    {
      "entropy": 1.8608660578727723,
      "epoch": 0.27285429868741956,
      "grad_norm": 8.066084861755371,
      "learning_rate": 4.843170367447909e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8490109965205193,
      "num_tokens": 105828552.0,
      "step": 88020
    },
    {
      "entropy": 1.89578920006752,
      "epoch": 0.27288529781246923,
      "grad_norm": 6.772432327270508,
      "learning_rate": 4.842895270154972e-06,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8335303783416748,
      "num_tokens": 105840037.0,
      "step": 88030
    },
    {
      "entropy": 1.7588407546281815,
      "epoch": 0.27291629693751895,
      "grad_norm": 1.8999171257019043,
      "learning_rate": 4.842620219734178e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8491491839289665,
      "num_tokens": 105854237.0,
      "step": 88040
    },
    {
      "entropy": 1.7782690718770027,
      "epoch": 0.2729472960625686,
      "grad_norm": 8.12587833404541,
      "learning_rate": 4.842345216172217e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8528685718774796,
      "num_tokens": 105866817.0,
      "step": 88050
    },
    {
      "entropy": 1.8298373103141785,
      "epoch": 0.27297829518761835,
      "grad_norm": 7.805505752563477,
      "learning_rate": 4.8420702594557855e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8397832661867142,
      "num_tokens": 105878717.0,
      "step": 88060
    },
    {
      "entropy": 1.8683532044291495,
      "epoch": 0.273009294312668,
      "grad_norm": 3.9877965450286865,
      "learning_rate": 4.841795349571587e-06,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8447675094008446,
      "num_tokens": 105890128.0,
      "step": 88070
    },
    {
      "entropy": 1.8671092882752418,
      "epoch": 0.27304029343771774,
      "grad_norm": 8.055480003356934,
      "learning_rate": 4.841520486506328e-06,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8493238598108291,
      "num_tokens": 105901651.0,
      "step": 88080
    },
    {
      "entropy": 1.9300762385129928,
      "epoch": 0.2730712925627674,
      "grad_norm": 8.968958854675293,
      "learning_rate": 4.841245670246719e-06,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8325337320566177,
      "num_tokens": 105912601.0,
      "step": 88090
    },
    {
      "entropy": 1.7842349156737327,
      "epoch": 0.27310229168781713,
      "grad_norm": 2.736703634262085,
      "learning_rate": 4.840970900779478e-06,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8453866109251976,
      "num_tokens": 105926027.0,
      "step": 88100
    },
    {
      "entropy": 1.8227722927927972,
      "epoch": 0.2731332908128668,
      "grad_norm": 7.752776622772217,
      "learning_rate": 4.840696178091329e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8525062575936317,
      "num_tokens": 105938989.0,
      "step": 88110
    },
    {
      "entropy": 1.85889712870121,
      "epoch": 0.2731642899379165,
      "grad_norm": 7.539937973022461,
      "learning_rate": 4.840421502169e-06,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8374249458312988,
      "num_tokens": 105950886.0,
      "step": 88120
    },
    {
      "entropy": 1.846862156689167,
      "epoch": 0.2731952890629662,
      "grad_norm": 8.776389122009277,
      "learning_rate": 4.840146872999224e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8551399633288383,
      "num_tokens": 105962835.0,
      "step": 88130
    },
    {
      "entropy": 1.7878411993384362,
      "epoch": 0.2732262881880159,
      "grad_norm": 7.770561695098877,
      "learning_rate": 4.839872290568737e-06,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8593245849013329,
      "num_tokens": 105975339.0,
      "step": 88140
    },
    {
      "entropy": 1.8643562525510788,
      "epoch": 0.2732572873130656,
      "grad_norm": 12.23235034942627,
      "learning_rate": 4.839597754864288e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8453767746686935,
      "num_tokens": 105987199.0,
      "step": 88150
    },
    {
      "entropy": 1.887287637591362,
      "epoch": 0.27328828643811526,
      "grad_norm": 7.948936462402344,
      "learning_rate": 4.839323265872622e-06,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8448623090982437,
      "num_tokens": 105998566.0,
      "step": 88160
    },
    {
      "entropy": 1.831418040394783,
      "epoch": 0.273319285563165,
      "grad_norm": 9.216192245483398,
      "learning_rate": 4.839048823580495e-06,
      "loss": 0.482,
      "mean_token_accuracy": 0.8426416292786598,
      "num_tokens": 106011036.0,
      "step": 88170
    },
    {
      "entropy": 1.8624485075473785,
      "epoch": 0.27335028468821465,
      "grad_norm": 8.75737476348877,
      "learning_rate": 4.838774427974665e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.837440450489521,
      "num_tokens": 106022579.0,
      "step": 88180
    },
    {
      "entropy": 1.87282153069973,
      "epoch": 0.2733812838132644,
      "grad_norm": 3.783078670501709,
      "learning_rate": 4.8385000790419005e-06,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8478828489780426,
      "num_tokens": 106033958.0,
      "step": 88190
    },
    {
      "entropy": 1.8366525799036026,
      "epoch": 0.27341228293831404,
      "grad_norm": 4.001052379608154,
      "learning_rate": 4.8382257767689696e-06,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8528647780418396,
      "num_tokens": 106046043.0,
      "step": 88200
    },
    {
      "entropy": 1.9092376694083213,
      "epoch": 0.27344328206336377,
      "grad_norm": 8.00796127319336,
      "learning_rate": 4.837951521142646e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8405633881688118,
      "num_tokens": 106057385.0,
      "step": 88210
    },
    {
      "entropy": 1.8548787474632262,
      "epoch": 0.27347428118841344,
      "grad_norm": 6.67507791519165,
      "learning_rate": 4.837677312149712e-06,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8439633920788765,
      "num_tokens": 106069738.0,
      "step": 88220
    },
    {
      "entropy": 1.7936237141489983,
      "epoch": 0.27350528031346316,
      "grad_norm": 10.650125503540039,
      "learning_rate": 4.837403149776953e-06,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8396651700139046,
      "num_tokens": 106083549.0,
      "step": 88230
    },
    {
      "entropy": 1.8829040303826332,
      "epoch": 0.27353627943851283,
      "grad_norm": 8.387964248657227,
      "learning_rate": 4.837129034011162e-06,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8502602845430374,
      "num_tokens": 106094686.0,
      "step": 88240
    },
    {
      "entropy": 1.8475438334047793,
      "epoch": 0.27356727856356255,
      "grad_norm": 4.180426597595215,
      "learning_rate": 4.836854964839133e-06,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8456549108028412,
      "num_tokens": 106108422.0,
      "step": 88250
    },
    {
      "entropy": 1.9390867114067079,
      "epoch": 0.2735982776886122,
      "grad_norm": 7.525297164916992,
      "learning_rate": 4.836580942247668e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8467183902859687,
      "num_tokens": 106119386.0,
      "step": 88260
    },
    {
      "entropy": 1.8454492062330246,
      "epoch": 0.27362927681366195,
      "grad_norm": 7.91515588760376,
      "learning_rate": 4.836306966223574e-06,
      "loss": 0.5361,
      "mean_token_accuracy": 0.84375009983778,
      "num_tokens": 106131751.0,
      "step": 88270
    },
    {
      "entropy": 1.8392127811908723,
      "epoch": 0.2736602759387116,
      "grad_norm": 4.0122270584106445,
      "learning_rate": 4.8360330367536644e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8450425997376442,
      "num_tokens": 106144785.0,
      "step": 88280
    },
    {
      "entropy": 1.902709110081196,
      "epoch": 0.27369127506376134,
      "grad_norm": 9.007641792297363,
      "learning_rate": 4.835759153824755e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8378954946994781,
      "num_tokens": 106156151.0,
      "step": 88290
    },
    {
      "entropy": 1.8917675152420999,
      "epoch": 0.273722274188811,
      "grad_norm": 8.450139045715332,
      "learning_rate": 4.835485317423669e-06,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8372221887111664,
      "num_tokens": 106167862.0,
      "step": 88300
    },
    {
      "entropy": 1.7971489533782006,
      "epoch": 0.27375327331386073,
      "grad_norm": 8.281867980957031,
      "learning_rate": 4.835211527537234e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8558110296726227,
      "num_tokens": 106181012.0,
      "step": 88310
    },
    {
      "entropy": 1.9332615464925766,
      "epoch": 0.2737842724389104,
      "grad_norm": 7.937662124633789,
      "learning_rate": 4.834937784152283e-06,
      "loss": 0.5597,
      "mean_token_accuracy": 0.8360219776630402,
      "num_tokens": 106192536.0,
      "step": 88320
    },
    {
      "entropy": 1.8722572714090346,
      "epoch": 0.2738152715639601,
      "grad_norm": 9.418374061584473,
      "learning_rate": 4.834664087255653e-06,
      "loss": 0.4561,
      "mean_token_accuracy": 0.8445107057690621,
      "num_tokens": 106204886.0,
      "step": 88330
    },
    {
      "entropy": 1.8897190272808075,
      "epoch": 0.2738462706890098,
      "grad_norm": 7.282384872436523,
      "learning_rate": 4.83439043683419e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8582799851894378,
      "num_tokens": 106216856.0,
      "step": 88340
    },
    {
      "entropy": 1.8999308928847314,
      "epoch": 0.2738772698140595,
      "grad_norm": 9.030301094055176,
      "learning_rate": 4.8341168328747395e-06,
      "loss": 0.496,
      "mean_token_accuracy": 0.8379582807421684,
      "num_tokens": 106228905.0,
      "step": 88350
    },
    {
      "entropy": 1.8722223863005638,
      "epoch": 0.2739082689391092,
      "grad_norm": 7.225384712219238,
      "learning_rate": 4.833843275364157e-06,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8510954797267913,
      "num_tokens": 106240136.0,
      "step": 88360
    },
    {
      "entropy": 1.7975994154810906,
      "epoch": 0.2739392680641589,
      "grad_norm": 7.694894790649414,
      "learning_rate": 4.833569764289303e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8697395831346512,
      "num_tokens": 106253669.0,
      "step": 88370
    },
    {
      "entropy": 1.7122723177075385,
      "epoch": 0.2739702671892086,
      "grad_norm": 4.0345540046691895,
      "learning_rate": 4.833296299637038e-06,
      "loss": 0.3687,
      "mean_token_accuracy": 0.8615933701395988,
      "num_tokens": 106268120.0,
      "step": 88380
    },
    {
      "entropy": 1.7913119062781333,
      "epoch": 0.2740012663142583,
      "grad_norm": 3.695681095123291,
      "learning_rate": 4.833022881394236e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8548486337065697,
      "num_tokens": 106281871.0,
      "step": 88390
    },
    {
      "entropy": 1.8856089636683464,
      "epoch": 0.274032265439308,
      "grad_norm": 7.232779026031494,
      "learning_rate": 4.832749509547768e-06,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8429864302277565,
      "num_tokens": 106293484.0,
      "step": 88400
    },
    {
      "entropy": 1.9002991870045662,
      "epoch": 0.27406326456435764,
      "grad_norm": 4.058182239532471,
      "learning_rate": 4.832476184084515e-06,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8347552672028542,
      "num_tokens": 106305347.0,
      "step": 88410
    },
    {
      "entropy": 1.8636735692620277,
      "epoch": 0.27409426368940737,
      "grad_norm": 7.725419998168945,
      "learning_rate": 4.832202904991362e-06,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8470685452222824,
      "num_tokens": 106317379.0,
      "step": 88420
    },
    {
      "entropy": 1.8224295750260353,
      "epoch": 0.27412526281445704,
      "grad_norm": 7.4432148933410645,
      "learning_rate": 4.8319296722552e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8520729199051857,
      "num_tokens": 106329618.0,
      "step": 88430
    },
    {
      "entropy": 1.9166969522833823,
      "epoch": 0.27415626193950676,
      "grad_norm": 8.651660919189453,
      "learning_rate": 4.8316564858629236e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.843331104516983,
      "num_tokens": 106341322.0,
      "step": 88440
    },
    {
      "entropy": 1.900239697098732,
      "epoch": 0.27418726106455643,
      "grad_norm": 9.139606475830078,
      "learning_rate": 4.831383345801432e-06,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8463894948363304,
      "num_tokens": 106352704.0,
      "step": 88450
    },
    {
      "entropy": 1.8953612327575684,
      "epoch": 0.27421826018960616,
      "grad_norm": 7.59618616104126,
      "learning_rate": 4.831110252057634e-06,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8437128469347954,
      "num_tokens": 106364147.0,
      "step": 88460
    },
    {
      "entropy": 1.8952149584889413,
      "epoch": 0.2742492593146558,
      "grad_norm": 7.970698356628418,
      "learning_rate": 4.830837204618439e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.856081509590149,
      "num_tokens": 106375733.0,
      "step": 88470
    },
    {
      "entropy": 1.8306088283658029,
      "epoch": 0.27428025843970555,
      "grad_norm": 8.874154090881348,
      "learning_rate": 4.830564203470762e-06,
      "loss": 0.4212,
      "mean_token_accuracy": 0.8551815196871757,
      "num_tokens": 106388345.0,
      "step": 88480
    },
    {
      "entropy": 1.8102877527475356,
      "epoch": 0.2743112575647552,
      "grad_norm": 5.486161231994629,
      "learning_rate": 4.830291248601526e-06,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8472759440541268,
      "num_tokens": 106401837.0,
      "step": 88490
    },
    {
      "entropy": 1.9017505258321763,
      "epoch": 0.27434225668980494,
      "grad_norm": 4.231834888458252,
      "learning_rate": 4.830018339997658e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8477098569273949,
      "num_tokens": 106414196.0,
      "step": 88500
    },
    {
      "entropy": 1.763375386595726,
      "epoch": 0.2743732558148546,
      "grad_norm": 3.8429412841796875,
      "learning_rate": 4.829745477646087e-06,
      "loss": 0.3896,
      "mean_token_accuracy": 0.8600920364260674,
      "num_tokens": 106427418.0,
      "step": 88510
    },
    {
      "entropy": 1.8455632477998734,
      "epoch": 0.27440425493990434,
      "grad_norm": 9.410290718078613,
      "learning_rate": 4.829472661533753e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8498677432537078,
      "num_tokens": 106439679.0,
      "step": 88520
    },
    {
      "entropy": 1.9347635477781295,
      "epoch": 0.274435254064954,
      "grad_norm": 8.736660957336426,
      "learning_rate": 4.829199891647595e-06,
      "loss": 0.5427,
      "mean_token_accuracy": 0.834366361796856,
      "num_tokens": 106450622.0,
      "step": 88530
    },
    {
      "entropy": 1.7944782301783562,
      "epoch": 0.27446625319000373,
      "grad_norm": 3.962170362472534,
      "learning_rate": 4.828927167974562e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.84948940128088,
      "num_tokens": 106464014.0,
      "step": 88540
    },
    {
      "entropy": 1.911988915503025,
      "epoch": 0.2744972523150534,
      "grad_norm": 8.544610977172852,
      "learning_rate": 4.828654490501605e-06,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8315621390938759,
      "num_tokens": 106475443.0,
      "step": 88550
    },
    {
      "entropy": 1.8577186211943626,
      "epoch": 0.2745282514401031,
      "grad_norm": 9.083486557006836,
      "learning_rate": 4.828381859215683e-06,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8416240692138672,
      "num_tokens": 106487802.0,
      "step": 88560
    },
    {
      "entropy": 1.8315035477280617,
      "epoch": 0.2745592505651528,
      "grad_norm": 8.581526756286621,
      "learning_rate": 4.828109274103759e-06,
      "loss": 0.4935,
      "mean_token_accuracy": 0.8431025877594948,
      "num_tokens": 106501150.0,
      "step": 88570
    },
    {
      "entropy": 1.8865200936794282,
      "epoch": 0.2745902496902025,
      "grad_norm": 11.377765655517578,
      "learning_rate": 4.8278367351527985e-06,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8460037097334862,
      "num_tokens": 106513566.0,
      "step": 88580
    },
    {
      "entropy": 1.7892251804471015,
      "epoch": 0.2746212488152522,
      "grad_norm": 6.965874671936035,
      "learning_rate": 4.8275642423497745e-06,
      "loss": 0.3847,
      "mean_token_accuracy": 0.8588668003678321,
      "num_tokens": 106526831.0,
      "step": 88590
    },
    {
      "entropy": 1.8974862158298493,
      "epoch": 0.2746522479403019,
      "grad_norm": 7.9539265632629395,
      "learning_rate": 4.827291795681668e-06,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8509656980633735,
      "num_tokens": 106538001.0,
      "step": 88600
    },
    {
      "entropy": 1.8913881599903106,
      "epoch": 0.2746832470653516,
      "grad_norm": 9.2141752243042,
      "learning_rate": 4.827019395135459e-06,
      "loss": 0.5634,
      "mean_token_accuracy": 0.8275657877326011,
      "num_tokens": 106549809.0,
      "step": 88610
    },
    {
      "entropy": 1.7867820590734482,
      "epoch": 0.2747142461904013,
      "grad_norm": 7.718073844909668,
      "learning_rate": 4.8267470406981375e-06,
      "loss": 0.4228,
      "mean_token_accuracy": 0.8527751803398133,
      "num_tokens": 106563083.0,
      "step": 88620
    },
    {
      "entropy": 1.838895745575428,
      "epoch": 0.27474524531545097,
      "grad_norm": 8.962431907653809,
      "learning_rate": 4.826474732356697e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8443746566772461,
      "num_tokens": 106575815.0,
      "step": 88630
    },
    {
      "entropy": 1.9002109482884406,
      "epoch": 0.2747762444405007,
      "grad_norm": 7.583240509033203,
      "learning_rate": 4.826202470098135e-06,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8451993703842163,
      "num_tokens": 106587141.0,
      "step": 88640
    },
    {
      "entropy": 1.7970835909247398,
      "epoch": 0.27480724356555036,
      "grad_norm": 3.407457113265991,
      "learning_rate": 4.825930253909458e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8588675916194916,
      "num_tokens": 106600173.0,
      "step": 88650
    },
    {
      "entropy": 1.8548505648970603,
      "epoch": 0.27483824269060003,
      "grad_norm": 7.602267742156982,
      "learning_rate": 4.825658083777671e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8400174841284752,
      "num_tokens": 106611441.0,
      "step": 88660
    },
    {
      "entropy": 1.8951510965824128,
      "epoch": 0.27486924181564976,
      "grad_norm": 9.224246978759766,
      "learning_rate": 4.82538595968979e-06,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8430482760071755,
      "num_tokens": 106622973.0,
      "step": 88670
    },
    {
      "entropy": 1.8475442111492157,
      "epoch": 0.2749002409406994,
      "grad_norm": 7.481077671051025,
      "learning_rate": 4.825113881632835e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8386109799146653,
      "num_tokens": 106635332.0,
      "step": 88680
    },
    {
      "entropy": 1.8375971369445323,
      "epoch": 0.27493124006574915,
      "grad_norm": 6.685962200164795,
      "learning_rate": 4.824841849593828e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.85401571393013,
      "num_tokens": 106647486.0,
      "step": 88690
    },
    {
      "entropy": 1.8508135929703713,
      "epoch": 0.2749622391907988,
      "grad_norm": 9.494312286376953,
      "learning_rate": 4.824569863559801e-06,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8352826073765754,
      "num_tokens": 106659227.0,
      "step": 88700
    },
    {
      "entropy": 1.8618362814188003,
      "epoch": 0.27499323831584854,
      "grad_norm": 8.583227157592773,
      "learning_rate": 4.824297923517787e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8454991012811661,
      "num_tokens": 106671337.0,
      "step": 88710
    },
    {
      "entropy": 1.8632221952080728,
      "epoch": 0.2750242374408982,
      "grad_norm": 4.273068428039551,
      "learning_rate": 4.824026029454825e-06,
      "loss": 0.477,
      "mean_token_accuracy": 0.8438972979784012,
      "num_tokens": 106683266.0,
      "step": 88720
    },
    {
      "entropy": 1.8759027153253556,
      "epoch": 0.27505523656594794,
      "grad_norm": 6.9859538078308105,
      "learning_rate": 4.823754181357961e-06,
      "loss": 0.5663,
      "mean_token_accuracy": 0.8387471958994865,
      "num_tokens": 106695075.0,
      "step": 88730
    },
    {
      "entropy": 1.9251735389232636,
      "epoch": 0.2750862356909976,
      "grad_norm": 8.672323226928711,
      "learning_rate": 4.823482379214244e-06,
      "loss": 0.5499,
      "mean_token_accuracy": 0.8304734826087952,
      "num_tokens": 106706762.0,
      "step": 88740
    },
    {
      "entropy": 1.8503682538866997,
      "epoch": 0.27511723481604733,
      "grad_norm": 9.310576438903809,
      "learning_rate": 4.8232106230107285e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8472747296094895,
      "num_tokens": 106718287.0,
      "step": 88750
    },
    {
      "entropy": 1.9084057167172432,
      "epoch": 0.275148233941097,
      "grad_norm": 8.345101356506348,
      "learning_rate": 4.822938912734476e-06,
      "loss": 0.5446,
      "mean_token_accuracy": 0.8372975274920463,
      "num_tokens": 106729907.0,
      "step": 88760
    },
    {
      "entropy": 1.8601343676447868,
      "epoch": 0.2751792330661467,
      "grad_norm": 3.5940239429473877,
      "learning_rate": 4.822667248372551e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8547569274902344,
      "num_tokens": 106742034.0,
      "step": 88770
    },
    {
      "entropy": 1.9169348627328873,
      "epoch": 0.2752102321911964,
      "grad_norm": 8.337597846984863,
      "learning_rate": 4.822395629912025e-06,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8388430058956147,
      "num_tokens": 106753068.0,
      "step": 88780
    },
    {
      "entropy": 1.872619953751564,
      "epoch": 0.2752412313162461,
      "grad_norm": 10.032551765441895,
      "learning_rate": 4.8221240573399705e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8413626179099083,
      "num_tokens": 106764665.0,
      "step": 88790
    },
    {
      "entropy": 1.8718515574932098,
      "epoch": 0.2752722304412958,
      "grad_norm": 4.433212757110596,
      "learning_rate": 4.82185253064347e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8516853898763657,
      "num_tokens": 106776221.0,
      "step": 88800
    },
    {
      "entropy": 1.8270509555935859,
      "epoch": 0.2753032295663455,
      "grad_norm": 8.920626640319824,
      "learning_rate": 4.821581049809608e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8488035008311272,
      "num_tokens": 106789268.0,
      "step": 88810
    },
    {
      "entropy": 1.8341209158301353,
      "epoch": 0.2753342286913952,
      "grad_norm": 9.364352226257324,
      "learning_rate": 4.821309614825477e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8486865177750588,
      "num_tokens": 106800990.0,
      "step": 88820
    },
    {
      "entropy": 1.7747258245944977,
      "epoch": 0.2753652278164449,
      "grad_norm": 3.747847557067871,
      "learning_rate": 4.82103822567817e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8474699661135674,
      "num_tokens": 106814661.0,
      "step": 88830
    },
    {
      "entropy": 1.8344976738095284,
      "epoch": 0.27539622694149457,
      "grad_norm": 3.948411226272583,
      "learning_rate": 4.8207668823547895e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8421635374426841,
      "num_tokens": 106827568.0,
      "step": 88840
    },
    {
      "entropy": 1.873189702630043,
      "epoch": 0.2754272260665443,
      "grad_norm": 7.882256507873535,
      "learning_rate": 4.82049558484244e-06,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8454801499843597,
      "num_tokens": 106839436.0,
      "step": 88850
    },
    {
      "entropy": 1.8929407209157945,
      "epoch": 0.27545822519159396,
      "grad_norm": 6.915173530578613,
      "learning_rate": 4.820224333128236e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8434187933802605,
      "num_tokens": 106850474.0,
      "step": 88860
    },
    {
      "entropy": 1.8293378427624702,
      "epoch": 0.2754892243166437,
      "grad_norm": 4.267940521240234,
      "learning_rate": 4.819953127199289e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8467979103326797,
      "num_tokens": 106862337.0,
      "step": 88870
    },
    {
      "entropy": 1.8841822102665902,
      "epoch": 0.27552022344169336,
      "grad_norm": 7.245059013366699,
      "learning_rate": 4.819681967042724e-06,
      "loss": 0.513,
      "mean_token_accuracy": 0.8383611172437668,
      "num_tokens": 106874237.0,
      "step": 88880
    },
    {
      "entropy": 1.8861842527985573,
      "epoch": 0.275551222566743,
      "grad_norm": 7.553530693054199,
      "learning_rate": 4.819410852645663e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8369044825434685,
      "num_tokens": 106886243.0,
      "step": 88890
    },
    {
      "entropy": 1.8612857922911643,
      "epoch": 0.27558222169179275,
      "grad_norm": 5.01759147644043,
      "learning_rate": 4.81913978399524e-06,
      "loss": 0.523,
      "mean_token_accuracy": 0.8347546473145485,
      "num_tokens": 106898922.0,
      "step": 88900
    },
    {
      "entropy": 1.789707398414612,
      "epoch": 0.2756132208168424,
      "grad_norm": 8.906004905700684,
      "learning_rate": 4.818868761078591e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8516422912478447,
      "num_tokens": 106911823.0,
      "step": 88910
    },
    {
      "entropy": 1.8937145173549652,
      "epoch": 0.27564421994189214,
      "grad_norm": 7.709479331970215,
      "learning_rate": 4.818597783882858e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8450556769967079,
      "num_tokens": 106923627.0,
      "step": 88920
    },
    {
      "entropy": 1.8715748369693757,
      "epoch": 0.2756752190669418,
      "grad_norm": 8.210858345031738,
      "learning_rate": 4.818326852395186e-06,
      "loss": 0.5139,
      "mean_token_accuracy": 0.843447645008564,
      "num_tokens": 106935011.0,
      "step": 88930
    },
    {
      "entropy": 1.8829000025987626,
      "epoch": 0.27570621819199154,
      "grad_norm": 9.785741806030273,
      "learning_rate": 4.818055966602728e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8352604553103447,
      "num_tokens": 106947124.0,
      "step": 88940
    },
    {
      "entropy": 1.9164473339915276,
      "epoch": 0.2757372173170412,
      "grad_norm": 8.271210670471191,
      "learning_rate": 4.817785126492638e-06,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8394878327846527,
      "num_tokens": 106957935.0,
      "step": 88950
    },
    {
      "entropy": 1.890079266577959,
      "epoch": 0.27576821644209093,
      "grad_norm": 8.78956127166748,
      "learning_rate": 4.817514332052081e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8357309713959694,
      "num_tokens": 106970004.0,
      "step": 88960
    },
    {
      "entropy": 1.9482278615236281,
      "epoch": 0.2757992155671406,
      "grad_norm": 8.85912799835205,
      "learning_rate": 4.817243583268221e-06,
      "loss": 0.5326,
      "mean_token_accuracy": 0.8404009014368057,
      "num_tokens": 106980930.0,
      "step": 88970
    },
    {
      "entropy": 1.8766920641064644,
      "epoch": 0.2758302146921903,
      "grad_norm": 7.513118743896484,
      "learning_rate": 4.8169728801282294e-06,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8502911329269409,
      "num_tokens": 106992422.0,
      "step": 88980
    },
    {
      "entropy": 1.8944596245884895,
      "epoch": 0.27586121381724,
      "grad_norm": 7.3327460289001465,
      "learning_rate": 4.816702222619286e-06,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8509733945131301,
      "num_tokens": 107004273.0,
      "step": 88990
    },
    {
      "entropy": 1.847128589451313,
      "epoch": 0.2758922129422897,
      "grad_norm": 4.272797584533691,
      "learning_rate": 4.816431610728571e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.8441042378544807,
      "num_tokens": 107016497.0,
      "step": 89000
    },
    {
      "entropy": 1.855496746301651,
      "epoch": 0.2759232120673394,
      "grad_norm": 4.222333908081055,
      "learning_rate": 4.816161044443269e-06,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8440208032727241,
      "num_tokens": 107029107.0,
      "step": 89010
    },
    {
      "entropy": 1.9156458109617234,
      "epoch": 0.2759542111923891,
      "grad_norm": 8.503098487854004,
      "learning_rate": 4.815890523750575e-06,
      "loss": 0.525,
      "mean_token_accuracy": 0.8486156970262527,
      "num_tokens": 107039998.0,
      "step": 89020
    },
    {
      "entropy": 1.8888219490647316,
      "epoch": 0.2759852103174388,
      "grad_norm": 2.975604772567749,
      "learning_rate": 4.8156200486376845e-06,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8388027891516685,
      "num_tokens": 107052621.0,
      "step": 89030
    },
    {
      "entropy": 1.9181819319725038,
      "epoch": 0.2760162094424885,
      "grad_norm": 3.860818386077881,
      "learning_rate": 4.8153496190918e-06,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8331426337361336,
      "num_tokens": 107063730.0,
      "step": 89040
    },
    {
      "entropy": 1.832422287762165,
      "epoch": 0.27604720856753817,
      "grad_norm": 7.345708847045898,
      "learning_rate": 4.815079235100127e-06,
      "loss": 0.4354,
      "mean_token_accuracy": 0.854414065182209,
      "num_tokens": 107076862.0,
      "step": 89050
    },
    {
      "entropy": 1.9153432667255401,
      "epoch": 0.2760782076925879,
      "grad_norm": 9.088801383972168,
      "learning_rate": 4.814808896649879e-06,
      "loss": 0.4634,
      "mean_token_accuracy": 0.853677150607109,
      "num_tokens": 107088395.0,
      "step": 89060
    },
    {
      "entropy": 1.8690629690885543,
      "epoch": 0.27610920681763756,
      "grad_norm": 7.586027145385742,
      "learning_rate": 4.814538603728274e-06,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8534372985363007,
      "num_tokens": 107100834.0,
      "step": 89070
    },
    {
      "entropy": 1.8856920048594474,
      "epoch": 0.2761402059426873,
      "grad_norm": 4.5250043869018555,
      "learning_rate": 4.814268356322531e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8607599556446075,
      "num_tokens": 107112389.0,
      "step": 89080
    },
    {
      "entropy": 1.9317937284708022,
      "epoch": 0.27617120506773696,
      "grad_norm": 8.948545455932617,
      "learning_rate": 4.813998154419879e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8472296670079231,
      "num_tokens": 107122894.0,
      "step": 89090
    },
    {
      "entropy": 1.8248413413763047,
      "epoch": 0.2762022041927867,
      "grad_norm": 11.74347972869873,
      "learning_rate": 4.813727998007552e-06,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8507599249482155,
      "num_tokens": 107134891.0,
      "step": 89100
    },
    {
      "entropy": 1.8651644140481949,
      "epoch": 0.27623320331783635,
      "grad_norm": 8.039405822753906,
      "learning_rate": 4.813457887072781e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.8547470927238464,
      "num_tokens": 107146700.0,
      "step": 89110
    },
    {
      "entropy": 1.8344761043787003,
      "epoch": 0.2762642024428861,
      "grad_norm": 8.795267105102539,
      "learning_rate": 4.813187821602815e-06,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8472502484917641,
      "num_tokens": 107159722.0,
      "step": 89120
    },
    {
      "entropy": 1.909415753185749,
      "epoch": 0.27629520156793574,
      "grad_norm": 9.062956809997559,
      "learning_rate": 4.812917801584898e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.850268816947937,
      "num_tokens": 107171135.0,
      "step": 89130
    },
    {
      "entropy": 1.886960855126381,
      "epoch": 0.2763262006929854,
      "grad_norm": 8.55114459991455,
      "learning_rate": 4.812647827006282e-06,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8514637067914009,
      "num_tokens": 107182863.0,
      "step": 89140
    },
    {
      "entropy": 1.862102809548378,
      "epoch": 0.27635719981803514,
      "grad_norm": 9.735430717468262,
      "learning_rate": 4.812377897854223e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8368513882160187,
      "num_tokens": 107195661.0,
      "step": 89150
    },
    {
      "entropy": 1.832503044605255,
      "epoch": 0.2763881989430848,
      "grad_norm": 7.541407108306885,
      "learning_rate": 4.812108014115985e-06,
      "loss": 0.4209,
      "mean_token_accuracy": 0.8562299177050591,
      "num_tokens": 107208363.0,
      "step": 89160
    },
    {
      "entropy": 1.884467676281929,
      "epoch": 0.27641919806813453,
      "grad_norm": 9.188913345336914,
      "learning_rate": 4.811838175778836e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.8432751014828682,
      "num_tokens": 107219939.0,
      "step": 89170
    },
    {
      "entropy": 1.8959971502423287,
      "epoch": 0.2764501971931842,
      "grad_norm": 4.789566993713379,
      "learning_rate": 4.8115683828300445e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8498433887958526,
      "num_tokens": 107231619.0,
      "step": 89180
    },
    {
      "entropy": 1.8323383823037147,
      "epoch": 0.2764811963182339,
      "grad_norm": 2.805758237838745,
      "learning_rate": 4.811298635256891e-06,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8528842240571975,
      "num_tokens": 107244259.0,
      "step": 89190
    },
    {
      "entropy": 1.7991958245635034,
      "epoch": 0.2765121954432836,
      "grad_norm": 6.88820219039917,
      "learning_rate": 4.811028933046656e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.872551740705967,
      "num_tokens": 107257190.0,
      "step": 89200
    },
    {
      "entropy": 1.8357098802924157,
      "epoch": 0.2765431945683333,
      "grad_norm": 4.887908935546875,
      "learning_rate": 4.810759276186628e-06,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8493363171815872,
      "num_tokens": 107269231.0,
      "step": 89210
    },
    {
      "entropy": 1.8650715343654156,
      "epoch": 0.276574193693383,
      "grad_norm": 8.015203475952148,
      "learning_rate": 4.810489664664098e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8378633111715317,
      "num_tokens": 107282063.0,
      "step": 89220
    },
    {
      "entropy": 1.8511658303439618,
      "epoch": 0.2766051928184327,
      "grad_norm": 8.239190101623535,
      "learning_rate": 4.810220098466364e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8476231142878532,
      "num_tokens": 107294849.0,
      "step": 89230
    },
    {
      "entropy": 1.8353752836585044,
      "epoch": 0.2766361919434824,
      "grad_norm": 11.316750526428223,
      "learning_rate": 4.809950577580724e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8560484200716019,
      "num_tokens": 107306875.0,
      "step": 89240
    },
    {
      "entropy": 1.8877992361783982,
      "epoch": 0.2766671910685321,
      "grad_norm": 10.842957496643066,
      "learning_rate": 4.809681101994492e-06,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8417460203170777,
      "num_tokens": 107318449.0,
      "step": 89250
    },
    {
      "entropy": 1.798448894917965,
      "epoch": 0.2766981901935818,
      "grad_norm": 7.134692192077637,
      "learning_rate": 4.809411671694974e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8509971871972084,
      "num_tokens": 107331888.0,
      "step": 89260
    },
    {
      "entropy": 1.8195422321558,
      "epoch": 0.2767291893186315,
      "grad_norm": 10.8891019821167,
      "learning_rate": 4.809142286669492e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8366403177380561,
      "num_tokens": 107344418.0,
      "step": 89270
    },
    {
      "entropy": 1.857719998061657,
      "epoch": 0.27676018844368117,
      "grad_norm": 7.889041423797607,
      "learning_rate": 4.808872946905363e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8492889553308487,
      "num_tokens": 107356113.0,
      "step": 89280
    },
    {
      "entropy": 1.7993003293871879,
      "epoch": 0.2767911875687309,
      "grad_norm": 7.292881011962891,
      "learning_rate": 4.808603652389917e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8571058794856071,
      "num_tokens": 107369163.0,
      "step": 89290
    },
    {
      "entropy": 1.8662144735455513,
      "epoch": 0.27682218669378056,
      "grad_norm": 10.856431007385254,
      "learning_rate": 4.808334403110485e-06,
      "loss": 0.5662,
      "mean_token_accuracy": 0.8313690677285195,
      "num_tokens": 107381585.0,
      "step": 89300
    },
    {
      "entropy": 1.8663518592715262,
      "epoch": 0.2768531858188303,
      "grad_norm": 7.93809700012207,
      "learning_rate": 4.808065199054404e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8461432874202728,
      "num_tokens": 107393815.0,
      "step": 89310
    },
    {
      "entropy": 1.9202445238828658,
      "epoch": 0.27688418494387995,
      "grad_norm": 8.869173049926758,
      "learning_rate": 4.8077960402090155e-06,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8392524033784866,
      "num_tokens": 107404393.0,
      "step": 89320
    },
    {
      "entropy": 1.9343400806188584,
      "epoch": 0.2769151840689297,
      "grad_norm": 8.072264671325684,
      "learning_rate": 4.807526926561667e-06,
      "loss": 0.509,
      "mean_token_accuracy": 0.8465714007616043,
      "num_tokens": 107414917.0,
      "step": 89330
    },
    {
      "entropy": 1.9184037640690803,
      "epoch": 0.27694618319397935,
      "grad_norm": 7.905839443206787,
      "learning_rate": 4.80725785809971e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8361783891916275,
      "num_tokens": 107425889.0,
      "step": 89340
    },
    {
      "entropy": 1.9122958168387414,
      "epoch": 0.27697718231902907,
      "grad_norm": 3.9599266052246094,
      "learning_rate": 4.806988834810501e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8507964372634887,
      "num_tokens": 107437118.0,
      "step": 89350
    },
    {
      "entropy": 1.7936722189188004,
      "epoch": 0.27700818144407874,
      "grad_norm": 9.577056884765625,
      "learning_rate": 4.806719856681402e-06,
      "loss": 0.4249,
      "mean_token_accuracy": 0.8579995214939118,
      "num_tokens": 107450413.0,
      "step": 89360
    },
    {
      "entropy": 1.8381581105291844,
      "epoch": 0.27703918056912846,
      "grad_norm": 7.809806823730469,
      "learning_rate": 4.806450923699778e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8484640643000603,
      "num_tokens": 107462904.0,
      "step": 89370
    },
    {
      "entropy": 1.878132027387619,
      "epoch": 0.27707017969417813,
      "grad_norm": 8.047992706298828,
      "learning_rate": 4.806182035853004e-06,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8498531639575958,
      "num_tokens": 107474624.0,
      "step": 89380
    },
    {
      "entropy": 1.9003929272294044,
      "epoch": 0.2771011788192278,
      "grad_norm": 7.774439334869385,
      "learning_rate": 4.805913193128452e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.864824341237545,
      "num_tokens": 107486478.0,
      "step": 89390
    },
    {
      "entropy": 1.885640236735344,
      "epoch": 0.2771321779442775,
      "grad_norm": 8.628144264221191,
      "learning_rate": 4.805644395513508e-06,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8425728008151054,
      "num_tokens": 107498133.0,
      "step": 89400
    },
    {
      "entropy": 1.9598473072052003,
      "epoch": 0.2771631770693272,
      "grad_norm": 9.253849983215332,
      "learning_rate": 4.805375642995554e-06,
      "loss": 0.5666,
      "mean_token_accuracy": 0.8246390670537949,
      "num_tokens": 107509154.0,
      "step": 89410
    },
    {
      "entropy": 1.8804704681038857,
      "epoch": 0.2771941761943769,
      "grad_norm": 3.7883799076080322,
      "learning_rate": 4.8051069355619846e-06,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8492748379707337,
      "num_tokens": 107521568.0,
      "step": 89420
    },
    {
      "entropy": 1.9423058658838273,
      "epoch": 0.2772251753194266,
      "grad_norm": 7.831295490264893,
      "learning_rate": 4.804838273200196e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.8353771314024925,
      "num_tokens": 107532573.0,
      "step": 89430
    },
    {
      "entropy": 1.9429254934191704,
      "epoch": 0.2772561744444763,
      "grad_norm": 7.812403202056885,
      "learning_rate": 4.804569655897587e-06,
      "loss": 0.514,
      "mean_token_accuracy": 0.8479711428284645,
      "num_tokens": 107544075.0,
      "step": 89440
    },
    {
      "entropy": 1.8462449744343759,
      "epoch": 0.277287173569526,
      "grad_norm": 7.212341785430908,
      "learning_rate": 4.8043010836415645e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8527239561080933,
      "num_tokens": 107556555.0,
      "step": 89450
    },
    {
      "entropy": 1.8698409616947174,
      "epoch": 0.2773181726945757,
      "grad_norm": 7.4296698570251465,
      "learning_rate": 4.804032556419541e-06,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8424155786633492,
      "num_tokens": 107569186.0,
      "step": 89460
    },
    {
      "entropy": 1.8821595564484597,
      "epoch": 0.2773491718196254,
      "grad_norm": 7.037583351135254,
      "learning_rate": 4.803764074218931e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8456917703151703,
      "num_tokens": 107580496.0,
      "step": 89470
    },
    {
      "entropy": 1.8518904522061348,
      "epoch": 0.2773801709446751,
      "grad_norm": 4.142475128173828,
      "learning_rate": 4.803495637027156e-06,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8462300166487694,
      "num_tokens": 107592528.0,
      "step": 89480
    },
    {
      "entropy": 1.9444121688604354,
      "epoch": 0.27741117006972477,
      "grad_norm": 8.443082809448242,
      "learning_rate": 4.803227244831642e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8497670993208886,
      "num_tokens": 107603347.0,
      "step": 89490
    },
    {
      "entropy": 1.9082912877202034,
      "epoch": 0.2774421691947745,
      "grad_norm": 8.889103889465332,
      "learning_rate": 4.80295889761982e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8479152485728264,
      "num_tokens": 107614464.0,
      "step": 89500
    },
    {
      "entropy": 1.9622692078351975,
      "epoch": 0.27747316831982416,
      "grad_norm": 9.245258331298828,
      "learning_rate": 4.802690595379124e-06,
      "loss": 0.5951,
      "mean_token_accuracy": 0.8277513101696968,
      "num_tokens": 107625283.0,
      "step": 89510
    },
    {
      "entropy": 1.8577041417360305,
      "epoch": 0.2775041674448739,
      "grad_norm": 8.274160385131836,
      "learning_rate": 4.802422338096995e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8520587861537934,
      "num_tokens": 107637211.0,
      "step": 89520
    },
    {
      "entropy": 1.8055953189730645,
      "epoch": 0.27753516656992355,
      "grad_norm": 9.126919746398926,
      "learning_rate": 4.80215412576088e-06,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8396184176206589,
      "num_tokens": 107650513.0,
      "step": 89530
    },
    {
      "entropy": 1.9063657209277154,
      "epoch": 0.2775661656949733,
      "grad_norm": 7.300408840179443,
      "learning_rate": 4.801885958358229e-06,
      "loss": 0.4567,
      "mean_token_accuracy": 0.853947177529335,
      "num_tokens": 107662885.0,
      "step": 89540
    },
    {
      "entropy": 2.0144054174423216,
      "epoch": 0.27759716482002295,
      "grad_norm": 8.333145141601562,
      "learning_rate": 4.801617835876496e-06,
      "loss": 0.5785,
      "mean_token_accuracy": 0.8276149451732635,
      "num_tokens": 107673365.0,
      "step": 89550
    },
    {
      "entropy": 1.8120394110679627,
      "epoch": 0.27762816394507267,
      "grad_norm": 5.053866863250732,
      "learning_rate": 4.801349758303142e-06,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8445941239595414,
      "num_tokens": 107686729.0,
      "step": 89560
    },
    {
      "entropy": 1.8656949549913406,
      "epoch": 0.27765916307012234,
      "grad_norm": 10.746817588806152,
      "learning_rate": 4.801081725625631e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8389295890927315,
      "num_tokens": 107699283.0,
      "step": 89570
    },
    {
      "entropy": 1.8939693599939347,
      "epoch": 0.27769016219517206,
      "grad_norm": 7.9915971755981445,
      "learning_rate": 4.800813737831435e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.836301201581955,
      "num_tokens": 107710971.0,
      "step": 89580
    },
    {
      "entropy": 1.8993390202522278,
      "epoch": 0.27772116132022173,
      "grad_norm": 3.7865583896636963,
      "learning_rate": 4.800545794908028e-06,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8455944269895553,
      "num_tokens": 107721957.0,
      "step": 89590
    },
    {
      "entropy": 1.9103139862418175,
      "epoch": 0.27775216044527146,
      "grad_norm": 7.072427749633789,
      "learning_rate": 4.800277896842888e-06,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8366509795188903,
      "num_tokens": 107733436.0,
      "step": 89600
    },
    {
      "entropy": 1.916232281923294,
      "epoch": 0.2777831595703211,
      "grad_norm": 7.880809307098389,
      "learning_rate": 4.8000100436235025e-06,
      "loss": 0.4882,
      "mean_token_accuracy": 0.846810282766819,
      "num_tokens": 107744724.0,
      "step": 89610
    },
    {
      "entropy": 1.8565625533461572,
      "epoch": 0.27781415869537085,
      "grad_norm": 9.63576889038086,
      "learning_rate": 4.799742235237359e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8563435986638069,
      "num_tokens": 107756737.0,
      "step": 89620
    },
    {
      "entropy": 1.8952525824308395,
      "epoch": 0.2778451578204205,
      "grad_norm": 7.304356575012207,
      "learning_rate": 4.799474471671954e-06,
      "loss": 0.4555,
      "mean_token_accuracy": 0.855880931019783,
      "num_tokens": 107767777.0,
      "step": 89630
    },
    {
      "entropy": 1.8737752437591553,
      "epoch": 0.2778761569454702,
      "grad_norm": 8.607854843139648,
      "learning_rate": 4.799206752914784e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8448594495654106,
      "num_tokens": 107780521.0,
      "step": 89640
    },
    {
      "entropy": 1.8830969855189323,
      "epoch": 0.2779071560705199,
      "grad_norm": 9.414170265197754,
      "learning_rate": 4.798939078953355e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8453325614333153,
      "num_tokens": 107792404.0,
      "step": 89650
    },
    {
      "entropy": 1.8424068495631218,
      "epoch": 0.2779381551955696,
      "grad_norm": 3.329559087753296,
      "learning_rate": 4.798671449775176e-06,
      "loss": 0.4264,
      "mean_token_accuracy": 0.851707661151886,
      "num_tokens": 107804701.0,
      "step": 89660
    },
    {
      "entropy": 1.8513332203030586,
      "epoch": 0.2779691543206193,
      "grad_norm": 9.445657730102539,
      "learning_rate": 4.798403865367761e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8407017186284065,
      "num_tokens": 107817031.0,
      "step": 89670
    },
    {
      "entropy": 1.861585983633995,
      "epoch": 0.278000153445669,
      "grad_norm": 8.74657154083252,
      "learning_rate": 4.798136325718627e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8404981315135955,
      "num_tokens": 107829450.0,
      "step": 89680
    },
    {
      "entropy": 1.8812636777758598,
      "epoch": 0.2780311525707187,
      "grad_norm": 8.068127632141113,
      "learning_rate": 4.797868830815301e-06,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8442062169313431,
      "num_tokens": 107842211.0,
      "step": 89690
    },
    {
      "entropy": 1.8865378215909003,
      "epoch": 0.27806215169576837,
      "grad_norm": 8.662693977355957,
      "learning_rate": 4.797601380645308e-06,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8458723932504654,
      "num_tokens": 107854429.0,
      "step": 89700
    },
    {
      "entropy": 1.9377282798290252,
      "epoch": 0.2780931508208181,
      "grad_norm": 7.293258190155029,
      "learning_rate": 4.797333975196185e-06,
      "loss": 0.5247,
      "mean_token_accuracy": 0.845015498995781,
      "num_tokens": 107866326.0,
      "step": 89710
    },
    {
      "entropy": 1.8403261929750443,
      "epoch": 0.27812414994586776,
      "grad_norm": 4.629557132720947,
      "learning_rate": 4.797066614455466e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8479184404015541,
      "num_tokens": 107879453.0,
      "step": 89720
    },
    {
      "entropy": 1.8934699580073358,
      "epoch": 0.2781551490709175,
      "grad_norm": 3.9676856994628906,
      "learning_rate": 4.796799298410698e-06,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8484136417508126,
      "num_tokens": 107891153.0,
      "step": 89730
    },
    {
      "entropy": 1.954687887430191,
      "epoch": 0.27818614819596715,
      "grad_norm": 7.991013526916504,
      "learning_rate": 4.796532027049428e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.85145123898983,
      "num_tokens": 107902323.0,
      "step": 89740
    },
    {
      "entropy": 1.9269502833485603,
      "epoch": 0.2782171473210169,
      "grad_norm": 10.174466133117676,
      "learning_rate": 4.796264800359207e-06,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8483559593558312,
      "num_tokens": 107913752.0,
      "step": 89750
    },
    {
      "entropy": 1.9451028525829315,
      "epoch": 0.27824814644606655,
      "grad_norm": 8.776559829711914,
      "learning_rate": 4.795997618327595e-06,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8392796277999878,
      "num_tokens": 107925468.0,
      "step": 89760
    },
    {
      "entropy": 1.9637413799762726,
      "epoch": 0.27827914557111627,
      "grad_norm": 8.385177612304688,
      "learning_rate": 4.795730480942153e-06,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8446138560771942,
      "num_tokens": 107936686.0,
      "step": 89770
    },
    {
      "entropy": 1.9453602582216263,
      "epoch": 0.27831014469616594,
      "grad_norm": 9.379023551940918,
      "learning_rate": 4.795463388190449e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8378261864185333,
      "num_tokens": 107947735.0,
      "step": 89780
    },
    {
      "entropy": 1.8662060409784318,
      "epoch": 0.27834114382121566,
      "grad_norm": 7.810781478881836,
      "learning_rate": 4.7951963400600565e-06,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8564840778708458,
      "num_tokens": 107958863.0,
      "step": 89790
    },
    {
      "entropy": 1.8019395358860493,
      "epoch": 0.27837214294626533,
      "grad_norm": 8.415993690490723,
      "learning_rate": 4.7949293365385505e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8503817662596702,
      "num_tokens": 107972692.0,
      "step": 89800
    },
    {
      "entropy": 1.7698960989713668,
      "epoch": 0.27840314207131506,
      "grad_norm": 8.861882209777832,
      "learning_rate": 4.794662377613515e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8597157940268516,
      "num_tokens": 107985556.0,
      "step": 89810
    },
    {
      "entropy": 1.7953774243593217,
      "epoch": 0.2784341411963647,
      "grad_norm": 7.687840461730957,
      "learning_rate": 4.794395463272534e-06,
      "loss": 0.4189,
      "mean_token_accuracy": 0.854545010626316,
      "num_tokens": 107998951.0,
      "step": 89820
    },
    {
      "entropy": 1.9417035043239594,
      "epoch": 0.27846514032141445,
      "grad_norm": 9.26181697845459,
      "learning_rate": 4.794128593503201e-06,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8420970395207406,
      "num_tokens": 108010519.0,
      "step": 89830
    },
    {
      "entropy": 1.8701779007911683,
      "epoch": 0.2784961394464641,
      "grad_norm": 8.857370376586914,
      "learning_rate": 4.793861768293114e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8434204265475274,
      "num_tokens": 108022433.0,
      "step": 89840
    },
    {
      "entropy": 1.9472114413976669,
      "epoch": 0.27852713857151384,
      "grad_norm": 8.817475318908691,
      "learning_rate": 4.793594987629871e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8399725124239922,
      "num_tokens": 108033484.0,
      "step": 89850
    },
    {
      "entropy": 1.9352742165327073,
      "epoch": 0.2785581376965635,
      "grad_norm": 7.381696701049805,
      "learning_rate": 4.7933282515010806e-06,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8374552443623543,
      "num_tokens": 108045222.0,
      "step": 89860
    },
    {
      "entropy": 1.8853474691510201,
      "epoch": 0.27858913682161324,
      "grad_norm": 8.79345417022705,
      "learning_rate": 4.793061559894352e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8477419853210449,
      "num_tokens": 108057330.0,
      "step": 89870
    },
    {
      "entropy": 1.8697724029421807,
      "epoch": 0.2786201359466629,
      "grad_norm": 8.972064018249512,
      "learning_rate": 4.792794912797302e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8470319598913193,
      "num_tokens": 108069573.0,
      "step": 89880
    },
    {
      "entropy": 1.8288535490632056,
      "epoch": 0.2786511350717126,
      "grad_norm": 3.9921863079071045,
      "learning_rate": 4.792528310197551e-06,
      "loss": 0.465,
      "mean_token_accuracy": 0.8418003395199776,
      "num_tokens": 108083028.0,
      "step": 89890
    },
    {
      "entropy": 1.9048464432358743,
      "epoch": 0.2786821341967623,
      "grad_norm": 8.971819877624512,
      "learning_rate": 4.792261752082724e-06,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8390678346157074,
      "num_tokens": 108094239.0,
      "step": 89900
    },
    {
      "entropy": 1.893374653160572,
      "epoch": 0.27871313332181197,
      "grad_norm": 8.717069625854492,
      "learning_rate": 4.791995238440452e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8402533918619156,
      "num_tokens": 108106085.0,
      "step": 89910
    },
    {
      "entropy": 1.9398944050073623,
      "epoch": 0.2787441324468617,
      "grad_norm": 9.565119743347168,
      "learning_rate": 4.79172876925837e-06,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8393180221319199,
      "num_tokens": 108116682.0,
      "step": 89920
    },
    {
      "entropy": 1.8258964017033577,
      "epoch": 0.27877513157191136,
      "grad_norm": 2.663538694381714,
      "learning_rate": 4.791462344524116e-06,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8499773174524308,
      "num_tokens": 108128639.0,
      "step": 89930
    },
    {
      "entropy": 1.833346499502659,
      "epoch": 0.2788061306969611,
      "grad_norm": 8.475390434265137,
      "learning_rate": 4.791195964225338e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8522562757134438,
      "num_tokens": 108141077.0,
      "step": 89940
    },
    {
      "entropy": 1.9464040279388428,
      "epoch": 0.27883712982201075,
      "grad_norm": 7.769806861877441,
      "learning_rate": 4.790929628349683e-06,
      "loss": 0.5211,
      "mean_token_accuracy": 0.838797104358673,
      "num_tokens": 108152453.0,
      "step": 89950
    },
    {
      "entropy": 1.8917251348495483,
      "epoch": 0.2788681289470605,
      "grad_norm": 7.752605438232422,
      "learning_rate": 4.790663336884804e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.8592374518513679,
      "num_tokens": 108164272.0,
      "step": 89960
    },
    {
      "entropy": 1.8999314427375793,
      "epoch": 0.27889912807211015,
      "grad_norm": 7.312995910644531,
      "learning_rate": 4.790397089818365e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8518079608678818,
      "num_tokens": 108175602.0,
      "step": 89970
    },
    {
      "entropy": 1.8947380736470223,
      "epoch": 0.27893012719715987,
      "grad_norm": 8.715499877929688,
      "learning_rate": 4.790130887138025e-06,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8417610317468643,
      "num_tokens": 108187295.0,
      "step": 89980
    },
    {
      "entropy": 1.8655477941036225,
      "epoch": 0.27896112632220954,
      "grad_norm": 8.097354888916016,
      "learning_rate": 4.789864728831455e-06,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8457545340061188,
      "num_tokens": 108199656.0,
      "step": 89990
    },
    {
      "entropy": 1.7254057943820953,
      "epoch": 0.27899212544725926,
      "grad_norm": 4.202314853668213,
      "learning_rate": 4.789598614886327e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8670785754919053,
      "num_tokens": 108213424.0,
      "step": 90000
    },
    {
      "entropy": 1.935379645228386,
      "epoch": 0.27902312457230893,
      "grad_norm": 4.619598865509033,
      "learning_rate": 4.789332545290321e-06,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8315211609005928,
      "num_tokens": 108224748.0,
      "step": 90010
    },
    {
      "entropy": 1.9281780689954757,
      "epoch": 0.27905412369735866,
      "grad_norm": 8.387505531311035,
      "learning_rate": 4.789066520031119e-06,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8352978438138962,
      "num_tokens": 108236160.0,
      "step": 90020
    },
    {
      "entropy": 1.9002047255635262,
      "epoch": 0.2790851228224083,
      "grad_norm": 8.676763534545898,
      "learning_rate": 4.7888005390964094e-06,
      "loss": 0.5236,
      "mean_token_accuracy": 0.834555535018444,
      "num_tokens": 108248149.0,
      "step": 90030
    },
    {
      "entropy": 1.8989999890327454,
      "epoch": 0.27911612194745805,
      "grad_norm": 9.32200813293457,
      "learning_rate": 4.788534602473885e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8412574380636215,
      "num_tokens": 108259859.0,
      "step": 90040
    },
    {
      "entropy": 1.8991701990365981,
      "epoch": 0.2791471210725077,
      "grad_norm": 7.9609694480896,
      "learning_rate": 4.788268710151243e-06,
      "loss": 0.4405,
      "mean_token_accuracy": 0.857192724943161,
      "num_tokens": 108271236.0,
      "step": 90050
    },
    {
      "entropy": 1.7937731251120568,
      "epoch": 0.27917812019755744,
      "grad_norm": 7.316855430603027,
      "learning_rate": 4.788002862116185e-06,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8526657864451408,
      "num_tokens": 108284781.0,
      "step": 90060
    },
    {
      "entropy": 1.9361607402563095,
      "epoch": 0.2792091193226071,
      "grad_norm": 7.928420066833496,
      "learning_rate": 4.787737058356419e-06,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8461529463529587,
      "num_tokens": 108295737.0,
      "step": 90070
    },
    {
      "entropy": 1.9357156157493591,
      "epoch": 0.27924011844765684,
      "grad_norm": 9.841268539428711,
      "learning_rate": 4.787471298859655e-06,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8431805863976478,
      "num_tokens": 108307828.0,
      "step": 90080
    },
    {
      "entropy": 1.8373299419879914,
      "epoch": 0.2792711175727065,
      "grad_norm": 8.356464385986328,
      "learning_rate": 4.78720558361361e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8455582305788993,
      "num_tokens": 108320724.0,
      "step": 90090
    },
    {
      "entropy": 1.9154756784439086,
      "epoch": 0.27930211669775623,
      "grad_norm": 3.6140196323394775,
      "learning_rate": 4.786939912606008e-06,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8417788296937943,
      "num_tokens": 108332507.0,
      "step": 90100
    },
    {
      "entropy": 1.8971551463007927,
      "epoch": 0.2793331158228059,
      "grad_norm": 8.835823059082031,
      "learning_rate": 4.786674285824571e-06,
      "loss": 0.484,
      "mean_token_accuracy": 0.8474727541208267,
      "num_tokens": 108343885.0,
      "step": 90110
    },
    {
      "entropy": 1.7721475332975387,
      "epoch": 0.2793641149478556,
      "grad_norm": 8.436137199401855,
      "learning_rate": 4.786408703257034e-06,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8569958195090294,
      "num_tokens": 108356767.0,
      "step": 90120
    },
    {
      "entropy": 1.8786687865853309,
      "epoch": 0.2793951140729053,
      "grad_norm": 7.322561740875244,
      "learning_rate": 4.78614316489113e-06,
      "loss": 0.474,
      "mean_token_accuracy": 0.8418064162135124,
      "num_tokens": 108369550.0,
      "step": 90130
    },
    {
      "entropy": 1.8602668032050134,
      "epoch": 0.27942611319795496,
      "grad_norm": 4.303462505340576,
      "learning_rate": 4.785877670714598e-06,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8566737055778504,
      "num_tokens": 108381931.0,
      "step": 90140
    },
    {
      "entropy": 1.869055911898613,
      "epoch": 0.2794571123230047,
      "grad_norm": 8.936867713928223,
      "learning_rate": 4.7856122207151874e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8521290734410286,
      "num_tokens": 108393434.0,
      "step": 90150
    },
    {
      "entropy": 1.9146001234650611,
      "epoch": 0.27948811144805435,
      "grad_norm": 10.076147079467773,
      "learning_rate": 4.7853468148806436e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8482694402337074,
      "num_tokens": 108404678.0,
      "step": 90160
    },
    {
      "entropy": 1.8209770336747169,
      "epoch": 0.2795191105731041,
      "grad_norm": 9.587752342224121,
      "learning_rate": 4.785081453198724e-06,
      "loss": 0.4182,
      "mean_token_accuracy": 0.8516710668802261,
      "num_tokens": 108416663.0,
      "step": 90170
    },
    {
      "entropy": 1.844451193511486,
      "epoch": 0.27955010969815375,
      "grad_norm": 10.955329895019531,
      "learning_rate": 4.784816135657187e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8469073712825775,
      "num_tokens": 108429075.0,
      "step": 90180
    },
    {
      "entropy": 1.8783362239599228,
      "epoch": 0.2795811088232035,
      "grad_norm": 8.352615356445312,
      "learning_rate": 4.784550862243798e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8481233865022659,
      "num_tokens": 108441399.0,
      "step": 90190
    },
    {
      "entropy": 1.9300066709518433,
      "epoch": 0.27961210794825314,
      "grad_norm": 7.3091206550598145,
      "learning_rate": 4.784285632946324e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8453634783625603,
      "num_tokens": 108452967.0,
      "step": 90200
    },
    {
      "entropy": 1.8706892415881158,
      "epoch": 0.27964310707330287,
      "grad_norm": 7.95927095413208,
      "learning_rate": 4.784020447752539e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8540063112974167,
      "num_tokens": 108464604.0,
      "step": 90210
    },
    {
      "entropy": 1.8064016848802567,
      "epoch": 0.27967410619835253,
      "grad_norm": 12.208187103271484,
      "learning_rate": 4.783755306650223e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8536765992641449,
      "num_tokens": 108477370.0,
      "step": 90220
    },
    {
      "entropy": 1.891729509830475,
      "epoch": 0.27970510532340226,
      "grad_norm": 7.544349193572998,
      "learning_rate": 4.783490209627159e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8449078306555748,
      "num_tokens": 108488578.0,
      "step": 90230
    },
    {
      "entropy": 1.839923305809498,
      "epoch": 0.2797361044484519,
      "grad_norm": 7.284799098968506,
      "learning_rate": 4.783225156671132e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8685827806591988,
      "num_tokens": 108500874.0,
      "step": 90240
    },
    {
      "entropy": 1.9023682996630669,
      "epoch": 0.27976710357350165,
      "grad_norm": 7.970053195953369,
      "learning_rate": 4.782960147769936e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8523323282599449,
      "num_tokens": 108512936.0,
      "step": 90250
    },
    {
      "entropy": 1.8668055430054664,
      "epoch": 0.2797981026985513,
      "grad_norm": 7.000694751739502,
      "learning_rate": 4.78269518291137e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8535307124257088,
      "num_tokens": 108524895.0,
      "step": 90260
    },
    {
      "entropy": 1.8698514148592948,
      "epoch": 0.27982910182360105,
      "grad_norm": 7.908920764923096,
      "learning_rate": 4.782430262083234e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8330338954925537,
      "num_tokens": 108537045.0,
      "step": 90270
    },
    {
      "entropy": 1.9352901756763459,
      "epoch": 0.2798601009486507,
      "grad_norm": 9.200328826904297,
      "learning_rate": 4.7821653852733365e-06,
      "loss": 0.4976,
      "mean_token_accuracy": 0.8495378881692887,
      "num_tokens": 108547879.0,
      "step": 90280
    },
    {
      "entropy": 1.8811449840664864,
      "epoch": 0.27989110007370044,
      "grad_norm": 8.855972290039062,
      "learning_rate": 4.781900552469487e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8424475163221359,
      "num_tokens": 108560224.0,
      "step": 90290
    },
    {
      "entropy": 1.7924820497632026,
      "epoch": 0.2799220991987501,
      "grad_norm": 3.9103078842163086,
      "learning_rate": 4.7816357636595036e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.856936690211296,
      "num_tokens": 108573946.0,
      "step": 90300
    },
    {
      "entropy": 1.9183641135692597,
      "epoch": 0.27995309832379983,
      "grad_norm": 4.683595180511475,
      "learning_rate": 4.781371018831206e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8507445871829986,
      "num_tokens": 108585889.0,
      "step": 90310
    },
    {
      "entropy": 1.882660059630871,
      "epoch": 0.2799840974488495,
      "grad_norm": 7.647256374359131,
      "learning_rate": 4.781106317972421e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8541607797145844,
      "num_tokens": 108597763.0,
      "step": 90320
    },
    {
      "entropy": 1.899367219209671,
      "epoch": 0.2800150965738992,
      "grad_norm": 7.681816577911377,
      "learning_rate": 4.780841661070978e-06,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8502333298325538,
      "num_tokens": 108609237.0,
      "step": 90330
    },
    {
      "entropy": 1.9607916057109833,
      "epoch": 0.2800460956989489,
      "grad_norm": 8.942265510559082,
      "learning_rate": 4.780577048114713e-06,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8450431019067765,
      "num_tokens": 108619807.0,
      "step": 90340
    },
    {
      "entropy": 1.925535424053669,
      "epoch": 0.2800770948239986,
      "grad_norm": 9.137941360473633,
      "learning_rate": 4.780312479091465e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8320911303162575,
      "num_tokens": 108630896.0,
      "step": 90350
    },
    {
      "entropy": 1.9415040105581283,
      "epoch": 0.2801080939490483,
      "grad_norm": 7.471439361572266,
      "learning_rate": 4.780047953989079e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.848981736600399,
      "num_tokens": 108642366.0,
      "step": 90360
    },
    {
      "entropy": 1.9614702731370925,
      "epoch": 0.280139093074098,
      "grad_norm": 9.385625839233398,
      "learning_rate": 4.779783472795404e-06,
      "loss": 0.5708,
      "mean_token_accuracy": 0.8339747205376625,
      "num_tokens": 108653416.0,
      "step": 90370
    },
    {
      "entropy": 1.7262531116604805,
      "epoch": 0.2801700921991477,
      "grad_norm": 8.864653587341309,
      "learning_rate": 4.779519035498294e-06,
      "loss": 0.3797,
      "mean_token_accuracy": 0.8626728132367134,
      "num_tokens": 108667678.0,
      "step": 90380
    },
    {
      "entropy": 1.9948803067207337,
      "epoch": 0.28020109132419735,
      "grad_norm": 8.669200897216797,
      "learning_rate": 4.779254642085608e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8431035861372947,
      "num_tokens": 108678385.0,
      "step": 90390
    },
    {
      "entropy": 1.9449063792824746,
      "epoch": 0.2802320904492471,
      "grad_norm": 9.006505966186523,
      "learning_rate": 4.778990292545207e-06,
      "loss": 0.5286,
      "mean_token_accuracy": 0.829691307246685,
      "num_tokens": 108689327.0,
      "step": 90400
    },
    {
      "entropy": 1.916630421578884,
      "epoch": 0.28026308957429674,
      "grad_norm": 8.61928653717041,
      "learning_rate": 4.7787259868649635e-06,
      "loss": 0.535,
      "mean_token_accuracy": 0.8399298146367074,
      "num_tokens": 108700905.0,
      "step": 90410
    },
    {
      "entropy": 1.9122631967067718,
      "epoch": 0.28029408869934647,
      "grad_norm": 7.261351108551025,
      "learning_rate": 4.778461725032747e-06,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8412989899516106,
      "num_tokens": 108713415.0,
      "step": 90420
    },
    {
      "entropy": 1.9562030732631683,
      "epoch": 0.28032508782439614,
      "grad_norm": 7.638638496398926,
      "learning_rate": 4.7781975070364375e-06,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8463377922773361,
      "num_tokens": 108723813.0,
      "step": 90430
    },
    {
      "entropy": 1.7865042075514794,
      "epoch": 0.28035608694944586,
      "grad_norm": 2.5158307552337646,
      "learning_rate": 4.7779333328639124e-06,
      "loss": 0.3875,
      "mean_token_accuracy": 0.8637947380542755,
      "num_tokens": 108737608.0,
      "step": 90440
    },
    {
      "entropy": 1.6872775062918663,
      "epoch": 0.28038708607449553,
      "grad_norm": 2.7983152866363525,
      "learning_rate": 4.777669202503063e-06,
      "loss": 0.3223,
      "mean_token_accuracy": 0.872826486825943,
      "num_tokens": 108752944.0,
      "step": 90450
    },
    {
      "entropy": 1.9655809059739113,
      "epoch": 0.28041808519954525,
      "grad_norm": 9.844002723693848,
      "learning_rate": 4.77740511594178e-06,
      "loss": 0.5496,
      "mean_token_accuracy": 0.8349737733602524,
      "num_tokens": 108764826.0,
      "step": 90460
    },
    {
      "entropy": 1.8685831755399704,
      "epoch": 0.2804490843245949,
      "grad_norm": 4.095294952392578,
      "learning_rate": 4.777141073167958e-06,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8481494188308716,
      "num_tokens": 108777223.0,
      "step": 90470
    },
    {
      "entropy": 1.9378087937831878,
      "epoch": 0.28048008344964465,
      "grad_norm": 3.9913816452026367,
      "learning_rate": 4.7768770741694985e-06,
      "loss": 0.521,
      "mean_token_accuracy": 0.8388440117239953,
      "num_tokens": 108788749.0,
      "step": 90480
    },
    {
      "entropy": 1.9008531831204891,
      "epoch": 0.2805110825746943,
      "grad_norm": 7.451270580291748,
      "learning_rate": 4.7766131189343075e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8565253868699074,
      "num_tokens": 108801607.0,
      "step": 90490
    },
    {
      "entropy": 1.960809737443924,
      "epoch": 0.28054208169974404,
      "grad_norm": 7.793551921844482,
      "learning_rate": 4.776349207450297e-06,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8530623838305473,
      "num_tokens": 108812575.0,
      "step": 90500
    },
    {
      "entropy": 1.9836690306663514,
      "epoch": 0.2805730808247937,
      "grad_norm": 9.390997886657715,
      "learning_rate": 4.776085339705378e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8447514802217484,
      "num_tokens": 108823707.0,
      "step": 90510
    },
    {
      "entropy": 1.8805051818490028,
      "epoch": 0.28060407994984343,
      "grad_norm": 6.977400302886963,
      "learning_rate": 4.775821515687472e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8450083822011948,
      "num_tokens": 108835768.0,
      "step": 90520
    },
    {
      "entropy": 1.8927241086959838,
      "epoch": 0.2806350790748931,
      "grad_norm": 8.844595909118652,
      "learning_rate": 4.775557735384503e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8455369025468826,
      "num_tokens": 108847447.0,
      "step": 90530
    },
    {
      "entropy": 1.9194593280553818,
      "epoch": 0.2806660781999428,
      "grad_norm": 9.10633659362793,
      "learning_rate": 4.775293998784402e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8458576589822769,
      "num_tokens": 108859245.0,
      "step": 90540
    },
    {
      "entropy": 1.9746450453996658,
      "epoch": 0.2806970773249925,
      "grad_norm": 8.2658052444458,
      "learning_rate": 4.775030305875099e-06,
      "loss": 0.5594,
      "mean_token_accuracy": 0.8276672974228859,
      "num_tokens": 108869967.0,
      "step": 90550
    },
    {
      "entropy": 1.8919481202960013,
      "epoch": 0.2807280764500422,
      "grad_norm": 6.916478157043457,
      "learning_rate": 4.774766656644536e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8547657161951066,
      "num_tokens": 108881653.0,
      "step": 90560
    },
    {
      "entropy": 1.8619554951786994,
      "epoch": 0.2807590755750919,
      "grad_norm": 7.139795780181885,
      "learning_rate": 4.774503051080653e-06,
      "loss": 0.4589,
      "mean_token_accuracy": 0.845686687529087,
      "num_tokens": 108893797.0,
      "step": 90570
    },
    {
      "entropy": 1.8943834751844406,
      "epoch": 0.2807900747001416,
      "grad_norm": 7.6240339279174805,
      "learning_rate": 4.7742394891713975e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.8450383573770524,
      "num_tokens": 108905569.0,
      "step": 90580
    },
    {
      "entropy": 1.8711055159568786,
      "epoch": 0.2808210738251913,
      "grad_norm": 7.671294689178467,
      "learning_rate": 4.773975970904725e-06,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8534323275089264,
      "num_tokens": 108918072.0,
      "step": 90590
    },
    {
      "entropy": 1.9521177858114243,
      "epoch": 0.280852072950241,
      "grad_norm": 9.559345245361328,
      "learning_rate": 4.773712496268588e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8375743925571442,
      "num_tokens": 108929589.0,
      "step": 90600
    },
    {
      "entropy": 1.8415945529937745,
      "epoch": 0.2808830720752907,
      "grad_norm": 4.66655969619751,
      "learning_rate": 4.773449065250952e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8464484736323357,
      "num_tokens": 108942319.0,
      "step": 90610
    },
    {
      "entropy": 1.8111029013991355,
      "epoch": 0.28091407120034034,
      "grad_norm": 7.22979211807251,
      "learning_rate": 4.77318567783978e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8576018720865249,
      "num_tokens": 108955141.0,
      "step": 90620
    },
    {
      "entropy": 1.910294608771801,
      "epoch": 0.28094507032539007,
      "grad_norm": 8.751194953918457,
      "learning_rate": 4.772922334023044e-06,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8544377252459526,
      "num_tokens": 108966874.0,
      "step": 90630
    },
    {
      "entropy": 1.8671040296554566,
      "epoch": 0.28097606945043974,
      "grad_norm": 3.474283218383789,
      "learning_rate": 4.7726590337887215e-06,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8581783235073089,
      "num_tokens": 108978553.0,
      "step": 90640
    },
    {
      "entropy": 1.8678077682852745,
      "epoch": 0.28100706857548946,
      "grad_norm": 10.035735130310059,
      "learning_rate": 4.772395777124789e-06,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8406292900443078,
      "num_tokens": 108990475.0,
      "step": 90650
    },
    {
      "entropy": 1.9179120391607285,
      "epoch": 0.28103806770053913,
      "grad_norm": 8.567628860473633,
      "learning_rate": 4.772132564019233e-06,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8443030267953873,
      "num_tokens": 109002069.0,
      "step": 90660
    },
    {
      "entropy": 1.9361174017190934,
      "epoch": 0.28106906682558885,
      "grad_norm": 8.406739234924316,
      "learning_rate": 4.7718693944600445e-06,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8500792890787124,
      "num_tokens": 109013697.0,
      "step": 90670
    },
    {
      "entropy": 1.8635503351688385,
      "epoch": 0.2811000659506385,
      "grad_norm": 9.684234619140625,
      "learning_rate": 4.771606268435215e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8503406763076782,
      "num_tokens": 109025070.0,
      "step": 90680
    },
    {
      "entropy": 1.773262333869934,
      "epoch": 0.28113106507568825,
      "grad_norm": 4.093050003051758,
      "learning_rate": 4.771343185932744e-06,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8514026969671249,
      "num_tokens": 109038769.0,
      "step": 90690
    },
    {
      "entropy": 1.8920626237988472,
      "epoch": 0.2811620642007379,
      "grad_norm": 8.63742446899414,
      "learning_rate": 4.771080146940636e-06,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8509817168116569,
      "num_tokens": 109050600.0,
      "step": 90700
    },
    {
      "entropy": 1.9327398404479026,
      "epoch": 0.28119306332578764,
      "grad_norm": 10.020679473876953,
      "learning_rate": 4.7708171514468965e-06,
      "loss": 0.528,
      "mean_token_accuracy": 0.842743456363678,
      "num_tokens": 109061795.0,
      "step": 90710
    },
    {
      "entropy": 1.8716965742409228,
      "epoch": 0.2812240624508373,
      "grad_norm": 9.19456958770752,
      "learning_rate": 4.770554199439541e-06,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8475604027509689,
      "num_tokens": 109073841.0,
      "step": 90720
    },
    {
      "entropy": 1.9363275811076164,
      "epoch": 0.28125506157588703,
      "grad_norm": 9.567742347717285,
      "learning_rate": 4.770291290906584e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8288316577672958,
      "num_tokens": 109084988.0,
      "step": 90730
    },
    {
      "entropy": 1.8694113805890082,
      "epoch": 0.2812860607009367,
      "grad_norm": 10.170321464538574,
      "learning_rate": 4.770028425836049e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.854419095814228,
      "num_tokens": 109097065.0,
      "step": 90740
    },
    {
      "entropy": 1.893789705634117,
      "epoch": 0.2813170598259864,
      "grad_norm": 8.507575035095215,
      "learning_rate": 4.769765604215961e-06,
      "loss": 0.4775,
      "mean_token_accuracy": 0.846349011361599,
      "num_tokens": 109108931.0,
      "step": 90750
    },
    {
      "entropy": 1.9919794470071792,
      "epoch": 0.2813480589510361,
      "grad_norm": 8.848701477050781,
      "learning_rate": 4.769502826034352e-06,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8403985217213631,
      "num_tokens": 109119893.0,
      "step": 90760
    },
    {
      "entropy": 1.826089233160019,
      "epoch": 0.2813790580760858,
      "grad_norm": 8.253718376159668,
      "learning_rate": 4.769240091279257e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8538560375571251,
      "num_tokens": 109132512.0,
      "step": 90770
    },
    {
      "entropy": 1.8694834753870964,
      "epoch": 0.2814100572011355,
      "grad_norm": 3.41013240814209,
      "learning_rate": 4.768977399938718e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8431166216731072,
      "num_tokens": 109145787.0,
      "step": 90780
    },
    {
      "entropy": 1.7938583612442016,
      "epoch": 0.2814410563261852,
      "grad_norm": 3.963634729385376,
      "learning_rate": 4.768714752000778e-06,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8582982420921326,
      "num_tokens": 109158903.0,
      "step": 90790
    },
    {
      "entropy": 1.9309082627296448,
      "epoch": 0.2814720554512349,
      "grad_norm": 9.012663841247559,
      "learning_rate": 4.768452147453487e-06,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8420499324798584,
      "num_tokens": 109169777.0,
      "step": 90800
    },
    {
      "entropy": 1.8647458493709563,
      "epoch": 0.2815030545762846,
      "grad_norm": 8.505784034729004,
      "learning_rate": 4.7681895862849e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8408839210867882,
      "num_tokens": 109182314.0,
      "step": 90810
    },
    {
      "entropy": 1.8462204396724702,
      "epoch": 0.2815340537013343,
      "grad_norm": 8.286667823791504,
      "learning_rate": 4.767927068483076e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8553971752524376,
      "num_tokens": 109195725.0,
      "step": 90820
    },
    {
      "entropy": 1.8974308609962462,
      "epoch": 0.281565052826384,
      "grad_norm": 9.34967041015625,
      "learning_rate": 4.767664594036074e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8490856289863586,
      "num_tokens": 109207842.0,
      "step": 90830
    },
    {
      "entropy": 1.8527883812785149,
      "epoch": 0.28159605195143367,
      "grad_norm": 3.996321439743042,
      "learning_rate": 4.767402162931967e-06,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8528962567448616,
      "num_tokens": 109220691.0,
      "step": 90840
    },
    {
      "entropy": 1.9433230310678482,
      "epoch": 0.2816270510764834,
      "grad_norm": 7.619417667388916,
      "learning_rate": 4.767139775158826e-06,
      "loss": 0.5239,
      "mean_token_accuracy": 0.843589824438095,
      "num_tokens": 109231579.0,
      "step": 90850
    },
    {
      "entropy": 1.8327447712421416,
      "epoch": 0.28165805020153306,
      "grad_norm": 8.732206344604492,
      "learning_rate": 4.766877430704727e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8554098963737488,
      "num_tokens": 109244327.0,
      "step": 90860
    },
    {
      "entropy": 1.9131013810634614,
      "epoch": 0.28168904932658273,
      "grad_norm": 8.691461563110352,
      "learning_rate": 4.766615129557752e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8473145887255669,
      "num_tokens": 109255888.0,
      "step": 90870
    },
    {
      "entropy": 1.9245838135480882,
      "epoch": 0.28172004845163245,
      "grad_norm": 7.54715633392334,
      "learning_rate": 4.766352871705987e-06,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8493126124143601,
      "num_tokens": 109266869.0,
      "step": 90880
    },
    {
      "entropy": 1.917286041378975,
      "epoch": 0.2817510475766821,
      "grad_norm": 4.127012252807617,
      "learning_rate": 4.7660906571375246e-06,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8439315423369408,
      "num_tokens": 109278743.0,
      "step": 90890
    },
    {
      "entropy": 1.897385112941265,
      "epoch": 0.28178204670173185,
      "grad_norm": 8.57779312133789,
      "learning_rate": 4.76582848584046e-06,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8471187829971314,
      "num_tokens": 109290444.0,
      "step": 90900
    },
    {
      "entropy": 1.9781205475330352,
      "epoch": 0.2818130458267815,
      "grad_norm": 9.53032398223877,
      "learning_rate": 4.765566357802891e-06,
      "loss": 0.5853,
      "mean_token_accuracy": 0.8293624818325043,
      "num_tokens": 109301356.0,
      "step": 90910
    },
    {
      "entropy": 1.9003041684627533,
      "epoch": 0.28184404495183124,
      "grad_norm": 8.813244819641113,
      "learning_rate": 4.765304273012924e-06,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8380188256502151,
      "num_tokens": 109313531.0,
      "step": 90920
    },
    {
      "entropy": 1.8687111303210258,
      "epoch": 0.2818750440768809,
      "grad_norm": 8.043350219726562,
      "learning_rate": 4.765042231458668e-06,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8497692689299583,
      "num_tokens": 109326158.0,
      "step": 90930
    },
    {
      "entropy": 1.9804033279418944,
      "epoch": 0.28190604320193063,
      "grad_norm": 8.524834632873535,
      "learning_rate": 4.764780233128236e-06,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8371674284338951,
      "num_tokens": 109337187.0,
      "step": 90940
    },
    {
      "entropy": 1.813350136578083,
      "epoch": 0.2819370423269803,
      "grad_norm": 7.912545204162598,
      "learning_rate": 4.764518278009748e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8572453141212464,
      "num_tokens": 109350343.0,
      "step": 90950
    },
    {
      "entropy": 1.9478640288114548,
      "epoch": 0.28196804145203,
      "grad_norm": 7.623589515686035,
      "learning_rate": 4.764256366091324e-06,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8418220117688179,
      "num_tokens": 109361694.0,
      "step": 90960
    },
    {
      "entropy": 1.8554902136325837,
      "epoch": 0.2819990405770797,
      "grad_norm": 7.590473651885986,
      "learning_rate": 4.763994497361095e-06,
      "loss": 0.435,
      "mean_token_accuracy": 0.8496796250343323,
      "num_tokens": 109374068.0,
      "step": 90970
    },
    {
      "entropy": 1.9092994675040245,
      "epoch": 0.2820300397021294,
      "grad_norm": 7.852284908294678,
      "learning_rate": 4.7637326718071905e-06,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8353525906801224,
      "num_tokens": 109385364.0,
      "step": 90980
    },
    {
      "entropy": 1.8934466361999511,
      "epoch": 0.2820610388271791,
      "grad_norm": 9.18950080871582,
      "learning_rate": 4.763470889417748e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8570297732949257,
      "num_tokens": 109397364.0,
      "step": 90990
    },
    {
      "entropy": 1.771867723762989,
      "epoch": 0.2820920379522288,
      "grad_norm": 9.84312629699707,
      "learning_rate": 4.763209150180908e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8599537044763566,
      "num_tokens": 109410971.0,
      "step": 91000
    },
    {
      "entropy": 1.9751540750265122,
      "epoch": 0.2821230370772785,
      "grad_norm": 10.806888580322266,
      "learning_rate": 4.762947454084818e-06,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8323762461543083,
      "num_tokens": 109421894.0,
      "step": 91010
    },
    {
      "entropy": 1.863322387635708,
      "epoch": 0.2821540362023282,
      "grad_norm": 9.505192756652832,
      "learning_rate": 4.7626858011176256e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8461381018161773,
      "num_tokens": 109433667.0,
      "step": 91020
    },
    {
      "entropy": 1.9372095853090285,
      "epoch": 0.2821850353273779,
      "grad_norm": 9.51073169708252,
      "learning_rate": 4.7624241912674885e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8425448566675187,
      "num_tokens": 109445506.0,
      "step": 91030
    },
    {
      "entropy": 1.8736591801047324,
      "epoch": 0.2822160344524276,
      "grad_norm": 9.874513626098633,
      "learning_rate": 4.762162624522564e-06,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8544930562376976,
      "num_tokens": 109456672.0,
      "step": 91040
    },
    {
      "entropy": 1.9017417326569557,
      "epoch": 0.28224703357747727,
      "grad_norm": 8.883088111877441,
      "learning_rate": 4.761901100871018e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8446088701486587,
      "num_tokens": 109468440.0,
      "step": 91050
    },
    {
      "entropy": 1.8893210887908936,
      "epoch": 0.282278032702527,
      "grad_norm": 10.458946228027344,
      "learning_rate": 4.7616396203010165e-06,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8509502872824669,
      "num_tokens": 109480575.0,
      "step": 91060
    },
    {
      "entropy": 1.9682177215814591,
      "epoch": 0.28230903182757666,
      "grad_norm": 7.0361151695251465,
      "learning_rate": 4.761378182800733e-06,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8456061512231827,
      "num_tokens": 109491254.0,
      "step": 91070
    },
    {
      "entropy": 1.9273748084902764,
      "epoch": 0.2823400309526264,
      "grad_norm": 7.263545989990234,
      "learning_rate": 4.761116788358349e-06,
      "loss": 0.501,
      "mean_token_accuracy": 0.8323819488286972,
      "num_tokens": 109503011.0,
      "step": 91080
    },
    {
      "entropy": 1.9739017739892006,
      "epoch": 0.28237103007767606,
      "grad_norm": 6.576552391052246,
      "learning_rate": 4.760855436962041e-06,
      "loss": 0.5487,
      "mean_token_accuracy": 0.8396146357059479,
      "num_tokens": 109514103.0,
      "step": 91090
    },
    {
      "entropy": 1.9086976170539856,
      "epoch": 0.2824020292027258,
      "grad_norm": 8.7454833984375,
      "learning_rate": 4.760594128599999e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.848183062672615,
      "num_tokens": 109526563.0,
      "step": 91100
    },
    {
      "entropy": 1.9339544415473937,
      "epoch": 0.28243302832777545,
      "grad_norm": 9.544751167297363,
      "learning_rate": 4.760332863260414e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8470303192734718,
      "num_tokens": 109538093.0,
      "step": 91110
    },
    {
      "entropy": 1.916434782743454,
      "epoch": 0.2824640274528251,
      "grad_norm": 4.14872407913208,
      "learning_rate": 4.7600716409314804e-06,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8473877355456352,
      "num_tokens": 109549378.0,
      "step": 91120
    },
    {
      "entropy": 1.766272282600403,
      "epoch": 0.28249502657787484,
      "grad_norm": 8.772634506225586,
      "learning_rate": 4.7598104616014005e-06,
      "loss": 0.3786,
      "mean_token_accuracy": 0.8664262443780899,
      "num_tokens": 109563346.0,
      "step": 91130
    },
    {
      "entropy": 1.8620783016085625,
      "epoch": 0.2825260257029245,
      "grad_norm": 5.1654887199401855,
      "learning_rate": 4.759549325258377e-06,
      "loss": 0.4479,
      "mean_token_accuracy": 0.8528469145298004,
      "num_tokens": 109575000.0,
      "step": 91140
    },
    {
      "entropy": 1.827536989748478,
      "epoch": 0.28255702482797423,
      "grad_norm": 4.431619644165039,
      "learning_rate": 4.759288231890621e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.849473150074482,
      "num_tokens": 109587804.0,
      "step": 91150
    },
    {
      "entropy": 1.865426352620125,
      "epoch": 0.2825880239530239,
      "grad_norm": 7.909317493438721,
      "learning_rate": 4.759027181486346e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8499869540333748,
      "num_tokens": 109600294.0,
      "step": 91160
    },
    {
      "entropy": 1.9297534614801406,
      "epoch": 0.28261902307807363,
      "grad_norm": 9.111772537231445,
      "learning_rate": 4.758766174033769e-06,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8378744632005691,
      "num_tokens": 109611503.0,
      "step": 91170
    },
    {
      "entropy": 1.8704712957143783,
      "epoch": 0.2826500222031233,
      "grad_norm": 4.173186779022217,
      "learning_rate": 4.758505209521114e-06,
      "loss": 0.4431,
      "mean_token_accuracy": 0.855632272362709,
      "num_tokens": 109624455.0,
      "step": 91180
    },
    {
      "entropy": 1.9280048042535782,
      "epoch": 0.282681021328173,
      "grad_norm": 8.439362525939941,
      "learning_rate": 4.758244287936609e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8422107562422753,
      "num_tokens": 109635564.0,
      "step": 91190
    },
    {
      "entropy": 1.8654653757810593,
      "epoch": 0.2827120204532227,
      "grad_norm": 4.882840156555176,
      "learning_rate": 4.757983409268485e-06,
      "loss": 0.47,
      "mean_token_accuracy": 0.8396533340215683,
      "num_tokens": 109648550.0,
      "step": 91200
    },
    {
      "entropy": 1.8784069836139679,
      "epoch": 0.2827430195782724,
      "grad_norm": 7.505212783813477,
      "learning_rate": 4.757722573504979e-06,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8495945394039154,
      "num_tokens": 109660589.0,
      "step": 91210
    },
    {
      "entropy": 1.9778222680091857,
      "epoch": 0.2827740187033221,
      "grad_norm": 10.172992706298828,
      "learning_rate": 4.757461780634332e-06,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8417175248265266,
      "num_tokens": 109671973.0,
      "step": 91220
    },
    {
      "entropy": 1.8196046486496926,
      "epoch": 0.2828050178283718,
      "grad_norm": 8.086063385009766,
      "learning_rate": 4.757201030644789e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8687186688184738,
      "num_tokens": 109684750.0,
      "step": 91230
    },
    {
      "entropy": 1.8343065902590752,
      "epoch": 0.2828360169534215,
      "grad_norm": 9.45622730255127,
      "learning_rate": 4.7569403235246005e-06,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8519006326794625,
      "num_tokens": 109697003.0,
      "step": 91240
    },
    {
      "entropy": 1.838625229895115,
      "epoch": 0.2828670160784712,
      "grad_norm": 3.578289270401001,
      "learning_rate": 4.756679659262021e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8600620925426483,
      "num_tokens": 109709959.0,
      "step": 91250
    },
    {
      "entropy": 1.9180777609348296,
      "epoch": 0.28289801520352087,
      "grad_norm": 9.219250679016113,
      "learning_rate": 4.756419037845309e-06,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8345475569367409,
      "num_tokens": 109721437.0,
      "step": 91260
    },
    {
      "entropy": 1.914302496612072,
      "epoch": 0.2829290143285706,
      "grad_norm": 3.8321692943573,
      "learning_rate": 4.756158459262729e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8463889390230179,
      "num_tokens": 109733406.0,
      "step": 91270
    },
    {
      "entropy": 1.841091763973236,
      "epoch": 0.28296001345362026,
      "grad_norm": 8.805869102478027,
      "learning_rate": 4.755897923502547e-06,
      "loss": 0.4353,
      "mean_token_accuracy": 0.8465237930417061,
      "num_tokens": 109745414.0,
      "step": 91280
    },
    {
      "entropy": 1.9549955561757089,
      "epoch": 0.28299101257867,
      "grad_norm": 4.771459102630615,
      "learning_rate": 4.755637430553038e-06,
      "loss": 0.5449,
      "mean_token_accuracy": 0.8281950682401658,
      "num_tokens": 109756473.0,
      "step": 91290
    },
    {
      "entropy": 1.9814363867044449,
      "epoch": 0.28302201170371966,
      "grad_norm": 9.486200332641602,
      "learning_rate": 4.755376980402479e-06,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8435119092464447,
      "num_tokens": 109767033.0,
      "step": 91300
    },
    {
      "entropy": 1.8801933750510216,
      "epoch": 0.2830530108287694,
      "grad_norm": 8.066265106201172,
      "learning_rate": 4.755116573039149e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8474684238433838,
      "num_tokens": 109779009.0,
      "step": 91310
    },
    {
      "entropy": 1.8336575701832771,
      "epoch": 0.28308400995381905,
      "grad_norm": 9.574748039245605,
      "learning_rate": 4.754856208451337e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8556732803583145,
      "num_tokens": 109791334.0,
      "step": 91320
    },
    {
      "entropy": 1.9349465638399124,
      "epoch": 0.2831150090788688,
      "grad_norm": 8.522025108337402,
      "learning_rate": 4.75459588662733e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.851339441537857,
      "num_tokens": 109802925.0,
      "step": 91330
    },
    {
      "entropy": 1.8921233609318733,
      "epoch": 0.28314600820391844,
      "grad_norm": 8.379258155822754,
      "learning_rate": 4.754335607555427e-06,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8478038221597671,
      "num_tokens": 109814130.0,
      "step": 91340
    },
    {
      "entropy": 1.8570956602692603,
      "epoch": 0.28317700732896817,
      "grad_norm": 9.117257118225098,
      "learning_rate": 4.754075371223925e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.8518408805131912,
      "num_tokens": 109826200.0,
      "step": 91350
    },
    {
      "entropy": 1.91699261367321,
      "epoch": 0.28320800645401784,
      "grad_norm": 8.286026000976562,
      "learning_rate": 4.753815177621128e-06,
      "loss": 0.539,
      "mean_token_accuracy": 0.8423243030905724,
      "num_tokens": 109837290.0,
      "step": 91360
    },
    {
      "entropy": 1.940002153813839,
      "epoch": 0.2832390055790675,
      "grad_norm": 9.196231842041016,
      "learning_rate": 4.753555026735344e-06,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8430562511086463,
      "num_tokens": 109848408.0,
      "step": 91370
    },
    {
      "entropy": 1.9086173102259636,
      "epoch": 0.28327000470411723,
      "grad_norm": 7.4503560066223145,
      "learning_rate": 4.753294918554887e-06,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8556687757372856,
      "num_tokens": 109859863.0,
      "step": 91380
    },
    {
      "entropy": 1.8716721430420875,
      "epoch": 0.2833010038291669,
      "grad_norm": 7.7145795822143555,
      "learning_rate": 4.753034853068076e-06,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8355464920401573,
      "num_tokens": 109872862.0,
      "step": 91390
    },
    {
      "entropy": 1.959967464208603,
      "epoch": 0.2833320029542166,
      "grad_norm": 8.618791580200195,
      "learning_rate": 4.752774830263229e-06,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8505872398614883,
      "num_tokens": 109884502.0,
      "step": 91400
    },
    {
      "entropy": 1.8670057207345963,
      "epoch": 0.2833630020792663,
      "grad_norm": 2.768303155899048,
      "learning_rate": 4.7525148501286754e-06,
      "loss": 0.514,
      "mean_token_accuracy": 0.841192239522934,
      "num_tokens": 109896997.0,
      "step": 91410
    },
    {
      "entropy": 1.942854182422161,
      "epoch": 0.283394001204316,
      "grad_norm": 9.517045974731445,
      "learning_rate": 4.752254912652746e-06,
      "loss": 0.5407,
      "mean_token_accuracy": 0.8302263349294663,
      "num_tokens": 109908513.0,
      "step": 91420
    },
    {
      "entropy": 1.8470983251929283,
      "epoch": 0.2834250003293657,
      "grad_norm": 3.9292147159576416,
      "learning_rate": 4.751995017823772e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.847145140171051,
      "num_tokens": 109921038.0,
      "step": 91430
    },
    {
      "entropy": 1.889025342464447,
      "epoch": 0.2834559994544154,
      "grad_norm": 7.8180952072143555,
      "learning_rate": 4.751735165630099e-06,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8388582825660705,
      "num_tokens": 109933509.0,
      "step": 91440
    },
    {
      "entropy": 1.926556906104088,
      "epoch": 0.2834869985794651,
      "grad_norm": 9.13294506072998,
      "learning_rate": 4.751475356060067e-06,
      "loss": 0.502,
      "mean_token_accuracy": 0.8347544342279434,
      "num_tokens": 109945589.0,
      "step": 91450
    },
    {
      "entropy": 1.8450331330299377,
      "epoch": 0.2835179977045148,
      "grad_norm": 4.0597100257873535,
      "learning_rate": 4.751215589102026e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8560441970825196,
      "num_tokens": 109958238.0,
      "step": 91460
    },
    {
      "entropy": 1.8735743075609208,
      "epoch": 0.28354899682956447,
      "grad_norm": 9.06523323059082,
      "learning_rate": 4.75095586474433e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8512053444981575,
      "num_tokens": 109970897.0,
      "step": 91470
    },
    {
      "entropy": 1.9218522995710372,
      "epoch": 0.2835799959546142,
      "grad_norm": 8.556892395019531,
      "learning_rate": 4.750696182975335e-06,
      "loss": 0.5451,
      "mean_token_accuracy": 0.8374419540166855,
      "num_tokens": 109982852.0,
      "step": 91480
    },
    {
      "entropy": 1.8760737299919128,
      "epoch": 0.28361099507966386,
      "grad_norm": 7.687764644622803,
      "learning_rate": 4.750436543783403e-06,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8507323205471039,
      "num_tokens": 109993887.0,
      "step": 91490
    },
    {
      "entropy": 1.8525379657745362,
      "epoch": 0.2836419942047136,
      "grad_norm": 8.57304573059082,
      "learning_rate": 4.750176947156903e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8570681139826775,
      "num_tokens": 110005619.0,
      "step": 91500
    },
    {
      "entropy": 1.9060514703392983,
      "epoch": 0.28367299332976326,
      "grad_norm": 6.858503818511963,
      "learning_rate": 4.749917393084203e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8408376634120941,
      "num_tokens": 110018405.0,
      "step": 91510
    },
    {
      "entropy": 1.8777115240693092,
      "epoch": 0.283703992454813,
      "grad_norm": 7.576192855834961,
      "learning_rate": 4.74965788155368e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8432316944003105,
      "num_tokens": 110030683.0,
      "step": 91520
    },
    {
      "entropy": 1.6769248962402343,
      "epoch": 0.28373499157986265,
      "grad_norm": 8.123210906982422,
      "learning_rate": 4.749398412553713e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8673816755414009,
      "num_tokens": 110045730.0,
      "step": 91530
    },
    {
      "entropy": 1.8561891838908195,
      "epoch": 0.2837659907049124,
      "grad_norm": 4.347743511199951,
      "learning_rate": 4.749138986072685e-06,
      "loss": 0.4555,
      "mean_token_accuracy": 0.849002268910408,
      "num_tokens": 110058293.0,
      "step": 91540
    },
    {
      "entropy": 1.87572433501482,
      "epoch": 0.28379698982996204,
      "grad_norm": 8.649298667907715,
      "learning_rate": 4.748879602098988e-06,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8493537470698357,
      "num_tokens": 110070569.0,
      "step": 91550
    },
    {
      "entropy": 1.8268972262740135,
      "epoch": 0.28382798895501177,
      "grad_norm": 3.4365265369415283,
      "learning_rate": 4.748620260621013e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8539086386561394,
      "num_tokens": 110083590.0,
      "step": 91560
    },
    {
      "entropy": 1.8347914576530457,
      "epoch": 0.28385898808006144,
      "grad_norm": 7.83189582824707,
      "learning_rate": 4.748360961627159e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8629049167037011,
      "num_tokens": 110096116.0,
      "step": 91570
    },
    {
      "entropy": 1.9295135840773583,
      "epoch": 0.28388998720511116,
      "grad_norm": 10.846009254455566,
      "learning_rate": 4.748101705105827e-06,
      "loss": 0.537,
      "mean_token_accuracy": 0.8269555777311325,
      "num_tokens": 110107595.0,
      "step": 91580
    },
    {
      "entropy": 1.9408522367477417,
      "epoch": 0.28392098633016083,
      "grad_norm": 7.8314056396484375,
      "learning_rate": 4.747842491045421e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8457581490278244,
      "num_tokens": 110118966.0,
      "step": 91590
    },
    {
      "entropy": 1.8445065826177598,
      "epoch": 0.28395198545521055,
      "grad_norm": 9.18549633026123,
      "learning_rate": 4.747583319434357e-06,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8425332620739937,
      "num_tokens": 110131790.0,
      "step": 91600
    },
    {
      "entropy": 1.8247267931699753,
      "epoch": 0.2839829845802602,
      "grad_norm": 9.114663124084473,
      "learning_rate": 4.747324190261046e-06,
      "loss": 0.491,
      "mean_token_accuracy": 0.8539474830031395,
      "num_tokens": 110144807.0,
      "step": 91610
    },
    {
      "entropy": 1.859029544889927,
      "epoch": 0.2840139837053099,
      "grad_norm": 9.042014122009277,
      "learning_rate": 4.74706510351391e-06,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8466860115528106,
      "num_tokens": 110157391.0,
      "step": 91620
    },
    {
      "entropy": 1.8583563596010209,
      "epoch": 0.2840449828303596,
      "grad_norm": 8.334424018859863,
      "learning_rate": 4.746806059181373e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8355572551488877,
      "num_tokens": 110170333.0,
      "step": 91630
    },
    {
      "entropy": 1.8378588289022446,
      "epoch": 0.2840759819554093,
      "grad_norm": 8.690698623657227,
      "learning_rate": 4.746547057251862e-06,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8404997318983078,
      "num_tokens": 110183049.0,
      "step": 91640
    },
    {
      "entropy": 1.8678978830575943,
      "epoch": 0.284106981080459,
      "grad_norm": 4.2940473556518555,
      "learning_rate": 4.7462880977138126e-06,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8339786469936371,
      "num_tokens": 110196329.0,
      "step": 91650
    },
    {
      "entropy": 1.7718224942684173,
      "epoch": 0.2841379802055087,
      "grad_norm": 8.046103477478027,
      "learning_rate": 4.74602918055566e-06,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8618276312947273,
      "num_tokens": 110209525.0,
      "step": 91660
    },
    {
      "entropy": 1.9796247810125351,
      "epoch": 0.2841689793305584,
      "grad_norm": 6.6610870361328125,
      "learning_rate": 4.745770305765847e-06,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8357976496219635,
      "num_tokens": 110220480.0,
      "step": 91670
    },
    {
      "entropy": 1.9403215855360032,
      "epoch": 0.28419997845560807,
      "grad_norm": 8.975007057189941,
      "learning_rate": 4.745511473332818e-06,
      "loss": 0.5478,
      "mean_token_accuracy": 0.8363938242197037,
      "num_tokens": 110231525.0,
      "step": 91680
    },
    {
      "entropy": 1.891119834780693,
      "epoch": 0.2842309775806578,
      "grad_norm": 9.538240432739258,
      "learning_rate": 4.745252683245027e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8425269886851311,
      "num_tokens": 110243198.0,
      "step": 91690
    },
    {
      "entropy": 1.8871520176529883,
      "epoch": 0.28426197670570746,
      "grad_norm": 9.343374252319336,
      "learning_rate": 4.744993935490928e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8517936706542969,
      "num_tokens": 110255593.0,
      "step": 91700
    },
    {
      "entropy": 1.8104426577687263,
      "epoch": 0.2842929758307572,
      "grad_norm": 8.839599609375,
      "learning_rate": 4.744735230058977e-06,
      "loss": 0.4135,
      "mean_token_accuracy": 0.8551136195659638,
      "num_tokens": 110268721.0,
      "step": 91710
    },
    {
      "entropy": 1.9426888212561608,
      "epoch": 0.28432397495580686,
      "grad_norm": 8.519753456115723,
      "learning_rate": 4.744476566937642e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8359502226114273,
      "num_tokens": 110280087.0,
      "step": 91720
    },
    {
      "entropy": 1.9278660222887993,
      "epoch": 0.2843549740808566,
      "grad_norm": 8.69306468963623,
      "learning_rate": 4.74421794611539e-06,
      "loss": 0.497,
      "mean_token_accuracy": 0.8463233426213265,
      "num_tokens": 110291541.0,
      "step": 91730
    },
    {
      "entropy": 1.9058746635913848,
      "epoch": 0.28438597320590625,
      "grad_norm": 4.05926513671875,
      "learning_rate": 4.743959367580693e-06,
      "loss": 0.4611,
      "mean_token_accuracy": 0.8437731295824051,
      "num_tokens": 110303659.0,
      "step": 91740
    },
    {
      "entropy": 1.9244377925992011,
      "epoch": 0.284416972330956,
      "grad_norm": 8.101446151733398,
      "learning_rate": 4.743700831322029e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8459056586027145,
      "num_tokens": 110315031.0,
      "step": 91750
    },
    {
      "entropy": 1.9023540601134301,
      "epoch": 0.28444797145600564,
      "grad_norm": 6.504104137420654,
      "learning_rate": 4.74344233732788e-06,
      "loss": 0.4976,
      "mean_token_accuracy": 0.8424684196710587,
      "num_tokens": 110327386.0,
      "step": 91760
    },
    {
      "entropy": 1.9569046169519424,
      "epoch": 0.28447897058105537,
      "grad_norm": 9.210738182067871,
      "learning_rate": 4.743183885586729e-06,
      "loss": 0.537,
      "mean_token_accuracy": 0.8266186848282814,
      "num_tokens": 110338648.0,
      "step": 91770
    },
    {
      "entropy": 1.8510024085640908,
      "epoch": 0.28450996970610504,
      "grad_norm": 8.181108474731445,
      "learning_rate": 4.74292547608707e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8562486320734024,
      "num_tokens": 110350736.0,
      "step": 91780
    },
    {
      "entropy": 1.9186692774295806,
      "epoch": 0.28454096883115476,
      "grad_norm": 9.964049339294434,
      "learning_rate": 4.7426671088173945e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8494500458240509,
      "num_tokens": 110362275.0,
      "step": 91790
    },
    {
      "entropy": 1.7676223665475845,
      "epoch": 0.28457196795620443,
      "grad_norm": 2.426920175552368,
      "learning_rate": 4.742408783766203e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.863847254216671,
      "num_tokens": 110375617.0,
      "step": 91800
    },
    {
      "entropy": 1.8544678494334221,
      "epoch": 0.28460296708125415,
      "grad_norm": 9.724370956420898,
      "learning_rate": 4.742150500922e-06,
      "loss": 0.484,
      "mean_token_accuracy": 0.8445950224995613,
      "num_tokens": 110387605.0,
      "step": 91810
    },
    {
      "entropy": 1.8677785605192185,
      "epoch": 0.2846339662063038,
      "grad_norm": 8.32010269165039,
      "learning_rate": 4.741892260273291e-06,
      "loss": 0.4568,
      "mean_token_accuracy": 0.845884545147419,
      "num_tokens": 110399589.0,
      "step": 91820
    },
    {
      "entropy": 1.8281821206212043,
      "epoch": 0.28466496533135355,
      "grad_norm": 8.002693176269531,
      "learning_rate": 4.741634061808588e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8603383213281631,
      "num_tokens": 110412247.0,
      "step": 91830
    },
    {
      "entropy": 1.8158651649951936,
      "epoch": 0.2846959644564032,
      "grad_norm": 8.582942008972168,
      "learning_rate": 4.741375905516411e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.846869707107544,
      "num_tokens": 110424355.0,
      "step": 91840
    },
    {
      "entropy": 1.7574502289295197,
      "epoch": 0.28472696358145294,
      "grad_norm": 4.905912399291992,
      "learning_rate": 4.741117791385276e-06,
      "loss": 0.3859,
      "mean_token_accuracy": 0.8498795077204704,
      "num_tokens": 110437595.0,
      "step": 91850
    },
    {
      "entropy": 1.898454374074936,
      "epoch": 0.2847579627065026,
      "grad_norm": 7.610432147979736,
      "learning_rate": 4.740859719403713e-06,
      "loss": 0.492,
      "mean_token_accuracy": 0.8487632632255554,
      "num_tokens": 110448902.0,
      "step": 91860
    },
    {
      "entropy": 1.7625425659120082,
      "epoch": 0.2847889618315523,
      "grad_norm": 4.429518699645996,
      "learning_rate": 4.740601689560249e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8575651466846466,
      "num_tokens": 110461975.0,
      "step": 91870
    },
    {
      "entropy": 1.8556936159729958,
      "epoch": 0.284819960956602,
      "grad_norm": 9.066593170166016,
      "learning_rate": 4.74034370184342e-06,
      "loss": 0.4796,
      "mean_token_accuracy": 0.847280016541481,
      "num_tokens": 110473152.0,
      "step": 91880
    },
    {
      "entropy": 1.880302868783474,
      "epoch": 0.28485096008165167,
      "grad_norm": 9.606684684753418,
      "learning_rate": 4.740085756241761e-06,
      "loss": 0.5205,
      "mean_token_accuracy": 0.831180626153946,
      "num_tokens": 110484515.0,
      "step": 91890
    },
    {
      "entropy": 1.768574671447277,
      "epoch": 0.2848819592067014,
      "grad_norm": 3.897165298461914,
      "learning_rate": 4.7398278527438175e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8571337282657623,
      "num_tokens": 110496957.0,
      "step": 91900
    },
    {
      "entropy": 1.791210974752903,
      "epoch": 0.28491295833175106,
      "grad_norm": 9.069717407226562,
      "learning_rate": 4.739569991338137e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.851265873014927,
      "num_tokens": 110509726.0,
      "step": 91910
    },
    {
      "entropy": 1.807637719810009,
      "epoch": 0.2849439574568008,
      "grad_norm": 7.900302886962891,
      "learning_rate": 4.739312172013269e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8464860022068024,
      "num_tokens": 110522419.0,
      "step": 91920
    },
    {
      "entropy": 1.8903899610042572,
      "epoch": 0.28497495658185046,
      "grad_norm": 6.663649559020996,
      "learning_rate": 4.7390543947577705e-06,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8373034983873368,
      "num_tokens": 110533897.0,
      "step": 91930
    },
    {
      "entropy": 1.8422768160700798,
      "epoch": 0.2850059557069002,
      "grad_norm": 7.8179826736450195,
      "learning_rate": 4.7387966595602014e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8479554176330566,
      "num_tokens": 110546623.0,
      "step": 91940
    },
    {
      "entropy": 1.7804616317152977,
      "epoch": 0.28503695483194985,
      "grad_norm": 8.71053409576416,
      "learning_rate": 4.738538966409126e-06,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8510273039340973,
      "num_tokens": 110559478.0,
      "step": 91950
    },
    {
      "entropy": 1.859451201558113,
      "epoch": 0.2850679539569996,
      "grad_norm": 4.298933029174805,
      "learning_rate": 4.738281315293114e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8425625443458558,
      "num_tokens": 110571704.0,
      "step": 91960
    },
    {
      "entropy": 1.850043423473835,
      "epoch": 0.28509895308204924,
      "grad_norm": 7.2485480308532715,
      "learning_rate": 4.738023706200738e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.851256474852562,
      "num_tokens": 110583945.0,
      "step": 91970
    },
    {
      "entropy": 1.9050755083560944,
      "epoch": 0.28512995220709897,
      "grad_norm": 9.636146545410156,
      "learning_rate": 4.737766139120575e-06,
      "loss": 0.5271,
      "mean_token_accuracy": 0.832217988371849,
      "num_tokens": 110595560.0,
      "step": 91980
    },
    {
      "entropy": 1.9206005334854126,
      "epoch": 0.28516095133214864,
      "grad_norm": 8.515999794006348,
      "learning_rate": 4.73750861404121e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8435544535517693,
      "num_tokens": 110606577.0,
      "step": 91990
    },
    {
      "entropy": 1.869492068886757,
      "epoch": 0.28519195045719836,
      "grad_norm": 8.025008201599121,
      "learning_rate": 4.737251130951226e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8449269160628319,
      "num_tokens": 110618391.0,
      "step": 92000
    },
    {
      "entropy": 1.845357683300972,
      "epoch": 0.28522294958224803,
      "grad_norm": 8.134385108947754,
      "learning_rate": 4.736993689839216e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8439294084906578,
      "num_tokens": 110631021.0,
      "step": 92010
    },
    {
      "entropy": 1.8097378730773925,
      "epoch": 0.28525394870729776,
      "grad_norm": 10.888206481933594,
      "learning_rate": 4.736736290693772e-06,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8568091303110122,
      "num_tokens": 110643427.0,
      "step": 92020
    },
    {
      "entropy": 1.9278713300824166,
      "epoch": 0.2852849478323474,
      "grad_norm": 8.461037635803223,
      "learning_rate": 4.736478933503496e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.8417813435196877,
      "num_tokens": 110654675.0,
      "step": 92030
    },
    {
      "entropy": 1.8513173662126063,
      "epoch": 0.28531594695739715,
      "grad_norm": 8.097729682922363,
      "learning_rate": 4.7362216182569906e-06,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8500329554080963,
      "num_tokens": 110667292.0,
      "step": 92040
    },
    {
      "entropy": 1.9195297732949257,
      "epoch": 0.2853469460824468,
      "grad_norm": 8.111653327941895,
      "learning_rate": 4.735964344942864e-06,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8443229928612709,
      "num_tokens": 110678569.0,
      "step": 92050
    },
    {
      "entropy": 1.8122631691396236,
      "epoch": 0.28537794520749654,
      "grad_norm": 8.886815071105957,
      "learning_rate": 4.735707113549729e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8556604027748108,
      "num_tokens": 110691490.0,
      "step": 92060
    },
    {
      "entropy": 1.884895347058773,
      "epoch": 0.2854089443325462,
      "grad_norm": 10.424093246459961,
      "learning_rate": 4.735449924066201e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8439397796988487,
      "num_tokens": 110703884.0,
      "step": 92070
    },
    {
      "entropy": 1.9233110576868058,
      "epoch": 0.28543994345759593,
      "grad_norm": 9.017784118652344,
      "learning_rate": 4.735192776480902e-06,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8422451555728913,
      "num_tokens": 110716070.0,
      "step": 92080
    },
    {
      "entropy": 1.8380617439746856,
      "epoch": 0.2854709425826456,
      "grad_norm": 8.3659086227417,
      "learning_rate": 4.734935670782457e-06,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8482602387666702,
      "num_tokens": 110728016.0,
      "step": 92090
    },
    {
      "entropy": 1.9306314051151277,
      "epoch": 0.2855019417076953,
      "grad_norm": 9.452094078063965,
      "learning_rate": 4.7346786069594955e-06,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8452840596437454,
      "num_tokens": 110739218.0,
      "step": 92100
    },
    {
      "entropy": 1.7319670930504798,
      "epoch": 0.285532940832745,
      "grad_norm": 7.807188034057617,
      "learning_rate": 4.734421585000652e-06,
      "loss": 0.4277,
      "mean_token_accuracy": 0.8550496265292168,
      "num_tokens": 110752633.0,
      "step": 92110
    },
    {
      "entropy": 1.9415203884243966,
      "epoch": 0.28556393995779467,
      "grad_norm": 8.066431999206543,
      "learning_rate": 4.7341646048945645e-06,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8382659062743187,
      "num_tokens": 110764376.0,
      "step": 92120
    },
    {
      "entropy": 1.801632682979107,
      "epoch": 0.2855949390828444,
      "grad_norm": 6.095812797546387,
      "learning_rate": 4.733907666629874e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8537136375904083,
      "num_tokens": 110776672.0,
      "step": 92130
    },
    {
      "entropy": 1.8623081862926483,
      "epoch": 0.28562593820789406,
      "grad_norm": 8.639974594116211,
      "learning_rate": 4.733650770195231e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.848575672507286,
      "num_tokens": 110788872.0,
      "step": 92140
    },
    {
      "entropy": 1.7804166600108147,
      "epoch": 0.2856569373329438,
      "grad_norm": 3.511837959289551,
      "learning_rate": 4.733393915579283e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8584547877311707,
      "num_tokens": 110801677.0,
      "step": 92150
    },
    {
      "entropy": 1.8777981102466583,
      "epoch": 0.28568793645799345,
      "grad_norm": 8.046910285949707,
      "learning_rate": 4.733137102770687e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8518513917922974,
      "num_tokens": 110813470.0,
      "step": 92160
    },
    {
      "entropy": 1.9628359898924828,
      "epoch": 0.2857189355830432,
      "grad_norm": 9.675116539001465,
      "learning_rate": 4.732880331758104e-06,
      "loss": 0.5357,
      "mean_token_accuracy": 0.8325561985373497,
      "num_tokens": 110824903.0,
      "step": 92170
    },
    {
      "entropy": 1.9065320461988449,
      "epoch": 0.28574993470809285,
      "grad_norm": 9.257568359375,
      "learning_rate": 4.732623602530196e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8543298453092575,
      "num_tokens": 110836341.0,
      "step": 92180
    },
    {
      "entropy": 1.8834924966096878,
      "epoch": 0.28578093383314257,
      "grad_norm": 3.791853189468384,
      "learning_rate": 4.732366915075634e-06,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8327832892537117,
      "num_tokens": 110848011.0,
      "step": 92190
    },
    {
      "entropy": 1.9523876518011094,
      "epoch": 0.28581193295819224,
      "grad_norm": 8.7767972946167,
      "learning_rate": 4.732110269383088e-06,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8447451606392861,
      "num_tokens": 110858754.0,
      "step": 92200
    },
    {
      "entropy": 1.9425658136606216,
      "epoch": 0.28584293208324196,
      "grad_norm": 6.863429069519043,
      "learning_rate": 4.731853665441238e-06,
      "loss": 0.5456,
      "mean_token_accuracy": 0.835406644642353,
      "num_tokens": 110869747.0,
      "step": 92210
    },
    {
      "entropy": 1.8428900748491288,
      "epoch": 0.28587393120829163,
      "grad_norm": 8.834907531738281,
      "learning_rate": 4.731597103238762e-06,
      "loss": 0.4663,
      "mean_token_accuracy": 0.844575323164463,
      "num_tokens": 110881672.0,
      "step": 92220
    },
    {
      "entropy": 1.9306576699018478,
      "epoch": 0.28590493033334136,
      "grad_norm": 9.883462905883789,
      "learning_rate": 4.731340582764347e-06,
      "loss": 0.498,
      "mean_token_accuracy": 0.8402498573064804,
      "num_tokens": 110893015.0,
      "step": 92230
    },
    {
      "entropy": 1.9201725766062736,
      "epoch": 0.285935929458391,
      "grad_norm": 6.858789443969727,
      "learning_rate": 4.731084104006684e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.841082276403904,
      "num_tokens": 110905179.0,
      "step": 92240
    },
    {
      "entropy": 1.8505833253264428,
      "epoch": 0.28596692858344075,
      "grad_norm": 7.575535297393799,
      "learning_rate": 4.730827666954467e-06,
      "loss": 0.4116,
      "mean_token_accuracy": 0.8532150328159332,
      "num_tokens": 110918892.0,
      "step": 92250
    },
    {
      "entropy": 1.9996703028678895,
      "epoch": 0.2859979277084904,
      "grad_norm": 9.380273818969727,
      "learning_rate": 4.730571271596393e-06,
      "loss": 0.547,
      "mean_token_accuracy": 0.8315587803721428,
      "num_tokens": 110929571.0,
      "step": 92260
    },
    {
      "entropy": 1.724966013431549,
      "epoch": 0.28602892683354014,
      "grad_norm": 8.106704711914062,
      "learning_rate": 4.730314917921165e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8658282339572907,
      "num_tokens": 110942972.0,
      "step": 92270
    },
    {
      "entropy": 1.912028570473194,
      "epoch": 0.2860599259585898,
      "grad_norm": 3.8293659687042236,
      "learning_rate": 4.730058605917492e-06,
      "loss": 0.489,
      "mean_token_accuracy": 0.847729179263115,
      "num_tokens": 110954086.0,
      "step": 92280
    },
    {
      "entropy": 1.9026774257421493,
      "epoch": 0.28609092508363954,
      "grad_norm": 3.5246219635009766,
      "learning_rate": 4.729802335574084e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8511073097586632,
      "num_tokens": 110965849.0,
      "step": 92290
    },
    {
      "entropy": 1.839532507956028,
      "epoch": 0.2861219242086892,
      "grad_norm": 7.714004039764404,
      "learning_rate": 4.729546106879656e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8525758549571038,
      "num_tokens": 110978775.0,
      "step": 92300
    },
    {
      "entropy": 1.8683712184429169,
      "epoch": 0.28615292333373893,
      "grad_norm": 4.254873752593994,
      "learning_rate": 4.729289919822929e-06,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8314826056361199,
      "num_tokens": 110990837.0,
      "step": 92310
    },
    {
      "entropy": 1.9552190572023391,
      "epoch": 0.2861839224587886,
      "grad_norm": 8.179940223693848,
      "learning_rate": 4.729033774392628e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8429278552532196,
      "num_tokens": 111002090.0,
      "step": 92320
    },
    {
      "entropy": 1.8049072623252869,
      "epoch": 0.2862149215838383,
      "grad_norm": 3.7070064544677734,
      "learning_rate": 4.728777670577479e-06,
      "loss": 0.3892,
      "mean_token_accuracy": 0.8596838563680649,
      "num_tokens": 111014690.0,
      "step": 92330
    },
    {
      "entropy": 1.865850919485092,
      "epoch": 0.286245920708888,
      "grad_norm": 9.017657279968262,
      "learning_rate": 4.7285216083662165e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.84141965508461,
      "num_tokens": 111026626.0,
      "step": 92340
    },
    {
      "entropy": 1.802014322578907,
      "epoch": 0.28627691983393766,
      "grad_norm": 4.565001010894775,
      "learning_rate": 4.728265587747578e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8524199083447457,
      "num_tokens": 111039897.0,
      "step": 92350
    },
    {
      "entropy": 1.8838436871767044,
      "epoch": 0.2863079189589874,
      "grad_norm": 10.405354499816895,
      "learning_rate": 4.728009608710304e-06,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8382696464657784,
      "num_tokens": 111052245.0,
      "step": 92360
    },
    {
      "entropy": 1.8829987928271295,
      "epoch": 0.28633891808403705,
      "grad_norm": 8.90793514251709,
      "learning_rate": 4.727753671243139e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8499641686677932,
      "num_tokens": 111063937.0,
      "step": 92370
    },
    {
      "entropy": 1.8485404312610627,
      "epoch": 0.2863699172090868,
      "grad_norm": 7.828176975250244,
      "learning_rate": 4.727497775334834e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8384749412536621,
      "num_tokens": 111076807.0,
      "step": 92380
    },
    {
      "entropy": 1.9129365399479865,
      "epoch": 0.28640091633413645,
      "grad_norm": 10.41478443145752,
      "learning_rate": 4.727241920974142e-06,
      "loss": 0.5466,
      "mean_token_accuracy": 0.8334106773138046,
      "num_tokens": 111089253.0,
      "step": 92390
    },
    {
      "entropy": 1.9112588971853257,
      "epoch": 0.28643191545918617,
      "grad_norm": 9.177438735961914,
      "learning_rate": 4.726986108149824e-06,
      "loss": 0.5223,
      "mean_token_accuracy": 0.8404966652393341,
      "num_tokens": 111100456.0,
      "step": 92400
    },
    {
      "entropy": 1.9098269432783126,
      "epoch": 0.28646291458423584,
      "grad_norm": 7.330765724182129,
      "learning_rate": 4.7267303368506395e-06,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8523991569876671,
      "num_tokens": 111111456.0,
      "step": 92410
    },
    {
      "entropy": 1.9250774174928664,
      "epoch": 0.28649391370928556,
      "grad_norm": 8.76259708404541,
      "learning_rate": 4.726474607065357e-06,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8517454132437706,
      "num_tokens": 111122408.0,
      "step": 92420
    },
    {
      "entropy": 1.8456337764859199,
      "epoch": 0.28652491283433523,
      "grad_norm": 8.982091903686523,
      "learning_rate": 4.726218918782747e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8434979751706123,
      "num_tokens": 111135014.0,
      "step": 92430
    },
    {
      "entropy": 1.8489980950951577,
      "epoch": 0.28655591195938496,
      "grad_norm": 8.749741554260254,
      "learning_rate": 4.725963271991586e-06,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8554965123534203,
      "num_tokens": 111147294.0,
      "step": 92440
    },
    {
      "entropy": 1.9318351566791534,
      "epoch": 0.2865869110844346,
      "grad_norm": 8.458632469177246,
      "learning_rate": 4.725707666680653e-06,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8387140676379203,
      "num_tokens": 111158516.0,
      "step": 92450
    },
    {
      "entropy": 1.8340228885412215,
      "epoch": 0.28661791020948435,
      "grad_norm": 3.4830410480499268,
      "learning_rate": 4.72545210283873e-06,
      "loss": 0.3714,
      "mean_token_accuracy": 0.8590243220329284,
      "num_tokens": 111171043.0,
      "step": 92460
    },
    {
      "entropy": 1.8585921421647071,
      "epoch": 0.286648909334534,
      "grad_norm": 10.120370864868164,
      "learning_rate": 4.725196580454608e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8484177976846695,
      "num_tokens": 111183642.0,
      "step": 92470
    },
    {
      "entropy": 1.924575427174568,
      "epoch": 0.28667990845958374,
      "grad_norm": 9.668998718261719,
      "learning_rate": 4.724941099517078e-06,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8406669244170188,
      "num_tokens": 111195086.0,
      "step": 92480
    },
    {
      "entropy": 1.8797391682863236,
      "epoch": 0.2867109075846334,
      "grad_norm": 7.982089996337891,
      "learning_rate": 4.724685660014936e-06,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8439215019345283,
      "num_tokens": 111206895.0,
      "step": 92490
    },
    {
      "entropy": 1.906739890575409,
      "epoch": 0.28674190670968314,
      "grad_norm": 8.40061092376709,
      "learning_rate": 4.724430261936984e-06,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8415932491421699,
      "num_tokens": 111218419.0,
      "step": 92500
    },
    {
      "entropy": 1.7755264952778815,
      "epoch": 0.2867729058347328,
      "grad_norm": 8.411188125610352,
      "learning_rate": 4.724174905272025e-06,
      "loss": 0.3824,
      "mean_token_accuracy": 0.8556288599967956,
      "num_tokens": 111232066.0,
      "step": 92510
    },
    {
      "entropy": 1.8620268389582635,
      "epoch": 0.28680390495978253,
      "grad_norm": 3.8714070320129395,
      "learning_rate": 4.72391959000887e-06,
      "loss": 0.4188,
      "mean_token_accuracy": 0.8555803820490837,
      "num_tokens": 111244082.0,
      "step": 92520
    },
    {
      "entropy": 1.925992988049984,
      "epoch": 0.2868349040848322,
      "grad_norm": 8.029500007629395,
      "learning_rate": 4.723664316136334e-06,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8394767045974731,
      "num_tokens": 111255877.0,
      "step": 92530
    },
    {
      "entropy": 1.972803682088852,
      "epoch": 0.2868659032098819,
      "grad_norm": 9.798595428466797,
      "learning_rate": 4.723409083643231e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.831108058989048,
      "num_tokens": 111266990.0,
      "step": 92540
    },
    {
      "entropy": 1.9300013601779937,
      "epoch": 0.2868969023349316,
      "grad_norm": 8.934462547302246,
      "learning_rate": 4.7231538925183875e-06,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8401208460330963,
      "num_tokens": 111277932.0,
      "step": 92550
    },
    {
      "entropy": 1.8719871819019318,
      "epoch": 0.2869279014599813,
      "grad_norm": 8.938404083251953,
      "learning_rate": 4.722898742750625e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8495993599295616,
      "num_tokens": 111290308.0,
      "step": 92560
    },
    {
      "entropy": 1.9398197084665298,
      "epoch": 0.286958900585031,
      "grad_norm": 9.505714416503906,
      "learning_rate": 4.7226436343287775e-06,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8359267324209213,
      "num_tokens": 111301797.0,
      "step": 92570
    },
    {
      "entropy": 1.8432279601693153,
      "epoch": 0.2869898997100807,
      "grad_norm": 3.82196307182312,
      "learning_rate": 4.7223885672416784e-06,
      "loss": 0.469,
      "mean_token_accuracy": 0.850963968038559,
      "num_tokens": 111314214.0,
      "step": 92580
    },
    {
      "entropy": 1.844231453537941,
      "epoch": 0.2870208988351304,
      "grad_norm": 8.75102710723877,
      "learning_rate": 4.722133541478166e-06,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8547771289944649,
      "num_tokens": 111326679.0,
      "step": 92590
    },
    {
      "entropy": 1.9384073466062546,
      "epoch": 0.28705189796018005,
      "grad_norm": 7.945824146270752,
      "learning_rate": 4.721878557027084e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.835817402601242,
      "num_tokens": 111338404.0,
      "step": 92600
    },
    {
      "entropy": 1.9293118342757225,
      "epoch": 0.28708289708522977,
      "grad_norm": 8.206449508666992,
      "learning_rate": 4.7216236138772795e-06,
      "loss": 0.4973,
      "mean_token_accuracy": 0.8453877314925193,
      "num_tokens": 111350168.0,
      "step": 92610
    },
    {
      "entropy": 1.8235325343906879,
      "epoch": 0.28711389621027944,
      "grad_norm": 3.6279773712158203,
      "learning_rate": 4.721368712017605e-06,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8457371458411217,
      "num_tokens": 111363248.0,
      "step": 92620
    },
    {
      "entropy": 1.9850926041603087,
      "epoch": 0.28714489533532916,
      "grad_norm": 9.338789939880371,
      "learning_rate": 4.721113851436916e-06,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8441228061914444,
      "num_tokens": 111374112.0,
      "step": 92630
    },
    {
      "entropy": 1.8502737820148467,
      "epoch": 0.28717589446037883,
      "grad_norm": 8.026445388793945,
      "learning_rate": 4.72085903212407e-06,
      "loss": 0.423,
      "mean_token_accuracy": 0.8572929382324219,
      "num_tokens": 111386542.0,
      "step": 92640
    },
    {
      "entropy": 1.8364259883761407,
      "epoch": 0.28720689358542856,
      "grad_norm": 7.70156717300415,
      "learning_rate": 4.7206042540679335e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8466950342059135,
      "num_tokens": 111399115.0,
      "step": 92650
    },
    {
      "entropy": 1.8671451389789582,
      "epoch": 0.2872378927104782,
      "grad_norm": 9.978074073791504,
      "learning_rate": 4.720349517257375e-06,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8522609934210778,
      "num_tokens": 111411335.0,
      "step": 92660
    },
    {
      "entropy": 1.8687150657176972,
      "epoch": 0.28726889183552795,
      "grad_norm": 4.219553470611572,
      "learning_rate": 4.720094821681266e-06,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8571690008044243,
      "num_tokens": 111423197.0,
      "step": 92670
    },
    {
      "entropy": 1.839407466351986,
      "epoch": 0.2872998909605776,
      "grad_norm": 7.242602348327637,
      "learning_rate": 4.719840167328485e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8519873261451721,
      "num_tokens": 111436358.0,
      "step": 92680
    },
    {
      "entropy": 1.8508578151464463,
      "epoch": 0.28733089008562734,
      "grad_norm": 4.504565238952637,
      "learning_rate": 4.719585554187911e-06,
      "loss": 0.455,
      "mean_token_accuracy": 0.8483609542250633,
      "num_tokens": 111448749.0,
      "step": 92690
    },
    {
      "entropy": 1.8885752364993096,
      "epoch": 0.287361889210677,
      "grad_norm": 8.104568481445312,
      "learning_rate": 4.7193309822484295e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8618633911013603,
      "num_tokens": 111460055.0,
      "step": 92700
    },
    {
      "entropy": 1.8163581773638726,
      "epoch": 0.28739288833572674,
      "grad_norm": 7.430144786834717,
      "learning_rate": 4.719076451498931e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8476424887776375,
      "num_tokens": 111473204.0,
      "step": 92710
    },
    {
      "entropy": 1.9194232299923897,
      "epoch": 0.2874238874607764,
      "grad_norm": 8.560006141662598,
      "learning_rate": 4.718821961928308e-06,
      "loss": 0.551,
      "mean_token_accuracy": 0.825740373134613,
      "num_tokens": 111484969.0,
      "step": 92720
    },
    {
      "entropy": 1.8660444170236588,
      "epoch": 0.28745488658582613,
      "grad_norm": 8.867291450500488,
      "learning_rate": 4.71856751352546e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8471433222293854,
      "num_tokens": 111497653.0,
      "step": 92730
    },
    {
      "entropy": 1.8844849050045014,
      "epoch": 0.2874858857108758,
      "grad_norm": 8.284875869750977,
      "learning_rate": 4.7183131062792855e-06,
      "loss": 0.4806,
      "mean_token_accuracy": 0.8477038741111755,
      "num_tokens": 111510030.0,
      "step": 92740
    },
    {
      "entropy": 1.945139628648758,
      "epoch": 0.2875168848359255,
      "grad_norm": 10.998967170715332,
      "learning_rate": 4.718058740178694e-06,
      "loss": 0.5958,
      "mean_token_accuracy": 0.8271935939788818,
      "num_tokens": 111521480.0,
      "step": 92750
    },
    {
      "entropy": 1.9299291223287582,
      "epoch": 0.2875478839609752,
      "grad_norm": 8.494784355163574,
      "learning_rate": 4.717804415212594e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.8403868332505227,
      "num_tokens": 111532281.0,
      "step": 92760
    },
    {
      "entropy": 1.8992675706744193,
      "epoch": 0.2875788830860249,
      "grad_norm": 3.8736727237701416,
      "learning_rate": 4.717550131369901e-06,
      "loss": 0.4972,
      "mean_token_accuracy": 0.843301497399807,
      "num_tokens": 111544377.0,
      "step": 92770
    },
    {
      "entropy": 1.9680778548121451,
      "epoch": 0.2876098822110746,
      "grad_norm": 7.67625093460083,
      "learning_rate": 4.717295888639533e-06,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8245472773909569,
      "num_tokens": 111555728.0,
      "step": 92780
    },
    {
      "entropy": 1.8295420020818711,
      "epoch": 0.2876408813361243,
      "grad_norm": 6.913825988769531,
      "learning_rate": 4.717041687010413e-06,
      "loss": 0.454,
      "mean_token_accuracy": 0.8417760416865349,
      "num_tokens": 111569369.0,
      "step": 92790
    },
    {
      "entropy": 1.8156803011894227,
      "epoch": 0.287671880461174,
      "grad_norm": 8.341958045959473,
      "learning_rate": 4.716787526471468e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.846878944337368,
      "num_tokens": 111582611.0,
      "step": 92800
    },
    {
      "entropy": 1.9245857551693917,
      "epoch": 0.2877028795862237,
      "grad_norm": 8.205559730529785,
      "learning_rate": 4.716533407011631e-06,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8327091008424758,
      "num_tokens": 111594361.0,
      "step": 92810
    },
    {
      "entropy": 1.818833366036415,
      "epoch": 0.28773387871127337,
      "grad_norm": 3.7494218349456787,
      "learning_rate": 4.716279328619835e-06,
      "loss": 0.414,
      "mean_token_accuracy": 0.8558287933468819,
      "num_tokens": 111607793.0,
      "step": 92820
    },
    {
      "entropy": 1.8438904002308845,
      "epoch": 0.2877648778363231,
      "grad_norm": 8.092823028564453,
      "learning_rate": 4.716025291285019e-06,
      "loss": 0.496,
      "mean_token_accuracy": 0.8389999002218247,
      "num_tokens": 111620382.0,
      "step": 92830
    },
    {
      "entropy": 1.8870263323187828,
      "epoch": 0.28779587696137277,
      "grad_norm": 2.7017006874084473,
      "learning_rate": 4.715771294996129e-06,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8476079568266869,
      "num_tokens": 111632680.0,
      "step": 92840
    },
    {
      "entropy": 1.841055366396904,
      "epoch": 0.28782687608642243,
      "grad_norm": 7.183150768280029,
      "learning_rate": 4.715517339742112e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8533366426825524,
      "num_tokens": 111644869.0,
      "step": 92850
    },
    {
      "entropy": 1.763474926352501,
      "epoch": 0.28785787521147216,
      "grad_norm": 3.5942344665527344,
      "learning_rate": 4.7152634255119215e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.8623644202947617,
      "num_tokens": 111658019.0,
      "step": 92860
    },
    {
      "entropy": 1.9290682673454285,
      "epoch": 0.2878888743365218,
      "grad_norm": 11.06643295288086,
      "learning_rate": 4.71500955229451e-06,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8362339481711387,
      "num_tokens": 111669612.0,
      "step": 92870
    },
    {
      "entropy": 1.9505492269992828,
      "epoch": 0.28791987346157155,
      "grad_norm": 7.899135112762451,
      "learning_rate": 4.7147557200788414e-06,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8352016389369965,
      "num_tokens": 111680635.0,
      "step": 92880
    },
    {
      "entropy": 1.8193707883358001,
      "epoch": 0.2879508725866212,
      "grad_norm": 5.0817975997924805,
      "learning_rate": 4.714501928853879e-06,
      "loss": 0.4464,
      "mean_token_accuracy": 0.849019393324852,
      "num_tokens": 111693588.0,
      "step": 92890
    },
    {
      "entropy": 1.8536582559347152,
      "epoch": 0.28798187171167094,
      "grad_norm": 3.8260529041290283,
      "learning_rate": 4.714248178608591e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8521668612957001,
      "num_tokens": 111705649.0,
      "step": 92900
    },
    {
      "entropy": 1.9044562801718712,
      "epoch": 0.2880128708367206,
      "grad_norm": 4.923748016357422,
      "learning_rate": 4.713994469331952e-06,
      "loss": 0.471,
      "mean_token_accuracy": 0.8511299833655357,
      "num_tokens": 111717116.0,
      "step": 92910
    },
    {
      "entropy": 1.8554534062743187,
      "epoch": 0.28804386996177034,
      "grad_norm": 3.183706521987915,
      "learning_rate": 4.713740801012937e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8666115805506707,
      "num_tokens": 111729411.0,
      "step": 92920
    },
    {
      "entropy": 1.8407445654273034,
      "epoch": 0.28807486908682,
      "grad_norm": 9.04386043548584,
      "learning_rate": 4.713487173640529e-06,
      "loss": 0.4396,
      "mean_token_accuracy": 0.848953865468502,
      "num_tokens": 111742147.0,
      "step": 92930
    },
    {
      "entropy": 1.8394930571317674,
      "epoch": 0.28810586821186973,
      "grad_norm": 3.4830808639526367,
      "learning_rate": 4.7132335872037114e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8509487017989159,
      "num_tokens": 111754776.0,
      "step": 92940
    },
    {
      "entropy": 1.8994437158107758,
      "epoch": 0.2881368673369194,
      "grad_norm": 9.240650177001953,
      "learning_rate": 4.712980041691476e-06,
      "loss": 0.5243,
      "mean_token_accuracy": 0.843148159980774,
      "num_tokens": 111765930.0,
      "step": 92950
    },
    {
      "entropy": 1.8866792187094688,
      "epoch": 0.2881678664619691,
      "grad_norm": 7.509947299957275,
      "learning_rate": 4.7127265370928134e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8502578973770142,
      "num_tokens": 111777482.0,
      "step": 92960
    },
    {
      "entropy": 1.8861197009682655,
      "epoch": 0.2881988655870188,
      "grad_norm": 7.608978271484375,
      "learning_rate": 4.712473073396724e-06,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8533033058047295,
      "num_tokens": 111789308.0,
      "step": 92970
    },
    {
      "entropy": 1.8014571815729141,
      "epoch": 0.2882298647120685,
      "grad_norm": 5.268444061279297,
      "learning_rate": 4.7122196505922085e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8505001902580261,
      "num_tokens": 111803076.0,
      "step": 92980
    },
    {
      "entropy": 1.9110738933086395,
      "epoch": 0.2882608638371182,
      "grad_norm": 7.389989376068115,
      "learning_rate": 4.711966268668274e-06,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8408216923475266,
      "num_tokens": 111814444.0,
      "step": 92990
    },
    {
      "entropy": 1.8892789036035538,
      "epoch": 0.2882918629621679,
      "grad_norm": 8.411653518676758,
      "learning_rate": 4.711712927613929e-06,
      "loss": 0.4976,
      "mean_token_accuracy": 0.8434664875268936,
      "num_tokens": 111826241.0,
      "step": 93000
    },
    {
      "entropy": 1.929236751794815,
      "epoch": 0.2883228620872176,
      "grad_norm": 9.464624404907227,
      "learning_rate": 4.711459627418189e-06,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8482060879468918,
      "num_tokens": 111838666.0,
      "step": 93010
    },
    {
      "entropy": 1.8663489505648614,
      "epoch": 0.2883538612122673,
      "grad_norm": 7.735058784484863,
      "learning_rate": 4.711206368070072e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8607504948973655,
      "num_tokens": 111851236.0,
      "step": 93020
    },
    {
      "entropy": 1.8760851591825485,
      "epoch": 0.288384860337317,
      "grad_norm": 7.6005754470825195,
      "learning_rate": 4.710953149558602e-06,
      "loss": 0.477,
      "mean_token_accuracy": 0.8493797600269317,
      "num_tokens": 111863060.0,
      "step": 93030
    },
    {
      "entropy": 1.8153146281838417,
      "epoch": 0.2884158594623667,
      "grad_norm": 8.263833999633789,
      "learning_rate": 4.710699971872803e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.851483790576458,
      "num_tokens": 111876458.0,
      "step": 93040
    },
    {
      "entropy": 1.8655553236603737,
      "epoch": 0.28844685858741637,
      "grad_norm": 5.136788368225098,
      "learning_rate": 4.710446835001707e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8393679440021515,
      "num_tokens": 111889068.0,
      "step": 93050
    },
    {
      "entropy": 1.8693054020404816,
      "epoch": 0.2884778577124661,
      "grad_norm": 6.886144638061523,
      "learning_rate": 4.71019373893435e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8552120968699455,
      "num_tokens": 111901836.0,
      "step": 93060
    },
    {
      "entropy": 1.8734416976571082,
      "epoch": 0.28850885683751576,
      "grad_norm": 10.7568941116333,
      "learning_rate": 4.709940683659771e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8552261263132095,
      "num_tokens": 111913647.0,
      "step": 93070
    },
    {
      "entropy": 1.868347629904747,
      "epoch": 0.2885398559625655,
      "grad_norm": 4.4855852127075195,
      "learning_rate": 4.709687669167011e-06,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8452133953571319,
      "num_tokens": 111926603.0,
      "step": 93080
    },
    {
      "entropy": 1.9437709406018258,
      "epoch": 0.28857085508761515,
      "grad_norm": 9.099634170532227,
      "learning_rate": 4.7094346954451196e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8310608744621277,
      "num_tokens": 111937928.0,
      "step": 93090
    },
    {
      "entropy": 1.8436076626181603,
      "epoch": 0.2886018542126648,
      "grad_norm": 4.898632526397705,
      "learning_rate": 4.709181762483149e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8579155907034874,
      "num_tokens": 111950393.0,
      "step": 93100
    },
    {
      "entropy": 1.8545200631022454,
      "epoch": 0.28863285333771455,
      "grad_norm": 6.226985931396484,
      "learning_rate": 4.708928870270152e-06,
      "loss": 0.4087,
      "mean_token_accuracy": 0.8614425033330917,
      "num_tokens": 111962194.0,
      "step": 93110
    },
    {
      "entropy": 1.9305305495858192,
      "epoch": 0.2886638524627642,
      "grad_norm": 8.529313087463379,
      "learning_rate": 4.70867601879519e-06,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8401212841272354,
      "num_tokens": 111973375.0,
      "step": 93120
    },
    {
      "entropy": 1.8925403907895089,
      "epoch": 0.28869485158781394,
      "grad_norm": 5.547452449798584,
      "learning_rate": 4.7084232080473254e-06,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8335786327719689,
      "num_tokens": 111985430.0,
      "step": 93130
    },
    {
      "entropy": 1.9314084231853486,
      "epoch": 0.2887258507128636,
      "grad_norm": 8.339520454406738,
      "learning_rate": 4.7081704380156275e-06,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8429270505905151,
      "num_tokens": 111996483.0,
      "step": 93140
    },
    {
      "entropy": 1.8602096036076545,
      "epoch": 0.28875684983791333,
      "grad_norm": 7.6653523445129395,
      "learning_rate": 4.7079177086891694e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8460811406373978,
      "num_tokens": 112008286.0,
      "step": 93150
    },
    {
      "entropy": 1.8911155819892884,
      "epoch": 0.288787848962963,
      "grad_norm": 8.817441940307617,
      "learning_rate": 4.7076650200570235e-06,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8515992224216461,
      "num_tokens": 112019993.0,
      "step": 93160
    },
    {
      "entropy": 1.9293202444911004,
      "epoch": 0.2888188480880127,
      "grad_norm": 7.830452919006348,
      "learning_rate": 4.707412372108274e-06,
      "loss": 0.5404,
      "mean_token_accuracy": 0.8322738707065582,
      "num_tokens": 112030697.0,
      "step": 93170
    },
    {
      "entropy": 1.9798065185546876,
      "epoch": 0.2888498472130624,
      "grad_norm": 8.459895133972168,
      "learning_rate": 4.707159764832003e-06,
      "loss": 0.5198,
      "mean_token_accuracy": 0.8400536015629768,
      "num_tokens": 112041497.0,
      "step": 93180
    },
    {
      "entropy": 1.879485437273979,
      "epoch": 0.2888808463381121,
      "grad_norm": 6.744016647338867,
      "learning_rate": 4.7069071982172985e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8491860061883927,
      "num_tokens": 112054220.0,
      "step": 93190
    },
    {
      "entropy": 1.9080306202173234,
      "epoch": 0.2889118454631618,
      "grad_norm": 7.247501373291016,
      "learning_rate": 4.706654672253255e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8472570210695267,
      "num_tokens": 112066105.0,
      "step": 93200
    },
    {
      "entropy": 1.8720626473426818,
      "epoch": 0.2889428445882115,
      "grad_norm": 7.93571138381958,
      "learning_rate": 4.706402186928967e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.8489278420805931,
      "num_tokens": 112078575.0,
      "step": 93210
    },
    {
      "entropy": 1.8863236621022224,
      "epoch": 0.2889738437132612,
      "grad_norm": 8.153715133666992,
      "learning_rate": 4.706149742233537e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8377825424075127,
      "num_tokens": 112090038.0,
      "step": 93220
    },
    {
      "entropy": 1.8599166497588158,
      "epoch": 0.2890048428383109,
      "grad_norm": 8.849479675292969,
      "learning_rate": 4.705897338156069e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8481295198202133,
      "num_tokens": 112101547.0,
      "step": 93230
    },
    {
      "entropy": 1.7532717436552048,
      "epoch": 0.2890358419633606,
      "grad_norm": 8.861629486083984,
      "learning_rate": 4.705644974685672e-06,
      "loss": 0.3496,
      "mean_token_accuracy": 0.8671473681926727,
      "num_tokens": 112114998.0,
      "step": 93240
    },
    {
      "entropy": 1.8831738129258155,
      "epoch": 0.2890668410884103,
      "grad_norm": 7.922811031341553,
      "learning_rate": 4.705392651811459e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8377567693591118,
      "num_tokens": 112126554.0,
      "step": 93250
    },
    {
      "entropy": 1.8818488538265228,
      "epoch": 0.28909784021345997,
      "grad_norm": 8.488920211791992,
      "learning_rate": 4.705140369522546e-06,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8422826811671257,
      "num_tokens": 112138514.0,
      "step": 93260
    },
    {
      "entropy": 1.9045387908816338,
      "epoch": 0.2891288393385097,
      "grad_norm": 7.4267425537109375,
      "learning_rate": 4.704888127808055e-06,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8422675371170044,
      "num_tokens": 112149688.0,
      "step": 93270
    },
    {
      "entropy": 1.9745795860886575,
      "epoch": 0.28915983846355936,
      "grad_norm": 7.358890056610107,
      "learning_rate": 4.704635926657112e-06,
      "loss": 0.5901,
      "mean_token_accuracy": 0.8308152720332146,
      "num_tokens": 112161678.0,
      "step": 93280
    },
    {
      "entropy": 1.8890698000788688,
      "epoch": 0.2891908375886091,
      "grad_norm": 7.141902446746826,
      "learning_rate": 4.704383766058845e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8586556881666183,
      "num_tokens": 112173914.0,
      "step": 93290
    },
    {
      "entropy": 1.9566715627908706,
      "epoch": 0.28922183671365875,
      "grad_norm": 8.638976097106934,
      "learning_rate": 4.70413164600239e-06,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8295210391283036,
      "num_tokens": 112185080.0,
      "step": 93300
    },
    {
      "entropy": 1.8873969689011574,
      "epoch": 0.2892528358387085,
      "grad_norm": 3.9027388095855713,
      "learning_rate": 4.70387956647688e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8503093495965004,
      "num_tokens": 112197377.0,
      "step": 93310
    },
    {
      "entropy": 1.8474943891167641,
      "epoch": 0.28928383496375815,
      "grad_norm": 4.076306343078613,
      "learning_rate": 4.703627527471461e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8605047658085823,
      "num_tokens": 112209297.0,
      "step": 93320
    },
    {
      "entropy": 1.9313654646277427,
      "epoch": 0.28931483408880787,
      "grad_norm": 11.894732475280762,
      "learning_rate": 4.703375528975276e-06,
      "loss": 0.4935,
      "mean_token_accuracy": 0.8404849201440812,
      "num_tokens": 112220394.0,
      "step": 93330
    },
    {
      "entropy": 1.9135547280311584,
      "epoch": 0.28934583321385754,
      "grad_norm": 7.225008487701416,
      "learning_rate": 4.703123570977474e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8464812904596328,
      "num_tokens": 112231383.0,
      "step": 93340
    },
    {
      "entropy": 1.9131829939782619,
      "epoch": 0.2893768323389072,
      "grad_norm": 10.179394721984863,
      "learning_rate": 4.702871653467211e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8461228340864182,
      "num_tokens": 112243480.0,
      "step": 93350
    },
    {
      "entropy": 1.9151297047734261,
      "epoch": 0.28940783146395693,
      "grad_norm": 7.998013019561768,
      "learning_rate": 4.702619776433645e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8424739927053452,
      "num_tokens": 112254846.0,
      "step": 93360
    },
    {
      "entropy": 1.8562763080000877,
      "epoch": 0.2894388305890066,
      "grad_norm": 3.337523937225342,
      "learning_rate": 4.702367939865935e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8496193781495094,
      "num_tokens": 112266394.0,
      "step": 93370
    },
    {
      "entropy": 1.9171319857239724,
      "epoch": 0.2894698297140563,
      "grad_norm": 8.044168472290039,
      "learning_rate": 4.70211614375325e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8436752587556839,
      "num_tokens": 112277740.0,
      "step": 93380
    },
    {
      "entropy": 1.8200967207551002,
      "epoch": 0.289500828839106,
      "grad_norm": 7.529294013977051,
      "learning_rate": 4.701864388084757e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8523975953459739,
      "num_tokens": 112290792.0,
      "step": 93390
    },
    {
      "entropy": 1.8970361724495888,
      "epoch": 0.2895318279641557,
      "grad_norm": 10.518641471862793,
      "learning_rate": 4.701612672849634e-06,
      "loss": 0.5585,
      "mean_token_accuracy": 0.8354102879762649,
      "num_tokens": 112302522.0,
      "step": 93400
    },
    {
      "entropy": 1.827714842557907,
      "epoch": 0.2895628270892054,
      "grad_norm": 8.352113723754883,
      "learning_rate": 4.701360998037056e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.849641677737236,
      "num_tokens": 112315336.0,
      "step": 93410
    },
    {
      "entropy": 1.9102312728762627,
      "epoch": 0.2895938262142551,
      "grad_norm": 7.807504653930664,
      "learning_rate": 4.701109363636205e-06,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8295806169509887,
      "num_tokens": 112326917.0,
      "step": 93420
    },
    {
      "entropy": 1.850624245405197,
      "epoch": 0.2896248253393048,
      "grad_norm": 4.632590293884277,
      "learning_rate": 4.70085776963627e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8438147455453873,
      "num_tokens": 112339561.0,
      "step": 93430
    },
    {
      "entropy": 1.909179501235485,
      "epoch": 0.2896558244643545,
      "grad_norm": 8.325363159179688,
      "learning_rate": 4.700606216026438e-06,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8498807772994041,
      "num_tokens": 112350456.0,
      "step": 93440
    },
    {
      "entropy": 1.8641536325216292,
      "epoch": 0.2896868235894042,
      "grad_norm": 10.412588119506836,
      "learning_rate": 4.700354702795905e-06,
      "loss": 0.4899,
      "mean_token_accuracy": 0.8456828027963639,
      "num_tokens": 112363019.0,
      "step": 93450
    },
    {
      "entropy": 1.92139028608799,
      "epoch": 0.2897178227144539,
      "grad_norm": 10.26781940460205,
      "learning_rate": 4.700103229933871e-06,
      "loss": 0.5355,
      "mean_token_accuracy": 0.8238134995102883,
      "num_tokens": 112374578.0,
      "step": 93460
    },
    {
      "entropy": 1.9097017407417298,
      "epoch": 0.28974882183950357,
      "grad_norm": 6.589956283569336,
      "learning_rate": 4.699851797429535e-06,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8451741203665734,
      "num_tokens": 112385717.0,
      "step": 93470
    },
    {
      "entropy": 1.9064988225698472,
      "epoch": 0.2897798209645533,
      "grad_norm": 7.078739643096924,
      "learning_rate": 4.6996004052721055e-06,
      "loss": 0.5205,
      "mean_token_accuracy": 0.832171306014061,
      "num_tokens": 112397934.0,
      "step": 93480
    },
    {
      "entropy": 1.8057002767920494,
      "epoch": 0.28981082008960296,
      "grad_norm": 8.572505950927734,
      "learning_rate": 4.699349053450793e-06,
      "loss": 0.3998,
      "mean_token_accuracy": 0.858871404826641,
      "num_tokens": 112411952.0,
      "step": 93490
    },
    {
      "entropy": 1.9524729505181313,
      "epoch": 0.2898418192146527,
      "grad_norm": 8.818181037902832,
      "learning_rate": 4.699097741954811e-06,
      "loss": 0.5257,
      "mean_token_accuracy": 0.8438670292496682,
      "num_tokens": 112423702.0,
      "step": 93500
    },
    {
      "entropy": 1.9189194455742835,
      "epoch": 0.28987281833970235,
      "grad_norm": 3.635896682739258,
      "learning_rate": 4.698846470773379e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8519726365804672,
      "num_tokens": 112435673.0,
      "step": 93510
    },
    {
      "entropy": 1.9499738737940788,
      "epoch": 0.2899038174647521,
      "grad_norm": 8.249286651611328,
      "learning_rate": 4.698595239895718e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.844916258752346,
      "num_tokens": 112446773.0,
      "step": 93520
    },
    {
      "entropy": 1.948981523513794,
      "epoch": 0.28993481658980175,
      "grad_norm": 8.473828315734863,
      "learning_rate": 4.698344049311058e-06,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8404761984944343,
      "num_tokens": 112457606.0,
      "step": 93530
    },
    {
      "entropy": 1.8598048985004425,
      "epoch": 0.28996581571485147,
      "grad_norm": 7.43870735168457,
      "learning_rate": 4.698092899008628e-06,
      "loss": 0.475,
      "mean_token_accuracy": 0.8440364077687263,
      "num_tokens": 112470335.0,
      "step": 93540
    },
    {
      "entropy": 1.871220625936985,
      "epoch": 0.28999681483990114,
      "grad_norm": 7.426569938659668,
      "learning_rate": 4.69784178897766e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8520672276616097,
      "num_tokens": 112482248.0,
      "step": 93550
    },
    {
      "entropy": 1.86535192579031,
      "epoch": 0.29002781396495086,
      "grad_norm": 4.510183811187744,
      "learning_rate": 4.697590719207397e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8448464289307595,
      "num_tokens": 112494530.0,
      "step": 93560
    },
    {
      "entropy": 1.8161170035600662,
      "epoch": 0.29005881309000053,
      "grad_norm": 8.694151878356934,
      "learning_rate": 4.69733968968708e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8455665096640587,
      "num_tokens": 112508367.0,
      "step": 93570
    },
    {
      "entropy": 1.9283171832561492,
      "epoch": 0.29008981221505026,
      "grad_norm": 8.558956146240234,
      "learning_rate": 4.697088700405954e-06,
      "loss": 0.567,
      "mean_token_accuracy": 0.8320746123790741,
      "num_tokens": 112519148.0,
      "step": 93580
    },
    {
      "entropy": 1.9475929975509643,
      "epoch": 0.2901208113400999,
      "grad_norm": 7.119383811950684,
      "learning_rate": 4.696837751353273e-06,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8410884723067283,
      "num_tokens": 112529839.0,
      "step": 93590
    },
    {
      "entropy": 1.8843889951705932,
      "epoch": 0.2901518104651496,
      "grad_norm": 6.53165340423584,
      "learning_rate": 4.69658684251829e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8552757307887078,
      "num_tokens": 112541392.0,
      "step": 93600
    },
    {
      "entropy": 1.9115852415561676,
      "epoch": 0.2901828095901993,
      "grad_norm": 8.958046913146973,
      "learning_rate": 4.696335973890263e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8533832281827927,
      "num_tokens": 112551686.0,
      "step": 93610
    },
    {
      "entropy": 1.9201652556657791,
      "epoch": 0.290213808715249,
      "grad_norm": 9.570830345153809,
      "learning_rate": 4.696085145458457e-06,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8437472805380821,
      "num_tokens": 112562743.0,
      "step": 93620
    },
    {
      "entropy": 1.909856851398945,
      "epoch": 0.2902448078402987,
      "grad_norm": 3.7993521690368652,
      "learning_rate": 4.695834357212138e-06,
      "loss": 0.5146,
      "mean_token_accuracy": 0.8377780944108963,
      "num_tokens": 112574467.0,
      "step": 93630
    },
    {
      "entropy": 1.8987789198756218,
      "epoch": 0.2902758069653484,
      "grad_norm": 10.340612411499023,
      "learning_rate": 4.695583609140576e-06,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8381225094199181,
      "num_tokens": 112587453.0,
      "step": 93640
    },
    {
      "entropy": 1.767358809709549,
      "epoch": 0.2903068060903981,
      "grad_norm": 4.561580181121826,
      "learning_rate": 4.695332901233046e-06,
      "loss": 0.4904,
      "mean_token_accuracy": 0.845879316329956,
      "num_tokens": 112601795.0,
      "step": 93650
    },
    {
      "entropy": 1.8162215173244476,
      "epoch": 0.2903378052154478,
      "grad_norm": 6.31105899810791,
      "learning_rate": 4.695082233478828e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.8538815006613731,
      "num_tokens": 112614333.0,
      "step": 93660
    },
    {
      "entropy": 1.865521389245987,
      "epoch": 0.2903688043404975,
      "grad_norm": 8.672292709350586,
      "learning_rate": 4.694831605867206e-06,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8480850234627724,
      "num_tokens": 112627252.0,
      "step": 93670
    },
    {
      "entropy": 1.7771051108837128,
      "epoch": 0.29039980346554717,
      "grad_norm": 9.864136695861816,
      "learning_rate": 4.694581018387463e-06,
      "loss": 0.3924,
      "mean_token_accuracy": 0.8620004311203957,
      "num_tokens": 112640242.0,
      "step": 93680
    },
    {
      "entropy": 1.8782151356339454,
      "epoch": 0.2904308025905969,
      "grad_norm": 6.314033031463623,
      "learning_rate": 4.694330471028893e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8508936673402786,
      "num_tokens": 112652406.0,
      "step": 93690
    },
    {
      "entropy": 1.9099466919898986,
      "epoch": 0.29046180171564656,
      "grad_norm": 4.040074348449707,
      "learning_rate": 4.694079963780791e-06,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8430052191019058,
      "num_tokens": 112664170.0,
      "step": 93700
    },
    {
      "entropy": 1.863590781390667,
      "epoch": 0.2904928008406963,
      "grad_norm": 8.13239860534668,
      "learning_rate": 4.693829496632454e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8332739755511284,
      "num_tokens": 112676541.0,
      "step": 93710
    },
    {
      "entropy": 1.8954721093177795,
      "epoch": 0.29052379996574595,
      "grad_norm": 9.620070457458496,
      "learning_rate": 4.693579069573186e-06,
      "loss": 0.5071,
      "mean_token_accuracy": 0.8386843577027321,
      "num_tokens": 112689006.0,
      "step": 93720
    },
    {
      "entropy": 1.8437098309397697,
      "epoch": 0.2905547990907957,
      "grad_norm": 8.268927574157715,
      "learning_rate": 4.693328682592294e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.860004435479641,
      "num_tokens": 112701042.0,
      "step": 93730
    },
    {
      "entropy": 1.8573900401592254,
      "epoch": 0.29058579821584535,
      "grad_norm": 3.780688524246216,
      "learning_rate": 4.693078335679089e-06,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8458506971597671,
      "num_tokens": 112713236.0,
      "step": 93740
    },
    {
      "entropy": 1.81138436794281,
      "epoch": 0.29061679734089507,
      "grad_norm": 9.589875221252441,
      "learning_rate": 4.692828028822885e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8497409135103225,
      "num_tokens": 112725893.0,
      "step": 93750
    },
    {
      "entropy": 1.835260456800461,
      "epoch": 0.29064779646594474,
      "grad_norm": 9.189959526062012,
      "learning_rate": 4.692577762013002e-06,
      "loss": 0.5272,
      "mean_token_accuracy": 0.8347838670015335,
      "num_tokens": 112738315.0,
      "step": 93760
    },
    {
      "entropy": 1.8794139876961709,
      "epoch": 0.29067879559099447,
      "grad_norm": 3.8947536945343018,
      "learning_rate": 4.692327535238763e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.851905120909214,
      "num_tokens": 112750430.0,
      "step": 93770
    },
    {
      "entropy": 1.7996224954724311,
      "epoch": 0.29070979471604413,
      "grad_norm": 9.7765474319458,
      "learning_rate": 4.6920773484894935e-06,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8449171632528305,
      "num_tokens": 112763272.0,
      "step": 93780
    },
    {
      "entropy": 1.8897378101944924,
      "epoch": 0.29074079384109386,
      "grad_norm": 9.47298526763916,
      "learning_rate": 4.6918272017545255e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8373202964663505,
      "num_tokens": 112775724.0,
      "step": 93790
    },
    {
      "entropy": 1.9278543174266816,
      "epoch": 0.2907717929661435,
      "grad_norm": 7.359293460845947,
      "learning_rate": 4.691577095023192e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8393516659736633,
      "num_tokens": 112787066.0,
      "step": 93800
    },
    {
      "entropy": 1.9417521178722381,
      "epoch": 0.29080279209119325,
      "grad_norm": 8.523365020751953,
      "learning_rate": 4.691327028284835e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.835085253417492,
      "num_tokens": 112798938.0,
      "step": 93810
    },
    {
      "entropy": 1.830255390703678,
      "epoch": 0.2908337912162429,
      "grad_norm": 8.126928329467773,
      "learning_rate": 4.691077001528794e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8412661746144294,
      "num_tokens": 112812321.0,
      "step": 93820
    },
    {
      "entropy": 1.8606264024972916,
      "epoch": 0.2908647903412926,
      "grad_norm": 10.298184394836426,
      "learning_rate": 4.690827014744417e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.840786500275135,
      "num_tokens": 112824184.0,
      "step": 93830
    },
    {
      "entropy": 1.9569880202412606,
      "epoch": 0.2908957894663423,
      "grad_norm": 10.95293140411377,
      "learning_rate": 4.690577067921055e-06,
      "loss": 0.5302,
      "mean_token_accuracy": 0.8427381262183189,
      "num_tokens": 112834958.0,
      "step": 93840
    },
    {
      "entropy": 1.879898366332054,
      "epoch": 0.290926788591392,
      "grad_norm": 8.868535041809082,
      "learning_rate": 4.690327161048064e-06,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8543467596173286,
      "num_tokens": 112847253.0,
      "step": 93850
    },
    {
      "entropy": 1.884432803094387,
      "epoch": 0.2909577877164417,
      "grad_norm": 3.7411601543426514,
      "learning_rate": 4.6900772941147994e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8459440425038338,
      "num_tokens": 112859034.0,
      "step": 93860
    },
    {
      "entropy": 1.8857886865735054,
      "epoch": 0.2909887868414914,
      "grad_norm": 4.058228492736816,
      "learning_rate": 4.689827467110626e-06,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8447174549102783,
      "num_tokens": 112870746.0,
      "step": 93870
    },
    {
      "entropy": 1.8584589600563048,
      "epoch": 0.2910197859665411,
      "grad_norm": 3.87026309967041,
      "learning_rate": 4.689577680024911e-06,
      "loss": 0.4473,
      "mean_token_accuracy": 0.846294179558754,
      "num_tokens": 112883210.0,
      "step": 93880
    },
    {
      "entropy": 1.867951761186123,
      "epoch": 0.29105078509159077,
      "grad_norm": 8.24725341796875,
      "learning_rate": 4.689327932847024e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8442766860127449,
      "num_tokens": 112895841.0,
      "step": 93890
    },
    {
      "entropy": 1.8113203912973403,
      "epoch": 0.2910817842166405,
      "grad_norm": 3.6465940475463867,
      "learning_rate": 4.689078225566338e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.8473143368959427,
      "num_tokens": 112909141.0,
      "step": 93900
    },
    {
      "entropy": 1.929970356822014,
      "epoch": 0.29111278334169016,
      "grad_norm": 8.883091926574707,
      "learning_rate": 4.688828558172234e-06,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8561759814620018,
      "num_tokens": 112919787.0,
      "step": 93910
    },
    {
      "entropy": 1.9370476379990578,
      "epoch": 0.2911437824667399,
      "grad_norm": 7.68314790725708,
      "learning_rate": 4.688578930654094e-06,
      "loss": 0.5672,
      "mean_token_accuracy": 0.8174238324165344,
      "num_tokens": 112931094.0,
      "step": 93920
    },
    {
      "entropy": 1.8187593132257462,
      "epoch": 0.29117478159178956,
      "grad_norm": 9.00848388671875,
      "learning_rate": 4.688329343001302e-06,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8468037515878677,
      "num_tokens": 112944183.0,
      "step": 93930
    },
    {
      "entropy": 1.7754915788769723,
      "epoch": 0.2912057807168393,
      "grad_norm": 8.143677711486816,
      "learning_rate": 4.688079795203251e-06,
      "loss": 0.3993,
      "mean_token_accuracy": 0.8577578738331795,
      "num_tokens": 112957991.0,
      "step": 93940
    },
    {
      "entropy": 1.8825334414839745,
      "epoch": 0.29123677984188895,
      "grad_norm": 4.188181400299072,
      "learning_rate": 4.687830287249335e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8422155871987342,
      "num_tokens": 112970170.0,
      "step": 93950
    },
    {
      "entropy": 1.8940382510423661,
      "epoch": 0.2912677789669387,
      "grad_norm": 9.72235107421875,
      "learning_rate": 4.68758081912895e-06,
      "loss": 0.5222,
      "mean_token_accuracy": 0.8427134841680527,
      "num_tokens": 112981192.0,
      "step": 93960
    },
    {
      "entropy": 1.8888707652688026,
      "epoch": 0.29129877809198834,
      "grad_norm": 8.664020538330078,
      "learning_rate": 4.6873313908315015e-06,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8318000584840775,
      "num_tokens": 112992870.0,
      "step": 93970
    },
    {
      "entropy": 1.850681571662426,
      "epoch": 0.29132977721703807,
      "grad_norm": 4.301815986633301,
      "learning_rate": 4.687082002346394e-06,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8496690198779107,
      "num_tokens": 113005123.0,
      "step": 93980
    },
    {
      "entropy": 1.7844605684280395,
      "epoch": 0.29136077634208774,
      "grad_norm": 8.043193817138672,
      "learning_rate": 4.686832653663037e-06,
      "loss": 0.4085,
      "mean_token_accuracy": 0.8557795941829681,
      "num_tokens": 113017593.0,
      "step": 93990
    },
    {
      "entropy": 1.7445304661989212,
      "epoch": 0.29139177546713746,
      "grad_norm": 3.7523953914642334,
      "learning_rate": 4.686583344770846e-06,
      "loss": 0.4032,
      "mean_token_accuracy": 0.8624411836266518,
      "num_tokens": 113031950.0,
      "step": 94000
    },
    {
      "entropy": 1.7966215670108796,
      "epoch": 0.29142277459218713,
      "grad_norm": 7.984848499298096,
      "learning_rate": 4.686334075659238e-06,
      "loss": 0.3961,
      "mean_token_accuracy": 0.8521535471081734,
      "num_tokens": 113045309.0,
      "step": 94010
    },
    {
      "entropy": 1.8039371758699416,
      "epoch": 0.29145377371723685,
      "grad_norm": 4.514188766479492,
      "learning_rate": 4.686084846317634e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8542758151888847,
      "num_tokens": 113058068.0,
      "step": 94020
    },
    {
      "entropy": 1.8654557079076768,
      "epoch": 0.2914847728422865,
      "grad_norm": 8.273473739624023,
      "learning_rate": 4.685835656735462e-06,
      "loss": 0.508,
      "mean_token_accuracy": 0.8452908799052239,
      "num_tokens": 113069283.0,
      "step": 94030
    },
    {
      "entropy": 1.8588725805282593,
      "epoch": 0.29151577196733625,
      "grad_norm": 4.237758636474609,
      "learning_rate": 4.685586506902148e-06,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8419048935174942,
      "num_tokens": 113081319.0,
      "step": 94040
    },
    {
      "entropy": 1.8301612615585328,
      "epoch": 0.2915467710923859,
      "grad_norm": 8.443975448608398,
      "learning_rate": 4.685337396807132e-06,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8360993027687073,
      "num_tokens": 113093388.0,
      "step": 94050
    },
    {
      "entropy": 1.8223912194371223,
      "epoch": 0.29157777021743564,
      "grad_norm": 8.578900337219238,
      "learning_rate": 4.685088326439846e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.856349365413189,
      "num_tokens": 113106077.0,
      "step": 94060
    },
    {
      "entropy": 1.8360855415463448,
      "epoch": 0.2916087693424853,
      "grad_norm": 8.660734176635742,
      "learning_rate": 4.684839295789734e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.848674775660038,
      "num_tokens": 113118493.0,
      "step": 94070
    },
    {
      "entropy": 1.8793357729911804,
      "epoch": 0.291639768467535,
      "grad_norm": 6.9419403076171875,
      "learning_rate": 4.684590304846241e-06,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8568961501121521,
      "num_tokens": 113129937.0,
      "step": 94080
    },
    {
      "entropy": 1.909253677725792,
      "epoch": 0.2916707675925847,
      "grad_norm": 8.138631820678711,
      "learning_rate": 4.684341353598818e-06,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8358252078294754,
      "num_tokens": 113140848.0,
      "step": 94090
    },
    {
      "entropy": 1.8075340047478676,
      "epoch": 0.29170176671763437,
      "grad_norm": 8.938015937805176,
      "learning_rate": 4.684092442036915e-06,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8573233857750893,
      "num_tokens": 113153007.0,
      "step": 94100
    },
    {
      "entropy": 1.8430316671729088,
      "epoch": 0.2917327658426841,
      "grad_norm": 7.955845355987549,
      "learning_rate": 4.683843570149992e-06,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8404422610998153,
      "num_tokens": 113165004.0,
      "step": 94110
    },
    {
      "entropy": 1.8731530100107192,
      "epoch": 0.29176376496773376,
      "grad_norm": 9.946492195129395,
      "learning_rate": 4.683594737927509e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8477850437164307,
      "num_tokens": 113176838.0,
      "step": 94120
    },
    {
      "entropy": 1.8446904599666596,
      "epoch": 0.2917947640927835,
      "grad_norm": 4.911070823669434,
      "learning_rate": 4.683345945358933e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8407913163304329,
      "num_tokens": 113189520.0,
      "step": 94130
    },
    {
      "entropy": 1.8978979125618936,
      "epoch": 0.29182576321783316,
      "grad_norm": 8.811211585998535,
      "learning_rate": 4.683097192433731e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8498449578881264,
      "num_tokens": 113201050.0,
      "step": 94140
    },
    {
      "entropy": 1.868384449183941,
      "epoch": 0.2918567623428829,
      "grad_norm": 8.619186401367188,
      "learning_rate": 4.682848479141376e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8416146129369736,
      "num_tokens": 113213345.0,
      "step": 94150
    },
    {
      "entropy": 1.8393750131130218,
      "epoch": 0.29188776146793255,
      "grad_norm": 2.398193836212158,
      "learning_rate": 4.682599805471346e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8392178237438201,
      "num_tokens": 113226489.0,
      "step": 94160
    },
    {
      "entropy": 1.9002157673239708,
      "epoch": 0.2919187605929823,
      "grad_norm": 9.483999252319336,
      "learning_rate": 4.68235117141312e-06,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8522425264120101,
      "num_tokens": 113238608.0,
      "step": 94170
    },
    {
      "entropy": 1.8459337919950485,
      "epoch": 0.29194975971803194,
      "grad_norm": 7.897007942199707,
      "learning_rate": 4.6821025769561855e-06,
      "loss": 0.4421,
      "mean_token_accuracy": 0.8493504419922828,
      "num_tokens": 113250584.0,
      "step": 94180
    },
    {
      "entropy": 1.9223339349031447,
      "epoch": 0.29198075884308167,
      "grad_norm": 8.27923583984375,
      "learning_rate": 4.681854022090028e-06,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8487838566303253,
      "num_tokens": 113261767.0,
      "step": 94190
    },
    {
      "entropy": 1.9111609309911728,
      "epoch": 0.29201175796813134,
      "grad_norm": 8.224406242370605,
      "learning_rate": 4.681605506804143e-06,
      "loss": 0.498,
      "mean_token_accuracy": 0.839901152253151,
      "num_tokens": 113273424.0,
      "step": 94200
    },
    {
      "entropy": 1.8766125679016112,
      "epoch": 0.29204275709318106,
      "grad_norm": 8.163731575012207,
      "learning_rate": 4.681357031088025e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8469099998474121,
      "num_tokens": 113285663.0,
      "step": 94210
    },
    {
      "entropy": 1.815425206720829,
      "epoch": 0.29207375621823073,
      "grad_norm": 8.935981750488281,
      "learning_rate": 4.681108594931173e-06,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8499750107526779,
      "num_tokens": 113298461.0,
      "step": 94220
    },
    {
      "entropy": 1.6602164059877396,
      "epoch": 0.29210475534328045,
      "grad_norm": 8.308954238891602,
      "learning_rate": 4.680860198323094e-06,
      "loss": 0.383,
      "mean_token_accuracy": 0.8591342657804489,
      "num_tokens": 113314031.0,
      "step": 94230
    },
    {
      "entropy": 1.838196623325348,
      "epoch": 0.2921357544683301,
      "grad_norm": 3.9183874130249023,
      "learning_rate": 4.6806118412532965e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8539181277155876,
      "num_tokens": 113326110.0,
      "step": 94240
    },
    {
      "entropy": 1.8387087047100068,
      "epoch": 0.29216675359337985,
      "grad_norm": 7.87742280960083,
      "learning_rate": 4.680363523711289e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.8537355214357376,
      "num_tokens": 113338311.0,
      "step": 94250
    },
    {
      "entropy": 1.8915387392044067,
      "epoch": 0.2921977527184295,
      "grad_norm": 8.182762145996094,
      "learning_rate": 4.680115245686591e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8396169424057007,
      "num_tokens": 113350148.0,
      "step": 94260
    },
    {
      "entropy": 1.8612817063927651,
      "epoch": 0.29222875184347924,
      "grad_norm": 6.411753177642822,
      "learning_rate": 4.679867007168719e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8560265928506852,
      "num_tokens": 113362183.0,
      "step": 94270
    },
    {
      "entropy": 1.8689087167382241,
      "epoch": 0.2922597509685289,
      "grad_norm": 8.272517204284668,
      "learning_rate": 4.6796188081472e-06,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8516158699989319,
      "num_tokens": 113374575.0,
      "step": 94280
    },
    {
      "entropy": 1.9062586069107055,
      "epoch": 0.29229075009357863,
      "grad_norm": 10.613982200622559,
      "learning_rate": 4.679370648611559e-06,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8422037899494171,
      "num_tokens": 113386126.0,
      "step": 94290
    },
    {
      "entropy": 1.9051846981048584,
      "epoch": 0.2923217492186283,
      "grad_norm": 8.695990562438965,
      "learning_rate": 4.679122528551329e-06,
      "loss": 0.5411,
      "mean_token_accuracy": 0.841190955042839,
      "num_tokens": 113397175.0,
      "step": 94300
    },
    {
      "entropy": 1.7903152495622634,
      "epoch": 0.292352748343678,
      "grad_norm": 4.608640670776367,
      "learning_rate": 4.678874447956044e-06,
      "loss": 0.391,
      "mean_token_accuracy": 0.8582356154918671,
      "num_tokens": 113410086.0,
      "step": 94310
    },
    {
      "entropy": 1.8944632783532143,
      "epoch": 0.2923837474687277,
      "grad_norm": 8.895879745483398,
      "learning_rate": 4.6786264068152445e-06,
      "loss": 0.47,
      "mean_token_accuracy": 0.8449664607644081,
      "num_tokens": 113422204.0,
      "step": 94320
    },
    {
      "entropy": 1.9663603961467744,
      "epoch": 0.29241474659377736,
      "grad_norm": 10.93663215637207,
      "learning_rate": 4.678378405118473e-06,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8348477125167847,
      "num_tokens": 113432962.0,
      "step": 94330
    },
    {
      "entropy": 1.8723286166787148,
      "epoch": 0.2924457457188271,
      "grad_norm": 8.900105476379395,
      "learning_rate": 4.6781304428552765e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8530125021934509,
      "num_tokens": 113444695.0,
      "step": 94340
    },
    {
      "entropy": 1.7696161583065986,
      "epoch": 0.29247674484387676,
      "grad_norm": 7.942965030670166,
      "learning_rate": 4.677882520015207e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.857183714210987,
      "num_tokens": 113458974.0,
      "step": 94350
    },
    {
      "entropy": 1.8979527205228806,
      "epoch": 0.2925077439689265,
      "grad_norm": 10.76806354522705,
      "learning_rate": 4.677634636587817e-06,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8429912239313125,
      "num_tokens": 113470746.0,
      "step": 94360
    },
    {
      "entropy": 1.8394412383437158,
      "epoch": 0.29253874309397615,
      "grad_norm": 7.812624931335449,
      "learning_rate": 4.677386792562667e-06,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8541227161884308,
      "num_tokens": 113482575.0,
      "step": 94370
    },
    {
      "entropy": 1.8561811774969101,
      "epoch": 0.2925697422190259,
      "grad_norm": 7.494784832000732,
      "learning_rate": 4.6771389879293185e-06,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8553539365530014,
      "num_tokens": 113493656.0,
      "step": 94380
    },
    {
      "entropy": 1.877823382616043,
      "epoch": 0.29260074134407554,
      "grad_norm": 7.919898986816406,
      "learning_rate": 4.676891222677338e-06,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8565265223383903,
      "num_tokens": 113504994.0,
      "step": 94390
    },
    {
      "entropy": 1.8452934324741364,
      "epoch": 0.29263174046912527,
      "grad_norm": 9.235838890075684,
      "learning_rate": 4.6766434967962945e-06,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8403481170535088,
      "num_tokens": 113517235.0,
      "step": 94400
    },
    {
      "entropy": 1.8997774094343185,
      "epoch": 0.29266273959417494,
      "grad_norm": 8.089444160461426,
      "learning_rate": 4.6763958102757665e-06,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8421216562390328,
      "num_tokens": 113528646.0,
      "step": 94410
    },
    {
      "entropy": 1.6933333344757557,
      "epoch": 0.29269373871922466,
      "grad_norm": 9.213237762451172,
      "learning_rate": 4.676148163105327e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.8442223116755485,
      "num_tokens": 113543849.0,
      "step": 94420
    },
    {
      "entropy": 1.74602971971035,
      "epoch": 0.29272473784427433,
      "grad_norm": 8.171969413757324,
      "learning_rate": 4.67590055527456e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.8524969726800918,
      "num_tokens": 113556535.0,
      "step": 94430
    },
    {
      "entropy": 1.899616888165474,
      "epoch": 0.29275573696932405,
      "grad_norm": 9.324505805969238,
      "learning_rate": 4.675652986773051e-06,
      "loss": 0.5368,
      "mean_token_accuracy": 0.8389844998717308,
      "num_tokens": 113567827.0,
      "step": 94440
    },
    {
      "entropy": 1.85470499843359,
      "epoch": 0.2927867360943737,
      "grad_norm": 7.886104106903076,
      "learning_rate": 4.675405457590389e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8475810378789902,
      "num_tokens": 113578947.0,
      "step": 94450
    },
    {
      "entropy": 1.9117970570921898,
      "epoch": 0.29281773521942345,
      "grad_norm": 8.605646133422852,
      "learning_rate": 4.675157967716168e-06,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8469821259379386,
      "num_tokens": 113590112.0,
      "step": 94460
    },
    {
      "entropy": 1.850995273888111,
      "epoch": 0.2928487343444731,
      "grad_norm": 9.236472129821777,
      "learning_rate": 4.6749105171399864e-06,
      "loss": 0.4171,
      "mean_token_accuracy": 0.8592243269085884,
      "num_tokens": 113602127.0,
      "step": 94470
    },
    {
      "entropy": 1.894811224937439,
      "epoch": 0.29287973346952284,
      "grad_norm": 7.549960136413574,
      "learning_rate": 4.674663105851442e-06,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8472552940249443,
      "num_tokens": 113613991.0,
      "step": 94480
    },
    {
      "entropy": 1.8211253330111503,
      "epoch": 0.2929107325945725,
      "grad_norm": 10.19444751739502,
      "learning_rate": 4.674415733840143e-06,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8452879965305329,
      "num_tokens": 113626732.0,
      "step": 94490
    },
    {
      "entropy": 1.870133863389492,
      "epoch": 0.29294173171962223,
      "grad_norm": 5.497763633728027,
      "learning_rate": 4.674168401095697e-06,
      "loss": 0.5169,
      "mean_token_accuracy": 0.8333605661988258,
      "num_tokens": 113638260.0,
      "step": 94500
    },
    {
      "entropy": 1.881983858346939,
      "epoch": 0.2929727308446719,
      "grad_norm": 7.503384590148926,
      "learning_rate": 4.673921107607716e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8480863809585572,
      "num_tokens": 113649729.0,
      "step": 94510
    },
    {
      "entropy": 1.7449950769543647,
      "epoch": 0.2930037299697216,
      "grad_norm": 6.960958003997803,
      "learning_rate": 4.673673853365818e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8576125666499138,
      "num_tokens": 113662822.0,
      "step": 94520
    },
    {
      "entropy": 1.6805919706821442,
      "epoch": 0.2930347290947713,
      "grad_norm": 7.576868534088135,
      "learning_rate": 4.673426638359622e-06,
      "loss": 0.3747,
      "mean_token_accuracy": 0.8681085333228111,
      "num_tokens": 113677064.0,
      "step": 94530
    },
    {
      "entropy": 1.8898225530982018,
      "epoch": 0.293065728219821,
      "grad_norm": 8.339954376220703,
      "learning_rate": 4.673179462578754e-06,
      "loss": 0.5254,
      "mean_token_accuracy": 0.834310744702816,
      "num_tokens": 113688785.0,
      "step": 94540
    },
    {
      "entropy": 1.822121039032936,
      "epoch": 0.2930967273448707,
      "grad_norm": 2.9981343746185303,
      "learning_rate": 4.672932326012839e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8623607367277145,
      "num_tokens": 113700761.0,
      "step": 94550
    },
    {
      "entropy": 1.8925957545638084,
      "epoch": 0.2931277264699204,
      "grad_norm": 8.26031494140625,
      "learning_rate": 4.672685228651511e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8499511271715164,
      "num_tokens": 113712498.0,
      "step": 94560
    },
    {
      "entropy": 1.92208162099123,
      "epoch": 0.2931587255949701,
      "grad_norm": 9.060463905334473,
      "learning_rate": 4.672438170484405e-06,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8314491465687752,
      "num_tokens": 113724082.0,
      "step": 94570
    },
    {
      "entropy": 1.8770214468240738,
      "epoch": 0.29318972472001975,
      "grad_norm": 8.743376731872559,
      "learning_rate": 4.672191151501161e-06,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8547257348895073,
      "num_tokens": 113736315.0,
      "step": 94580
    },
    {
      "entropy": 1.837056641280651,
      "epoch": 0.2932207238450695,
      "grad_norm": 4.969115257263184,
      "learning_rate": 4.671944171691422e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8457030534744263,
      "num_tokens": 113748972.0,
      "step": 94590
    },
    {
      "entropy": 1.9361342743039132,
      "epoch": 0.29325172297011914,
      "grad_norm": 8.590668678283691,
      "learning_rate": 4.671697231044837e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8455305263400078,
      "num_tokens": 113760795.0,
      "step": 94600
    },
    {
      "entropy": 1.775714547932148,
      "epoch": 0.29328272209516887,
      "grad_norm": 3.653806209564209,
      "learning_rate": 4.671450329551054e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8520709335803985,
      "num_tokens": 113774218.0,
      "step": 94610
    },
    {
      "entropy": 1.9447847425937652,
      "epoch": 0.29331372122021854,
      "grad_norm": 8.897464752197266,
      "learning_rate": 4.671203467199731e-06,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8491786658763886,
      "num_tokens": 113785173.0,
      "step": 94620
    },
    {
      "entropy": 1.8407832726836204,
      "epoch": 0.29334472034526826,
      "grad_norm": 6.721490383148193,
      "learning_rate": 4.670956643980524e-06,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8475279733538628,
      "num_tokens": 113797753.0,
      "step": 94630
    },
    {
      "entropy": 1.7925242587924004,
      "epoch": 0.29337571947031793,
      "grad_norm": 8.641847610473633,
      "learning_rate": 4.670709859883096e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8569319665431976,
      "num_tokens": 113809733.0,
      "step": 94640
    },
    {
      "entropy": 1.9358911573886872,
      "epoch": 0.29340671859536765,
      "grad_norm": 8.337167739868164,
      "learning_rate": 4.670463114897114e-06,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8383975982666015,
      "num_tokens": 113820542.0,
      "step": 94650
    },
    {
      "entropy": 1.8865822792053222,
      "epoch": 0.2934377177204173,
      "grad_norm": 8.994181632995605,
      "learning_rate": 4.670216409012248e-06,
      "loss": 0.49,
      "mean_token_accuracy": 0.8425557687878609,
      "num_tokens": 113832371.0,
      "step": 94660
    },
    {
      "entropy": 1.7770501509308816,
      "epoch": 0.29346871684546705,
      "grad_norm": 6.771418571472168,
      "learning_rate": 4.669969742218173e-06,
      "loss": 0.431,
      "mean_token_accuracy": 0.8530333042144775,
      "num_tokens": 113845284.0,
      "step": 94670
    },
    {
      "entropy": 1.7958509385585786,
      "epoch": 0.2934997159705167,
      "grad_norm": 7.999570369720459,
      "learning_rate": 4.669723114504565e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8648405924439431,
      "num_tokens": 113858418.0,
      "step": 94680
    },
    {
      "entropy": 1.9169925883412362,
      "epoch": 0.29353071509556644,
      "grad_norm": 8.945980072021484,
      "learning_rate": 4.669476525861107e-06,
      "loss": 0.5081,
      "mean_token_accuracy": 0.8366316080093383,
      "num_tokens": 113869941.0,
      "step": 94690
    },
    {
      "entropy": 1.804078498482704,
      "epoch": 0.2935617142206161,
      "grad_norm": 8.323823928833008,
      "learning_rate": 4.669229976277483e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8675397202372551,
      "num_tokens": 113882565.0,
      "step": 94700
    },
    {
      "entropy": 1.9035957649350166,
      "epoch": 0.29359271334566583,
      "grad_norm": 9.34669017791748,
      "learning_rate": 4.668983465743385e-06,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8365237683057785,
      "num_tokens": 113893898.0,
      "step": 94710
    },
    {
      "entropy": 1.9729483425617218,
      "epoch": 0.2936237124707155,
      "grad_norm": 6.514955043792725,
      "learning_rate": 4.668736994248504e-06,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8372229024767875,
      "num_tokens": 113904833.0,
      "step": 94720
    },
    {
      "entropy": 1.9153319388628005,
      "epoch": 0.2936547115957652,
      "grad_norm": 11.098676681518555,
      "learning_rate": 4.668490561782535e-06,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8335372865200043,
      "num_tokens": 113915799.0,
      "step": 94730
    },
    {
      "entropy": 1.9642977565526962,
      "epoch": 0.2936857107208149,
      "grad_norm": 8.632896423339844,
      "learning_rate": 4.668244168335182e-06,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8496646553277969,
      "num_tokens": 113926831.0,
      "step": 94740
    },
    {
      "entropy": 1.8451596662402152,
      "epoch": 0.2937167098458646,
      "grad_norm": 8.402206420898438,
      "learning_rate": 4.667997813896149e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8657459333539009,
      "num_tokens": 113938570.0,
      "step": 94750
    },
    {
      "entropy": 1.8689325533807277,
      "epoch": 0.2937477089709143,
      "grad_norm": 9.82726001739502,
      "learning_rate": 4.667751498455142e-06,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8491661697626114,
      "num_tokens": 113950581.0,
      "step": 94760
    },
    {
      "entropy": 1.81126269698143,
      "epoch": 0.293778708095964,
      "grad_norm": 2.5253586769104004,
      "learning_rate": 4.667505222001875e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.8603689581155777,
      "num_tokens": 113963539.0,
      "step": 94770
    },
    {
      "entropy": 1.8925784215331078,
      "epoch": 0.2938097072210137,
      "grad_norm": 7.797617435455322,
      "learning_rate": 4.667258984526063e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8523267209529877,
      "num_tokens": 113975014.0,
      "step": 94780
    },
    {
      "entropy": 1.8433985903859138,
      "epoch": 0.2938407063460634,
      "grad_norm": 10.589468002319336,
      "learning_rate": 4.667012786017426e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8405422657728195,
      "num_tokens": 113987158.0,
      "step": 94790
    },
    {
      "entropy": 1.8784984156489373,
      "epoch": 0.2938717054711131,
      "grad_norm": 3.9485971927642822,
      "learning_rate": 4.666766626465689e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8542310744524002,
      "num_tokens": 113998286.0,
      "step": 94800
    },
    {
      "entropy": 1.8718909628689289,
      "epoch": 0.2939027045961628,
      "grad_norm": 8.77500057220459,
      "learning_rate": 4.666520505860577e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.8534493952989578,
      "num_tokens": 114010433.0,
      "step": 94810
    },
    {
      "entropy": 1.9200672134757042,
      "epoch": 0.29393370372121247,
      "grad_norm": 8.568684577941895,
      "learning_rate": 4.666274424191821e-06,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8516236767172813,
      "num_tokens": 114021482.0,
      "step": 94820
    },
    {
      "entropy": 1.8536953687667848,
      "epoch": 0.29396470284626214,
      "grad_norm": 7.200621128082275,
      "learning_rate": 4.666028381449159e-06,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8599602654576302,
      "num_tokens": 114033985.0,
      "step": 94830
    },
    {
      "entropy": 1.894244983792305,
      "epoch": 0.29399570197131186,
      "grad_norm": 10.332306861877441,
      "learning_rate": 4.6657823776223255e-06,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8414657115936279,
      "num_tokens": 114045285.0,
      "step": 94840
    },
    {
      "entropy": 1.863986374437809,
      "epoch": 0.29402670109636153,
      "grad_norm": 4.247035980224609,
      "learning_rate": 4.6655364127010655e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8416002795100213,
      "num_tokens": 114057674.0,
      "step": 94850
    },
    {
      "entropy": 1.9077302902936935,
      "epoch": 0.29405770022141126,
      "grad_norm": 7.60083532333374,
      "learning_rate": 4.665290486675124e-06,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8370777696371079,
      "num_tokens": 114069600.0,
      "step": 94860
    },
    {
      "entropy": 1.8785944551229476,
      "epoch": 0.2940886993464609,
      "grad_norm": 7.80269718170166,
      "learning_rate": 4.665044599534251e-06,
      "loss": 0.4277,
      "mean_token_accuracy": 0.852102880179882,
      "num_tokens": 114081686.0,
      "step": 94870
    },
    {
      "entropy": 1.8737225562334061,
      "epoch": 0.29411969847151065,
      "grad_norm": 8.264700889587402,
      "learning_rate": 4.664798751268201e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8497600093483925,
      "num_tokens": 114093863.0,
      "step": 94880
    },
    {
      "entropy": 1.9434197053313256,
      "epoch": 0.2941506975965603,
      "grad_norm": 8.022650718688965,
      "learning_rate": 4.664552941866732e-06,
      "loss": 0.482,
      "mean_token_accuracy": 0.8454866543412208,
      "num_tokens": 114105246.0,
      "step": 94890
    },
    {
      "entropy": 1.7929591804742813,
      "epoch": 0.29418169672161004,
      "grad_norm": 4.1385908126831055,
      "learning_rate": 4.664307171319604e-06,
      "loss": 0.4195,
      "mean_token_accuracy": 0.84492729306221,
      "num_tokens": 114118876.0,
      "step": 94900
    },
    {
      "entropy": 1.9504995405673982,
      "epoch": 0.2942126958466597,
      "grad_norm": 7.590723037719727,
      "learning_rate": 4.664061439616583e-06,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8555523350834846,
      "num_tokens": 114129331.0,
      "step": 94910
    },
    {
      "entropy": 1.8454117804765702,
      "epoch": 0.29424369497170944,
      "grad_norm": 10.14136791229248,
      "learning_rate": 4.663815746747437e-06,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8382917091250419,
      "num_tokens": 114141760.0,
      "step": 94920
    },
    {
      "entropy": 1.8560292690992355,
      "epoch": 0.2942746940967591,
      "grad_norm": 7.97593879699707,
      "learning_rate": 4.66357009270194e-06,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8398404493927956,
      "num_tokens": 114154144.0,
      "step": 94930
    },
    {
      "entropy": 1.9188194185495377,
      "epoch": 0.29430569322180883,
      "grad_norm": 8.06023120880127,
      "learning_rate": 4.663324477469867e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8514153569936752,
      "num_tokens": 114165232.0,
      "step": 94940
    },
    {
      "entropy": 1.9558087676763534,
      "epoch": 0.2943366923468585,
      "grad_norm": 9.450501441955566,
      "learning_rate": 4.663078901040999e-06,
      "loss": 0.5441,
      "mean_token_accuracy": 0.8372850298881531,
      "num_tokens": 114176045.0,
      "step": 94950
    },
    {
      "entropy": 1.8351658523082732,
      "epoch": 0.2943676914719082,
      "grad_norm": 4.117894649505615,
      "learning_rate": 4.662833363405119e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8479064837098121,
      "num_tokens": 114189019.0,
      "step": 94960
    },
    {
      "entropy": 1.8557307586073875,
      "epoch": 0.2943986905969579,
      "grad_norm": 8.198458671569824,
      "learning_rate": 4.662587864552017e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8527988627552986,
      "num_tokens": 114201808.0,
      "step": 94970
    },
    {
      "entropy": 1.927764955163002,
      "epoch": 0.2944296897220076,
      "grad_norm": 9.65707015991211,
      "learning_rate": 4.662342404471482e-06,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8513452783226967,
      "num_tokens": 114213290.0,
      "step": 94980
    },
    {
      "entropy": 1.76514892578125,
      "epoch": 0.2944606888470573,
      "grad_norm": 8.054937362670898,
      "learning_rate": 4.662096983153311e-06,
      "loss": 0.3548,
      "mean_token_accuracy": 0.8645235538482666,
      "num_tokens": 114226877.0,
      "step": 94990
    },
    {
      "entropy": 1.8766354367136955,
      "epoch": 0.294491687972107,
      "grad_norm": 4.424029350280762,
      "learning_rate": 4.661851600587301e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8427970930933952,
      "num_tokens": 114239241.0,
      "step": 95000
    },
    {
      "entropy": 1.9168139308691026,
      "epoch": 0.2945226870971567,
      "grad_norm": 7.28840970993042,
      "learning_rate": 4.661606256763257e-06,
      "loss": 0.528,
      "mean_token_accuracy": 0.8504084393382072,
      "num_tokens": 114250255.0,
      "step": 95010
    },
    {
      "entropy": 1.7996359452605248,
      "epoch": 0.2945536862222064,
      "grad_norm": 8.724395751953125,
      "learning_rate": 4.661360951670983e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8572775185108185,
      "num_tokens": 114262731.0,
      "step": 95020
    },
    {
      "entropy": 1.847113935649395,
      "epoch": 0.29458468534725607,
      "grad_norm": 8.312970161437988,
      "learning_rate": 4.661115685300293e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8483634769916535,
      "num_tokens": 114274768.0,
      "step": 95030
    },
    {
      "entropy": 1.8922452926635742,
      "epoch": 0.2946156844723058,
      "grad_norm": 6.005066871643066,
      "learning_rate": 4.660870457640998e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.846984113752842,
      "num_tokens": 114286772.0,
      "step": 95040
    },
    {
      "entropy": 1.8506748765707015,
      "epoch": 0.29464668359735546,
      "grad_norm": 3.484794855117798,
      "learning_rate": 4.660625268682915e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8471810176968575,
      "num_tokens": 114298457.0,
      "step": 95050
    },
    {
      "entropy": 1.952257016301155,
      "epoch": 0.2946776827224052,
      "grad_norm": 9.706525802612305,
      "learning_rate": 4.660380118415869e-06,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8365617975592613,
      "num_tokens": 114309656.0,
      "step": 95060
    },
    {
      "entropy": 1.854885457456112,
      "epoch": 0.29470868184745486,
      "grad_norm": 7.192900657653809,
      "learning_rate": 4.660135006829682e-06,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8500230267643929,
      "num_tokens": 114321375.0,
      "step": 95070
    },
    {
      "entropy": 1.8961684226989746,
      "epoch": 0.2947396809725045,
      "grad_norm": 8.305758476257324,
      "learning_rate": 4.659889933914185e-06,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8406580999493599,
      "num_tokens": 114333189.0,
      "step": 95080
    },
    {
      "entropy": 1.914514508843422,
      "epoch": 0.29477068009755425,
      "grad_norm": 7.5317702293396,
      "learning_rate": 4.65964489965921e-06,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8434663847088814,
      "num_tokens": 114344937.0,
      "step": 95090
    },
    {
      "entropy": 1.8632069528102875,
      "epoch": 0.2948016792226039,
      "grad_norm": 8.437536239624023,
      "learning_rate": 4.659399904054594e-06,
      "loss": 0.512,
      "mean_token_accuracy": 0.8462415754795074,
      "num_tokens": 114357269.0,
      "step": 95100
    },
    {
      "entropy": 1.9126679688692092,
      "epoch": 0.29483267834765364,
      "grad_norm": 4.100780010223389,
      "learning_rate": 4.6591549470901755e-06,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8231175735592842,
      "num_tokens": 114369778.0,
      "step": 95110
    },
    {
      "entropy": 1.84993801638484,
      "epoch": 0.2948636774727033,
      "grad_norm": 8.060639381408691,
      "learning_rate": 4.6589100287558015e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8632707759737969,
      "num_tokens": 114383231.0,
      "step": 95120
    },
    {
      "entropy": 1.9371715486049652,
      "epoch": 0.29489467659775304,
      "grad_norm": 7.608775615692139,
      "learning_rate": 4.658665149041318e-06,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8381532579660416,
      "num_tokens": 114393685.0,
      "step": 95130
    },
    {
      "entropy": 1.8963730677962303,
      "epoch": 0.2949256757228027,
      "grad_norm": 7.8309550285339355,
      "learning_rate": 4.6584203079365756e-06,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8506500810384751,
      "num_tokens": 114404986.0,
      "step": 95140
    },
    {
      "entropy": 1.8873082131147385,
      "epoch": 0.29495667484785243,
      "grad_norm": 8.878643989562988,
      "learning_rate": 4.658175505431431e-06,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8476304590702057,
      "num_tokens": 114416424.0,
      "step": 95150
    },
    {
      "entropy": 1.7684010401368142,
      "epoch": 0.2949876739729021,
      "grad_norm": 4.694974422454834,
      "learning_rate": 4.657930741515742e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.853473761677742,
      "num_tokens": 114430299.0,
      "step": 95160
    },
    {
      "entropy": 1.8848821252584458,
      "epoch": 0.2950186730979518,
      "grad_norm": 9.460819244384766,
      "learning_rate": 4.657686016179372e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8388624176383018,
      "num_tokens": 114441497.0,
      "step": 95170
    },
    {
      "entropy": 1.8895638823509215,
      "epoch": 0.2950496722230015,
      "grad_norm": 7.845742702484131,
      "learning_rate": 4.6574413294121865e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.840872798860073,
      "num_tokens": 114453312.0,
      "step": 95180
    },
    {
      "entropy": 1.8671682730317116,
      "epoch": 0.2950806713480512,
      "grad_norm": 8.617532730102539,
      "learning_rate": 4.657196681204057e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.858017711341381,
      "num_tokens": 114464871.0,
      "step": 95190
    },
    {
      "entropy": 1.8859140858054162,
      "epoch": 0.2951116704731009,
      "grad_norm": 8.359928131103516,
      "learning_rate": 4.656952071544857e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8446494430303574,
      "num_tokens": 114476109.0,
      "step": 95200
    },
    {
      "entropy": 1.9119530409574508,
      "epoch": 0.2951426695981506,
      "grad_norm": 9.479840278625488,
      "learning_rate": 4.656707500424463e-06,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8339703768491745,
      "num_tokens": 114487221.0,
      "step": 95210
    },
    {
      "entropy": 1.8878633320331573,
      "epoch": 0.2951736687232003,
      "grad_norm": 3.638047218322754,
      "learning_rate": 4.656462967832758e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8457856431603432,
      "num_tokens": 114498809.0,
      "step": 95220
    },
    {
      "entropy": 1.893609069287777,
      "epoch": 0.29520466784825,
      "grad_norm": 7.487365245819092,
      "learning_rate": 4.656218473759623e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.849824258685112,
      "num_tokens": 114510866.0,
      "step": 95230
    },
    {
      "entropy": 1.9306818440556526,
      "epoch": 0.29523566697329967,
      "grad_norm": 7.329023838043213,
      "learning_rate": 4.655974018194953e-06,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8432904377579689,
      "num_tokens": 114521966.0,
      "step": 95240
    },
    {
      "entropy": 1.9246254205703734,
      "epoch": 0.2952666660983494,
      "grad_norm": 8.638476371765137,
      "learning_rate": 4.655729601128635e-06,
      "loss": 0.493,
      "mean_token_accuracy": 0.8454683497548103,
      "num_tokens": 114533785.0,
      "step": 95250
    },
    {
      "entropy": 1.8969884738326073,
      "epoch": 0.29529766522339906,
      "grad_norm": 7.14600944519043,
      "learning_rate": 4.655485222550568e-06,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8438623622059822,
      "num_tokens": 114545791.0,
      "step": 95260
    },
    {
      "entropy": 1.8377904728055001,
      "epoch": 0.2953286643484488,
      "grad_norm": 6.43510103225708,
      "learning_rate": 4.65524088245065e-06,
      "loss": 0.433,
      "mean_token_accuracy": 0.8514227852225303,
      "num_tokens": 114558462.0,
      "step": 95270
    },
    {
      "entropy": 1.9160344362258912,
      "epoch": 0.29535966347349846,
      "grad_norm": 10.534104347229004,
      "learning_rate": 4.654996580818786e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.845592126250267,
      "num_tokens": 114569485.0,
      "step": 95280
    },
    {
      "entropy": 1.8697632998228073,
      "epoch": 0.2953906625985482,
      "grad_norm": 8.796942710876465,
      "learning_rate": 4.654752317644883e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8387501269578934,
      "num_tokens": 114581277.0,
      "step": 95290
    },
    {
      "entropy": 1.831010890007019,
      "epoch": 0.29542166172359785,
      "grad_norm": 9.258670806884766,
      "learning_rate": 4.654508092918852e-06,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8350206702947617,
      "num_tokens": 114593427.0,
      "step": 95300
    },
    {
      "entropy": 1.8626672565937041,
      "epoch": 0.2954526608486475,
      "grad_norm": 4.265320777893066,
      "learning_rate": 4.6542639066306065e-06,
      "loss": 0.4421,
      "mean_token_accuracy": 0.8505603194236755,
      "num_tokens": 114605252.0,
      "step": 95310
    },
    {
      "entropy": 1.7814742475748062,
      "epoch": 0.29548365997369724,
      "grad_norm": 9.163718223571777,
      "learning_rate": 4.654019758770067e-06,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8585087850689888,
      "num_tokens": 114618618.0,
      "step": 95320
    },
    {
      "entropy": 1.8076193630695343,
      "epoch": 0.2955146590987469,
      "grad_norm": 7.5134477615356445,
      "learning_rate": 4.653775649327154e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8535345688462257,
      "num_tokens": 114632466.0,
      "step": 95330
    },
    {
      "entropy": 1.878467933833599,
      "epoch": 0.29554565822379664,
      "grad_norm": 3.287616729736328,
      "learning_rate": 4.653531578291793e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8520422443747521,
      "num_tokens": 114644129.0,
      "step": 95340
    },
    {
      "entropy": 1.7941744670271873,
      "epoch": 0.2955766573488463,
      "grad_norm": 8.888310432434082,
      "learning_rate": 4.653287545653915e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8592835500836372,
      "num_tokens": 114657506.0,
      "step": 95350
    },
    {
      "entropy": 1.842213323712349,
      "epoch": 0.29560765647389603,
      "grad_norm": 8.71075439453125,
      "learning_rate": 4.653043551403452e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8478049322962761,
      "num_tokens": 114669587.0,
      "step": 95360
    },
    {
      "entropy": 1.8689180329442023,
      "epoch": 0.2956386555989457,
      "grad_norm": 7.912755489349365,
      "learning_rate": 4.652799595530342e-06,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8550763368606568,
      "num_tokens": 114682059.0,
      "step": 95370
    },
    {
      "entropy": 1.9113721013069154,
      "epoch": 0.2956696547239954,
      "grad_norm": 9.693363189697266,
      "learning_rate": 4.652555678024524e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.853804387152195,
      "num_tokens": 114693470.0,
      "step": 95380
    },
    {
      "entropy": 1.8694558992981911,
      "epoch": 0.2957006538490451,
      "grad_norm": 8.779112815856934,
      "learning_rate": 4.652311798875943e-06,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8518740639090538,
      "num_tokens": 114705358.0,
      "step": 95390
    },
    {
      "entropy": 1.8623104050755501,
      "epoch": 0.2957316529740948,
      "grad_norm": 9.614526748657227,
      "learning_rate": 4.652067958074547e-06,
      "loss": 0.4875,
      "mean_token_accuracy": 0.8510076805949212,
      "num_tokens": 114718125.0,
      "step": 95400
    },
    {
      "entropy": 1.8787400797009468,
      "epoch": 0.2957626520991445,
      "grad_norm": 8.458456039428711,
      "learning_rate": 4.651824155610288e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8569131478667259,
      "num_tokens": 114729728.0,
      "step": 95410
    },
    {
      "entropy": 1.9131328269839287,
      "epoch": 0.2957936512241942,
      "grad_norm": 9.49421215057373,
      "learning_rate": 4.6515803914731215e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.845634426176548,
      "num_tokens": 114742431.0,
      "step": 95420
    },
    {
      "entropy": 1.7844593867659568,
      "epoch": 0.2958246503492439,
      "grad_norm": 8.483735084533691,
      "learning_rate": 4.651336665653005e-06,
      "loss": 0.3808,
      "mean_token_accuracy": 0.8685527250170708,
      "num_tokens": 114755668.0,
      "step": 95430
    },
    {
      "entropy": 1.922539620101452,
      "epoch": 0.2958556494742936,
      "grad_norm": 7.755213260650635,
      "learning_rate": 4.651092978139902e-06,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8362011343240738,
      "num_tokens": 114766944.0,
      "step": 95440
    },
    {
      "entropy": 1.8400570914149283,
      "epoch": 0.29588664859934327,
      "grad_norm": 7.850890636444092,
      "learning_rate": 4.650849328923779e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8529717952013016,
      "num_tokens": 114779470.0,
      "step": 95450
    },
    {
      "entropy": 1.8972987502813339,
      "epoch": 0.295917647724393,
      "grad_norm": 10.900968551635742,
      "learning_rate": 4.650605717994607e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8552507728338241,
      "num_tokens": 114791012.0,
      "step": 95460
    },
    {
      "entropy": 1.904093398153782,
      "epoch": 0.29594864684944266,
      "grad_norm": 7.401911735534668,
      "learning_rate": 4.650362145342358e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8534610956907273,
      "num_tokens": 114802550.0,
      "step": 95470
    },
    {
      "entropy": 1.8667383641004562,
      "epoch": 0.2959796459744924,
      "grad_norm": 7.704689025878906,
      "learning_rate": 4.650118610957009e-06,
      "loss": 0.468,
      "mean_token_accuracy": 0.8388277173042298,
      "num_tokens": 114814877.0,
      "step": 95480
    },
    {
      "entropy": 1.8931495070457458,
      "epoch": 0.29601064509954206,
      "grad_norm": 9.703409194946289,
      "learning_rate": 4.649875114828544e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8397621661424637,
      "num_tokens": 114826671.0,
      "step": 95490
    },
    {
      "entropy": 1.949988493323326,
      "epoch": 0.2960416442245918,
      "grad_norm": 7.655276298522949,
      "learning_rate": 4.6496316569469455e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.8395925149321556,
      "num_tokens": 114837908.0,
      "step": 95500
    },
    {
      "entropy": 1.9334761276841164,
      "epoch": 0.29607264334964145,
      "grad_norm": 8.131454467773438,
      "learning_rate": 4.649388237302203e-06,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8403454497456551,
      "num_tokens": 114849471.0,
      "step": 95510
    },
    {
      "entropy": 1.902390295267105,
      "epoch": 0.2961036424746912,
      "grad_norm": 8.340156555175781,
      "learning_rate": 4.6491448558843065e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8446046605706214,
      "num_tokens": 114861012.0,
      "step": 95520
    },
    {
      "entropy": 1.9234414175152779,
      "epoch": 0.29613464159974084,
      "grad_norm": 9.107486724853516,
      "learning_rate": 4.648901512683255e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.84765195697546,
      "num_tokens": 114872674.0,
      "step": 95530
    },
    {
      "entropy": 1.8331530943512917,
      "epoch": 0.29616564072479057,
      "grad_norm": 8.991997718811035,
      "learning_rate": 4.648658207689045e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8575505539774895,
      "num_tokens": 114885367.0,
      "step": 95540
    },
    {
      "entropy": 1.8234670519828797,
      "epoch": 0.29619663984984024,
      "grad_norm": 5.19529914855957,
      "learning_rate": 4.648414940891681e-06,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8523637726902962,
      "num_tokens": 114899055.0,
      "step": 95550
    },
    {
      "entropy": 1.8568585798144341,
      "epoch": 0.2962276389748899,
      "grad_norm": 4.264057636260986,
      "learning_rate": 4.648171712281169e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8576111733913422,
      "num_tokens": 114911683.0,
      "step": 95560
    },
    {
      "entropy": 1.9596162497997285,
      "epoch": 0.29625863809993963,
      "grad_norm": 9.092710494995117,
      "learning_rate": 4.64792852184752e-06,
      "loss": 0.5458,
      "mean_token_accuracy": 0.8393025726079941,
      "num_tokens": 114922365.0,
      "step": 95570
    },
    {
      "entropy": 1.8301913827657699,
      "epoch": 0.2962896372249893,
      "grad_norm": 7.440187454223633,
      "learning_rate": 4.647685369580747e-06,
      "loss": 0.4163,
      "mean_token_accuracy": 0.857902692258358,
      "num_tokens": 114935376.0,
      "step": 95580
    },
    {
      "entropy": 1.9472972556948662,
      "epoch": 0.296320636350039,
      "grad_norm": 8.745131492614746,
      "learning_rate": 4.64744225547087e-06,
      "loss": 0.5434,
      "mean_token_accuracy": 0.828769737482071,
      "num_tokens": 114946969.0,
      "step": 95590
    },
    {
      "entropy": 1.935869987308979,
      "epoch": 0.2963516354750887,
      "grad_norm": 7.479891777038574,
      "learning_rate": 4.647199179507909e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8483911573886871,
      "num_tokens": 114958473.0,
      "step": 95600
    },
    {
      "entropy": 1.899821263551712,
      "epoch": 0.2963826346001384,
      "grad_norm": 7.5076680183410645,
      "learning_rate": 4.646956141681888e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.845979979634285,
      "num_tokens": 114969919.0,
      "step": 95610
    },
    {
      "entropy": 1.9035681039094925,
      "epoch": 0.2964136337251881,
      "grad_norm": 8.332342147827148,
      "learning_rate": 4.646713141982837e-06,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8571434035897255,
      "num_tokens": 114982144.0,
      "step": 95620
    },
    {
      "entropy": 1.924658827483654,
      "epoch": 0.2964446328502378,
      "grad_norm": 9.30444622039795,
      "learning_rate": 4.646470180400788e-06,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8478532791137695,
      "num_tokens": 114993947.0,
      "step": 95630
    },
    {
      "entropy": 1.8619504556059838,
      "epoch": 0.2964756319752875,
      "grad_norm": 5.051435470581055,
      "learning_rate": 4.646227256925777e-06,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8470785349607468,
      "num_tokens": 115006275.0,
      "step": 95640
    },
    {
      "entropy": 1.8626134410500526,
      "epoch": 0.2965066311003372,
      "grad_norm": 4.338906288146973,
      "learning_rate": 4.645984371547844e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8533647701144218,
      "num_tokens": 115018973.0,
      "step": 95650
    },
    {
      "entropy": 1.8923157513141633,
      "epoch": 0.2965376302253869,
      "grad_norm": 8.258339881896973,
      "learning_rate": 4.645741524257032e-06,
      "loss": 0.5561,
      "mean_token_accuracy": 0.8394008025527,
      "num_tokens": 115030711.0,
      "step": 95660
    },
    {
      "entropy": 1.9015984013676643,
      "epoch": 0.2965686293504366,
      "grad_norm": 10.353814125061035,
      "learning_rate": 4.645498715043387e-06,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8356149435043335,
      "num_tokens": 115042545.0,
      "step": 95670
    },
    {
      "entropy": 1.9278481543064117,
      "epoch": 0.29659962847548627,
      "grad_norm": 8.37845516204834,
      "learning_rate": 4.645255943896961e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8446251779794693,
      "num_tokens": 115054273.0,
      "step": 95680
    },
    {
      "entropy": 1.9885007172822953,
      "epoch": 0.296630627600536,
      "grad_norm": 8.631330490112305,
      "learning_rate": 4.645013210807806e-06,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8454102978110314,
      "num_tokens": 115064894.0,
      "step": 95690
    },
    {
      "entropy": 1.8312623113393784,
      "epoch": 0.29666162672558566,
      "grad_norm": 8.99056339263916,
      "learning_rate": 4.644770515765983e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.8604427292943001,
      "num_tokens": 115077844.0,
      "step": 95700
    },
    {
      "entropy": 1.8540172457695008,
      "epoch": 0.2966926258506354,
      "grad_norm": 2.9684078693389893,
      "learning_rate": 4.64452785876155e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.844263382256031,
      "num_tokens": 115090181.0,
      "step": 95710
    },
    {
      "entropy": 1.9741147756576538,
      "epoch": 0.29672362497568505,
      "grad_norm": 8.870850563049316,
      "learning_rate": 4.644285239784575e-06,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8312768504023552,
      "num_tokens": 115100406.0,
      "step": 95720
    },
    {
      "entropy": 1.9829291343688964,
      "epoch": 0.2967546241007348,
      "grad_norm": 3.856050729751587,
      "learning_rate": 4.644042658825126e-06,
      "loss": 0.538,
      "mean_token_accuracy": 0.8350884988903999,
      "num_tokens": 115111439.0,
      "step": 95730
    },
    {
      "entropy": 1.9604087606072427,
      "epoch": 0.29678562322578445,
      "grad_norm": 8.375349998474121,
      "learning_rate": 4.643800115873274e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8406010925769806,
      "num_tokens": 115122511.0,
      "step": 95740
    },
    {
      "entropy": 1.8680056795477866,
      "epoch": 0.29681662235083417,
      "grad_norm": 3.6315958499908447,
      "learning_rate": 4.643557610919095e-06,
      "loss": 0.4559,
      "mean_token_accuracy": 0.851527801156044,
      "num_tokens": 115134569.0,
      "step": 95750
    },
    {
      "entropy": 1.860308752954006,
      "epoch": 0.29684762147588384,
      "grad_norm": 7.783955097198486,
      "learning_rate": 4.643315143952671e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8580260217189789,
      "num_tokens": 115146701.0,
      "step": 95760
    },
    {
      "entropy": 1.9309222459793092,
      "epoch": 0.29687862060093356,
      "grad_norm": 8.806280136108398,
      "learning_rate": 4.643072714964084e-06,
      "loss": 0.5613,
      "mean_token_accuracy": 0.836700190603733,
      "num_tokens": 115157808.0,
      "step": 95770
    },
    {
      "entropy": 1.9781962364912034,
      "epoch": 0.29690961972598323,
      "grad_norm": 7.529308319091797,
      "learning_rate": 4.64283032394342e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.8430225938558579,
      "num_tokens": 115168935.0,
      "step": 95780
    },
    {
      "entropy": 1.8590948715806008,
      "epoch": 0.29694061885103296,
      "grad_norm": 8.396674156188965,
      "learning_rate": 4.642587970880769e-06,
      "loss": 0.4035,
      "mean_token_accuracy": 0.8598534435033798,
      "num_tokens": 115180678.0,
      "step": 95790
    },
    {
      "entropy": 1.8222449347376823,
      "epoch": 0.2969716179760826,
      "grad_norm": 5.250387191772461,
      "learning_rate": 4.642345655766227e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8539089500904083,
      "num_tokens": 115193480.0,
      "step": 95800
    },
    {
      "entropy": 1.8699547871947289,
      "epoch": 0.2970026171011323,
      "grad_norm": 10.275331497192383,
      "learning_rate": 4.642103378589891e-06,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8331056758761406,
      "num_tokens": 115205319.0,
      "step": 95810
    },
    {
      "entropy": 1.8728644341230392,
      "epoch": 0.297033616226182,
      "grad_norm": 3.5322608947753906,
      "learning_rate": 4.641861139341863e-06,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8516603350639343,
      "num_tokens": 115217604.0,
      "step": 95820
    },
    {
      "entropy": 1.9369212195277214,
      "epoch": 0.2970646153512317,
      "grad_norm": 7.612384796142578,
      "learning_rate": 4.641618938012246e-06,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8410941004753113,
      "num_tokens": 115229163.0,
      "step": 95830
    },
    {
      "entropy": 1.930888931453228,
      "epoch": 0.2970956144762814,
      "grad_norm": 8.152316093444824,
      "learning_rate": 4.64137677459115e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8360361099243164,
      "num_tokens": 115241832.0,
      "step": 95840
    },
    {
      "entropy": 1.860244083404541,
      "epoch": 0.2971266136013311,
      "grad_norm": 2.479391574859619,
      "learning_rate": 4.641134649068688e-06,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8496681988239289,
      "num_tokens": 115254340.0,
      "step": 95850
    },
    {
      "entropy": 1.8632259294390678,
      "epoch": 0.2971576127263808,
      "grad_norm": 4.353872299194336,
      "learning_rate": 4.640892561434973e-06,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8551715791225434,
      "num_tokens": 115266838.0,
      "step": 95860
    },
    {
      "entropy": 1.999795663356781,
      "epoch": 0.2971886118514305,
      "grad_norm": 9.249789237976074,
      "learning_rate": 4.640650511680128e-06,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8440039098262787,
      "num_tokens": 115277215.0,
      "step": 95870
    },
    {
      "entropy": 1.954987433552742,
      "epoch": 0.2972196109764802,
      "grad_norm": 13.373074531555176,
      "learning_rate": 4.640408499794271e-06,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8423504129052162,
      "num_tokens": 115287841.0,
      "step": 95880
    },
    {
      "entropy": 1.821416835486889,
      "epoch": 0.29725061010152987,
      "grad_norm": 3.3793182373046875,
      "learning_rate": 4.640166525767535e-06,
      "loss": 0.3991,
      "mean_token_accuracy": 0.8553699105978012,
      "num_tokens": 115300828.0,
      "step": 95890
    },
    {
      "entropy": 1.9414212331175804,
      "epoch": 0.2972816092265796,
      "grad_norm": 3.7656142711639404,
      "learning_rate": 4.639924589590045e-06,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8337204396724701,
      "num_tokens": 115312598.0,
      "step": 95900
    },
    {
      "entropy": 1.9279131293296814,
      "epoch": 0.29731260835162926,
      "grad_norm": 10.359955787658691,
      "learning_rate": 4.639682691251938e-06,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8390220627188683,
      "num_tokens": 115324584.0,
      "step": 95910
    },
    {
      "entropy": 1.859154610335827,
      "epoch": 0.297343607476679,
      "grad_norm": 7.873774528503418,
      "learning_rate": 4.6394408307433494e-06,
      "loss": 0.3915,
      "mean_token_accuracy": 0.8545582070946693,
      "num_tokens": 115337114.0,
      "step": 95920
    },
    {
      "entropy": 1.9387451350688933,
      "epoch": 0.29737460660172865,
      "grad_norm": 10.652080535888672,
      "learning_rate": 4.639199008054421e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8364695340394974,
      "num_tokens": 115348545.0,
      "step": 95930
    },
    {
      "entropy": 2.010096028447151,
      "epoch": 0.2974056057267784,
      "grad_norm": 9.396890640258789,
      "learning_rate": 4.638957223175298e-06,
      "loss": 0.5485,
      "mean_token_accuracy": 0.8352597668766976,
      "num_tokens": 115359072.0,
      "step": 95940
    },
    {
      "entropy": 1.845473076403141,
      "epoch": 0.29743660485182805,
      "grad_norm": 7.396862030029297,
      "learning_rate": 4.638715476096127e-06,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8523558273911476,
      "num_tokens": 115372520.0,
      "step": 95950
    },
    {
      "entropy": 1.8642360031604768,
      "epoch": 0.29746760397687777,
      "grad_norm": 3.6565442085266113,
      "learning_rate": 4.638473766807061e-06,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8511311173439026,
      "num_tokens": 115385427.0,
      "step": 95960
    },
    {
      "entropy": 1.9817214131355285,
      "epoch": 0.29749860310192744,
      "grad_norm": 9.105077743530273,
      "learning_rate": 4.638232095298256e-06,
      "loss": 0.5501,
      "mean_token_accuracy": 0.8311897858977317,
      "num_tokens": 115396089.0,
      "step": 95970
    },
    {
      "entropy": 1.8148702546954154,
      "epoch": 0.29752960222697716,
      "grad_norm": 4.5015106201171875,
      "learning_rate": 4.6379904615598705e-06,
      "loss": 0.4107,
      "mean_token_accuracy": 0.8546622708439827,
      "num_tokens": 115408984.0,
      "step": 95980
    },
    {
      "entropy": 1.8370864361524581,
      "epoch": 0.29756060135202683,
      "grad_norm": 8.784217834472656,
      "learning_rate": 4.637748865582065e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8529459208250045,
      "num_tokens": 115421841.0,
      "step": 95990
    },
    {
      "entropy": 1.9331490993499756,
      "epoch": 0.29759160047707656,
      "grad_norm": 8.97538948059082,
      "learning_rate": 4.637507307355009e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8470474123954773,
      "num_tokens": 115432927.0,
      "step": 96000
    },
    {
      "entropy": 1.8707862794399261,
      "epoch": 0.2976225996021262,
      "grad_norm": 8.228986740112305,
      "learning_rate": 4.63726578686887e-06,
      "loss": 0.4487,
      "mean_token_accuracy": 0.857261911034584,
      "num_tokens": 115444664.0,
      "step": 96010
    },
    {
      "entropy": 1.9656662926077844,
      "epoch": 0.29765359872717595,
      "grad_norm": 8.407549858093262,
      "learning_rate": 4.637024304113822e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.847040732204914,
      "num_tokens": 115456093.0,
      "step": 96020
    },
    {
      "entropy": 2.015861451625824,
      "epoch": 0.2976845978522256,
      "grad_norm": 8.171123504638672,
      "learning_rate": 4.636782859080041e-06,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8443436101078987,
      "num_tokens": 115467201.0,
      "step": 96030
    },
    {
      "entropy": 2.0067085534334184,
      "epoch": 0.29771559697727534,
      "grad_norm": 9.067146301269531,
      "learning_rate": 4.636541451757711e-06,
      "loss": 0.5731,
      "mean_token_accuracy": 0.8257392793893814,
      "num_tokens": 115478314.0,
      "step": 96040
    },
    {
      "entropy": 1.8171662881970405,
      "epoch": 0.297746596102325,
      "grad_norm": 9.887383460998535,
      "learning_rate": 4.636300082137011e-06,
      "loss": 0.4195,
      "mean_token_accuracy": 0.860226346552372,
      "num_tokens": 115490891.0,
      "step": 96050
    },
    {
      "entropy": 1.8969256058335304,
      "epoch": 0.2977775952273747,
      "grad_norm": 3.9007980823516846,
      "learning_rate": 4.636058750208131e-06,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8481943354010582,
      "num_tokens": 115502955.0,
      "step": 96060
    },
    {
      "entropy": 1.855407053232193,
      "epoch": 0.2978085943524244,
      "grad_norm": 7.683308124542236,
      "learning_rate": 4.635817455961264e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8527497768402099,
      "num_tokens": 115516432.0,
      "step": 96070
    },
    {
      "entropy": 1.9460287556052207,
      "epoch": 0.2978395934774741,
      "grad_norm": 10.421119689941406,
      "learning_rate": 4.635576199386602e-06,
      "loss": 0.556,
      "mean_token_accuracy": 0.8432202488183975,
      "num_tokens": 115528309.0,
      "step": 96080
    },
    {
      "entropy": 1.9160909160971642,
      "epoch": 0.2978705926025238,
      "grad_norm": 7.117789268493652,
      "learning_rate": 4.635334980474345e-06,
      "loss": 0.5228,
      "mean_token_accuracy": 0.8440079033374787,
      "num_tokens": 115539915.0,
      "step": 96090
    },
    {
      "entropy": 1.9123343035578728,
      "epoch": 0.29790159172757347,
      "grad_norm": 7.615013122558594,
      "learning_rate": 4.635093799214693e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8447704032063484,
      "num_tokens": 115551725.0,
      "step": 96100
    },
    {
      "entropy": 1.8921069249510765,
      "epoch": 0.2979325908526232,
      "grad_norm": 8.272263526916504,
      "learning_rate": 4.634852655597854e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8503027930855751,
      "num_tokens": 115563104.0,
      "step": 96110
    },
    {
      "entropy": 1.9330130770802498,
      "epoch": 0.29796358997767286,
      "grad_norm": 7.9516143798828125,
      "learning_rate": 4.634611549614036e-06,
      "loss": 0.4694,
      "mean_token_accuracy": 0.846668167412281,
      "num_tokens": 115574420.0,
      "step": 96120
    },
    {
      "entropy": 1.930530808866024,
      "epoch": 0.2979945891027226,
      "grad_norm": 10.271747589111328,
      "learning_rate": 4.634370481253451e-06,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8329700931906701,
      "num_tokens": 115586106.0,
      "step": 96130
    },
    {
      "entropy": 1.9371146902441978,
      "epoch": 0.29802558822777225,
      "grad_norm": 8.164349555969238,
      "learning_rate": 4.634129450506316e-06,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8380087822675705,
      "num_tokens": 115597047.0,
      "step": 96140
    },
    {
      "entropy": 1.93800760358572,
      "epoch": 0.298056587352822,
      "grad_norm": 8.42188549041748,
      "learning_rate": 4.633888457362851e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8453196853399276,
      "num_tokens": 115608165.0,
      "step": 96150
    },
    {
      "entropy": 1.8701598271727562,
      "epoch": 0.29808758647787165,
      "grad_norm": 8.132607460021973,
      "learning_rate": 4.633647501813278e-06,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8482308521866798,
      "num_tokens": 115620759.0,
      "step": 96160
    },
    {
      "entropy": 1.8687379583716393,
      "epoch": 0.29811858560292137,
      "grad_norm": 3.990684747695923,
      "learning_rate": 4.633406583847825e-06,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8474869653582573,
      "num_tokens": 115633358.0,
      "step": 96170
    },
    {
      "entropy": 1.8821706905961038,
      "epoch": 0.29814958472797104,
      "grad_norm": 7.216353416442871,
      "learning_rate": 4.633165703456723e-06,
      "loss": 0.433,
      "mean_token_accuracy": 0.8504739284515381,
      "num_tokens": 115646131.0,
      "step": 96180
    },
    {
      "entropy": 1.98152334690094,
      "epoch": 0.29818058385302076,
      "grad_norm": 9.739982604980469,
      "learning_rate": 4.6329248606302045e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8421538904309273,
      "num_tokens": 115656444.0,
      "step": 96190
    },
    {
      "entropy": 1.7937684386968613,
      "epoch": 0.29821158297807043,
      "grad_norm": 4.097762107849121,
      "learning_rate": 4.6326840553585075e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.8663532704114913,
      "num_tokens": 115669849.0,
      "step": 96200
    },
    {
      "entropy": 1.960816130042076,
      "epoch": 0.29824258210312016,
      "grad_norm": 7.547084331512451,
      "learning_rate": 4.632443287631873e-06,
      "loss": 0.538,
      "mean_token_accuracy": 0.8430081829428673,
      "num_tokens": 115680669.0,
      "step": 96210
    },
    {
      "entropy": 1.8645331501960754,
      "epoch": 0.2982735812281698,
      "grad_norm": 4.153885364532471,
      "learning_rate": 4.632202557440546e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8473611772060394,
      "num_tokens": 115693855.0,
      "step": 96220
    },
    {
      "entropy": 1.932821586728096,
      "epoch": 0.29830458035321955,
      "grad_norm": 6.955663681030273,
      "learning_rate": 4.631961864774775e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.853759004175663,
      "num_tokens": 115705441.0,
      "step": 96230
    },
    {
      "entropy": 1.8800847560167313,
      "epoch": 0.2983355794782692,
      "grad_norm": 10.34147834777832,
      "learning_rate": 4.631721209624811e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8399926438927651,
      "num_tokens": 115718673.0,
      "step": 96240
    },
    {
      "entropy": 1.868877911567688,
      "epoch": 0.29836657860331894,
      "grad_norm": 9.02034854888916,
      "learning_rate": 4.63148059198091e-06,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8474683433771133,
      "num_tokens": 115731147.0,
      "step": 96250
    },
    {
      "entropy": 1.7052193224430083,
      "epoch": 0.2983975777283686,
      "grad_norm": 4.165556907653809,
      "learning_rate": 4.63124001183333e-06,
      "loss": 0.392,
      "mean_token_accuracy": 0.8648575574159623,
      "num_tokens": 115745846.0,
      "step": 96260
    },
    {
      "entropy": 1.8481420859694482,
      "epoch": 0.29842857685341834,
      "grad_norm": 8.495997428894043,
      "learning_rate": 4.630999469172333e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8595449611544609,
      "num_tokens": 115758179.0,
      "step": 96270
    },
    {
      "entropy": 1.8788498505949973,
      "epoch": 0.298459575978468,
      "grad_norm": 7.667300701141357,
      "learning_rate": 4.630758963988187e-06,
      "loss": 0.456,
      "mean_token_accuracy": 0.8503319293260574,
      "num_tokens": 115769809.0,
      "step": 96280
    },
    {
      "entropy": 1.9287839099764823,
      "epoch": 0.29849057510351773,
      "grad_norm": 8.444547653198242,
      "learning_rate": 4.63051849627116e-06,
      "loss": 0.4939,
      "mean_token_accuracy": 0.8440607368946076,
      "num_tokens": 115781425.0,
      "step": 96290
    },
    {
      "entropy": 1.9038120612502099,
      "epoch": 0.2985215742285674,
      "grad_norm": 7.381701946258545,
      "learning_rate": 4.630278066011525e-06,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8477517932653427,
      "num_tokens": 115793181.0,
      "step": 96300
    },
    {
      "entropy": 1.8960516452789307,
      "epoch": 0.29855257335361707,
      "grad_norm": 7.266051292419434,
      "learning_rate": 4.630037673199559e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8412499457597733,
      "num_tokens": 115805564.0,
      "step": 96310
    },
    {
      "entropy": 1.9981954544782639,
      "epoch": 0.2985835724786668,
      "grad_norm": 8.080859184265137,
      "learning_rate": 4.62979731782554e-06,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8427246376872063,
      "num_tokens": 115816757.0,
      "step": 96320
    },
    {
      "entropy": 1.824733631312847,
      "epoch": 0.29861457160371646,
      "grad_norm": 3.8523154258728027,
      "learning_rate": 4.629556999879755e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8527489557862282,
      "num_tokens": 115829066.0,
      "step": 96330
    },
    {
      "entropy": 1.9180657356977462,
      "epoch": 0.2986455707287662,
      "grad_norm": 9.480389595031738,
      "learning_rate": 4.629316719352488e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8383275628089905,
      "num_tokens": 115840829.0,
      "step": 96340
    },
    {
      "entropy": 1.8489946901798249,
      "epoch": 0.29867656985381585,
      "grad_norm": 4.126688003540039,
      "learning_rate": 4.629076476234032e-06,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8490898162126541,
      "num_tokens": 115853917.0,
      "step": 96350
    },
    {
      "entropy": 1.8371913895010947,
      "epoch": 0.2987075689788656,
      "grad_norm": 3.543614387512207,
      "learning_rate": 4.628836270514679e-06,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8514250427484512,
      "num_tokens": 115867183.0,
      "step": 96360
    },
    {
      "entropy": 1.853197917342186,
      "epoch": 0.29873856810391525,
      "grad_norm": 3.921837329864502,
      "learning_rate": 4.628596102184729e-06,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8422363549470901,
      "num_tokens": 115879237.0,
      "step": 96370
    },
    {
      "entropy": 1.818500466644764,
      "epoch": 0.29876956722896497,
      "grad_norm": 8.878031730651855,
      "learning_rate": 4.6283559712344825e-06,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8616446629166603,
      "num_tokens": 115892081.0,
      "step": 96380
    },
    {
      "entropy": 1.9318446889519691,
      "epoch": 0.29880056635401464,
      "grad_norm": 8.830316543579102,
      "learning_rate": 4.628115877654243e-06,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8325408533215523,
      "num_tokens": 115903514.0,
      "step": 96390
    },
    {
      "entropy": 1.8660483628511428,
      "epoch": 0.29883156547906436,
      "grad_norm": 8.622795104980469,
      "learning_rate": 4.627875821434319e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8535114452242851,
      "num_tokens": 115915306.0,
      "step": 96400
    },
    {
      "entropy": 1.8813782200217246,
      "epoch": 0.29886256460411403,
      "grad_norm": 9.0267333984375,
      "learning_rate": 4.627635802565024e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.846407724916935,
      "num_tokens": 115927498.0,
      "step": 96410
    },
    {
      "entropy": 1.9499717622995376,
      "epoch": 0.29889356372916376,
      "grad_norm": 9.516846656799316,
      "learning_rate": 4.627395821036672e-06,
      "loss": 0.4815,
      "mean_token_accuracy": 0.853526496887207,
      "num_tokens": 115938787.0,
      "step": 96420
    },
    {
      "entropy": 1.9189068511128426,
      "epoch": 0.2989245628542134,
      "grad_norm": 7.334571361541748,
      "learning_rate": 4.6271558768395816e-06,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8488484725356102,
      "num_tokens": 115950064.0,
      "step": 96430
    },
    {
      "entropy": 1.8531944811344148,
      "epoch": 0.29895556197926315,
      "grad_norm": 8.624135971069336,
      "learning_rate": 4.6269159699640755e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8516410380601883,
      "num_tokens": 115962309.0,
      "step": 96440
    },
    {
      "entropy": 1.903770998120308,
      "epoch": 0.2989865611043128,
      "grad_norm": 8.187854766845703,
      "learning_rate": 4.62667610040048e-06,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8519768640398979,
      "num_tokens": 115973845.0,
      "step": 96450
    },
    {
      "entropy": 1.8115949898958206,
      "epoch": 0.29901756022936254,
      "grad_norm": 8.5139799118042,
      "learning_rate": 4.626436268139122e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8568135395646095,
      "num_tokens": 115986832.0,
      "step": 96460
    },
    {
      "entropy": 1.979418794810772,
      "epoch": 0.2990485593544122,
      "grad_norm": 7.29787015914917,
      "learning_rate": 4.626196473170338e-06,
      "loss": 0.5356,
      "mean_token_accuracy": 0.8345787361264229,
      "num_tokens": 115997910.0,
      "step": 96470
    },
    {
      "entropy": 1.9397563070058823,
      "epoch": 0.29907955847946194,
      "grad_norm": 8.139474868774414,
      "learning_rate": 4.625956715484463e-06,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8472580030560494,
      "num_tokens": 116009904.0,
      "step": 96480
    },
    {
      "entropy": 1.8948927074670792,
      "epoch": 0.2991105576045116,
      "grad_norm": 9.053715705871582,
      "learning_rate": 4.625716995071836e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8533863142132759,
      "num_tokens": 116021106.0,
      "step": 96490
    },
    {
      "entropy": 1.8922299653291703,
      "epoch": 0.29914155672956133,
      "grad_norm": 9.0900239944458,
      "learning_rate": 4.6254773119228004e-06,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8474109992384911,
      "num_tokens": 116033322.0,
      "step": 96500
    },
    {
      "entropy": 1.847077339887619,
      "epoch": 0.299172555854611,
      "grad_norm": 3.035865068435669,
      "learning_rate": 4.625237666027704e-06,
      "loss": 0.4026,
      "mean_token_accuracy": 0.8617709457874299,
      "num_tokens": 116045141.0,
      "step": 96510
    },
    {
      "entropy": 1.8434469774365425,
      "epoch": 0.2992035549796607,
      "grad_norm": 10.367220878601074,
      "learning_rate": 4.624998057376896e-06,
      "loss": 0.43,
      "mean_token_accuracy": 0.859175056219101,
      "num_tokens": 116057311.0,
      "step": 96520
    },
    {
      "entropy": 1.922757549583912,
      "epoch": 0.2992345541047104,
      "grad_norm": 7.6684417724609375,
      "learning_rate": 4.624758485960731e-06,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8412060752511025,
      "num_tokens": 116068781.0,
      "step": 96530
    },
    {
      "entropy": 1.874141050875187,
      "epoch": 0.2992655532297601,
      "grad_norm": 5.786795139312744,
      "learning_rate": 4.624518951769568e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8453152433037758,
      "num_tokens": 116080384.0,
      "step": 96540
    },
    {
      "entropy": 1.9315826326608658,
      "epoch": 0.2992965523548098,
      "grad_norm": 7.588706016540527,
      "learning_rate": 4.624279454793765e-06,
      "loss": 0.5078,
      "mean_token_accuracy": 0.84673622995615,
      "num_tokens": 116090980.0,
      "step": 96550
    },
    {
      "entropy": 1.910808216035366,
      "epoch": 0.29932755147985945,
      "grad_norm": 4.097476482391357,
      "learning_rate": 4.624039995023688e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8506806045770645,
      "num_tokens": 116102714.0,
      "step": 96560
    },
    {
      "entropy": 1.8671618595719337,
      "epoch": 0.2993585506049092,
      "grad_norm": 4.14247465133667,
      "learning_rate": 4.623800572449704e-06,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8386751100420952,
      "num_tokens": 116114873.0,
      "step": 96570
    },
    {
      "entropy": 1.8730577558279038,
      "epoch": 0.29938954972995885,
      "grad_norm": 11.251909255981445,
      "learning_rate": 4.623561187062184e-06,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8561120167374611,
      "num_tokens": 116126620.0,
      "step": 96580
    },
    {
      "entropy": 1.8773257568478585,
      "epoch": 0.2994205488550086,
      "grad_norm": 9.164704322814941,
      "learning_rate": 4.623321838851505e-06,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8373346477746964,
      "num_tokens": 116137916.0,
      "step": 96590
    },
    {
      "entropy": 1.9126343131065369,
      "epoch": 0.29945154798005824,
      "grad_norm": 8.017783164978027,
      "learning_rate": 4.623082527808043e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.841637770831585,
      "num_tokens": 116149896.0,
      "step": 96600
    },
    {
      "entropy": 1.8838177442550659,
      "epoch": 0.29948254710510797,
      "grad_norm": 4.913340091705322,
      "learning_rate": 4.622843253922182e-06,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8604057088494301,
      "num_tokens": 116162580.0,
      "step": 96610
    },
    {
      "entropy": 1.8277122244238853,
      "epoch": 0.29951354623015763,
      "grad_norm": 8.11374282836914,
      "learning_rate": 4.622604017184304e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8429702982306481,
      "num_tokens": 116175332.0,
      "step": 96620
    },
    {
      "entropy": 1.9389643400907517,
      "epoch": 0.29954454535520736,
      "grad_norm": 8.148730278015137,
      "learning_rate": 4.622364817584801e-06,
      "loss": 0.541,
      "mean_token_accuracy": 0.8371641963720322,
      "num_tokens": 116186424.0,
      "step": 96630
    },
    {
      "entropy": 1.9510870546102523,
      "epoch": 0.29957554448025703,
      "grad_norm": 6.960958957672119,
      "learning_rate": 4.622125655114065e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.839006906747818,
      "num_tokens": 116197799.0,
      "step": 96640
    },
    {
      "entropy": 1.9510822862386703,
      "epoch": 0.29960654360530675,
      "grad_norm": 8.637950897216797,
      "learning_rate": 4.621886529762488e-06,
      "loss": 0.5395,
      "mean_token_accuracy": 0.8410506933927536,
      "num_tokens": 116208985.0,
      "step": 96650
    },
    {
      "entropy": 1.868352036178112,
      "epoch": 0.2996375427303564,
      "grad_norm": 10.132604598999023,
      "learning_rate": 4.621647441520475e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8492979109287262,
      "num_tokens": 116221911.0,
      "step": 96660
    },
    {
      "entropy": 1.9646795377135278,
      "epoch": 0.29966854185540615,
      "grad_norm": 9.418262481689453,
      "learning_rate": 4.621408390378424e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8449583351612091,
      "num_tokens": 116232932.0,
      "step": 96670
    },
    {
      "entropy": 1.9498365700244904,
      "epoch": 0.2996995409804558,
      "grad_norm": 7.607828617095947,
      "learning_rate": 4.621169376326742e-06,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8415804550051689,
      "num_tokens": 116243576.0,
      "step": 96680
    },
    {
      "entropy": 1.862950399518013,
      "epoch": 0.29973054010550554,
      "grad_norm": 8.767817497253418,
      "learning_rate": 4.620930399355841e-06,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8488745912909508,
      "num_tokens": 116256004.0,
      "step": 96690
    },
    {
      "entropy": 1.795606505870819,
      "epoch": 0.2997615392305552,
      "grad_norm": 4.170324325561523,
      "learning_rate": 4.620691459456132e-06,
      "loss": 0.4269,
      "mean_token_accuracy": 0.8523398399353027,
      "num_tokens": 116269297.0,
      "step": 96700
    },
    {
      "entropy": 1.8912088066339492,
      "epoch": 0.29979253835560493,
      "grad_norm": 7.616124153137207,
      "learning_rate": 4.620452556618031e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8536741435527802,
      "num_tokens": 116281401.0,
      "step": 96710
    },
    {
      "entropy": 1.9044504195451737,
      "epoch": 0.2998235374806546,
      "grad_norm": 3.9353320598602295,
      "learning_rate": 4.6202136908319606e-06,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8428740501403809,
      "num_tokens": 116293236.0,
      "step": 96720
    },
    {
      "entropy": 1.831810677051544,
      "epoch": 0.2998545366057043,
      "grad_norm": 9.96127986907959,
      "learning_rate": 4.6199748620883425e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.856151320040226,
      "num_tokens": 116305330.0,
      "step": 96730
    },
    {
      "entropy": 1.915688943862915,
      "epoch": 0.299885535730754,
      "grad_norm": 8.897388458251953,
      "learning_rate": 4.619736070377604e-06,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8483431145548821,
      "num_tokens": 116316018.0,
      "step": 96740
    },
    {
      "entropy": 1.8551687330007554,
      "epoch": 0.2999165348558037,
      "grad_norm": 3.1668596267700195,
      "learning_rate": 4.619497315690176e-06,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8362834542989731,
      "num_tokens": 116328555.0,
      "step": 96750
    },
    {
      "entropy": 1.8770240753889085,
      "epoch": 0.2999475339808534,
      "grad_norm": 2.5352485179901123,
      "learning_rate": 4.61925859801649e-06,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8435857251286507,
      "num_tokens": 116340178.0,
      "step": 96760
    },
    {
      "entropy": 1.9596908926963805,
      "epoch": 0.2999785331059031,
      "grad_norm": 8.359101295471191,
      "learning_rate": 4.619019917346987e-06,
      "loss": 0.5595,
      "mean_token_accuracy": 0.8392258763313294,
      "num_tokens": 116351512.0,
      "step": 96770
    },
    {
      "entropy": 1.7745133236050605,
      "epoch": 0.3000095322309528,
      "grad_norm": 6.980304718017578,
      "learning_rate": 4.618781273672105e-06,
      "loss": 0.3915,
      "mean_token_accuracy": 0.8550699412822723,
      "num_tokens": 116365493.0,
      "step": 96780
    },
    {
      "entropy": 1.8009425699710846,
      "epoch": 0.3000405313560025,
      "grad_norm": 9.106100082397461,
      "learning_rate": 4.618542666982291e-06,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8583745554089546,
      "num_tokens": 116378953.0,
      "step": 96790
    },
    {
      "entropy": 1.8405057787895203,
      "epoch": 0.3000715304810522,
      "grad_norm": 6.269810199737549,
      "learning_rate": 4.6183040972679905e-06,
      "loss": 0.455,
      "mean_token_accuracy": 0.853019006550312,
      "num_tokens": 116391720.0,
      "step": 96800
    },
    {
      "entropy": 1.8243362039327622,
      "epoch": 0.30010252960610184,
      "grad_norm": 8.785292625427246,
      "learning_rate": 4.618065564519655e-06,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8519068777561187,
      "num_tokens": 116404782.0,
      "step": 96810
    },
    {
      "entropy": 1.9162081688642503,
      "epoch": 0.30013352873115157,
      "grad_norm": 9.158159255981445,
      "learning_rate": 4.617827068727739e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8442169070243836,
      "num_tokens": 116415863.0,
      "step": 96820
    },
    {
      "entropy": 1.8956199631094932,
      "epoch": 0.30016452785620124,
      "grad_norm": 5.014525890350342,
      "learning_rate": 4.617588609882702e-06,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8389386609196663,
      "num_tokens": 116427960.0,
      "step": 96830
    },
    {
      "entropy": 1.890786738693714,
      "epoch": 0.30019552698125096,
      "grad_norm": 10.428237915039062,
      "learning_rate": 4.617350187975004e-06,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8487927287817001,
      "num_tokens": 116439841.0,
      "step": 96840
    },
    {
      "entropy": 1.8615357890725135,
      "epoch": 0.30022652610630063,
      "grad_norm": 9.423101425170898,
      "learning_rate": 4.617111802995109e-06,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8472002789378166,
      "num_tokens": 116451475.0,
      "step": 96850
    },
    {
      "entropy": 1.9217775925993918,
      "epoch": 0.30025752523135035,
      "grad_norm": 7.137108325958252,
      "learning_rate": 4.616873454933489e-06,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8377560988068581,
      "num_tokens": 116462859.0,
      "step": 96860
    },
    {
      "entropy": 1.8776028633117676,
      "epoch": 0.3002885243564,
      "grad_norm": 4.520118236541748,
      "learning_rate": 4.616635143780614e-06,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8304355323314667,
      "num_tokens": 116475060.0,
      "step": 96870
    },
    {
      "entropy": 1.8731580957770348,
      "epoch": 0.30031952348144975,
      "grad_norm": 7.3886942863464355,
      "learning_rate": 4.616396869526958e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8490657389163971,
      "num_tokens": 116487261.0,
      "step": 96880
    },
    {
      "entropy": 1.8840817973017692,
      "epoch": 0.3003505226064994,
      "grad_norm": 8.366750717163086,
      "learning_rate": 4.616158632163e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.857377803325653,
      "num_tokens": 116498999.0,
      "step": 96890
    },
    {
      "entropy": 1.8003995344042778,
      "epoch": 0.30038152173154914,
      "grad_norm": 2.8149616718292236,
      "learning_rate": 4.615920431679226e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8655429676175117,
      "num_tokens": 116512337.0,
      "step": 96900
    },
    {
      "entropy": 1.8423293381929398,
      "epoch": 0.3004125208565988,
      "grad_norm": 3.726378917694092,
      "learning_rate": 4.615682268066116e-06,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8466761097311973,
      "num_tokens": 116525215.0,
      "step": 96910
    },
    {
      "entropy": 1.8393646717071532,
      "epoch": 0.30044351998164853,
      "grad_norm": 9.186988830566406,
      "learning_rate": 4.615444141314163e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8458735197782516,
      "num_tokens": 116537163.0,
      "step": 96920
    },
    {
      "entropy": 1.9296279013156892,
      "epoch": 0.3004745191066982,
      "grad_norm": 9.516865730285645,
      "learning_rate": 4.615206051413857e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8410545736551285,
      "num_tokens": 116548242.0,
      "step": 96930
    },
    {
      "entropy": 1.867045633494854,
      "epoch": 0.3005055182317479,
      "grad_norm": 7.079854965209961,
      "learning_rate": 4.614967998355696e-06,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8517505764961243,
      "num_tokens": 116560243.0,
      "step": 96940
    },
    {
      "entropy": 1.9103495597839355,
      "epoch": 0.3005365173567976,
      "grad_norm": 5.358558177947998,
      "learning_rate": 4.614729982130179e-06,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8406784102320671,
      "num_tokens": 116571891.0,
      "step": 96950
    },
    {
      "entropy": 1.9171627387404442,
      "epoch": 0.3005675164818473,
      "grad_norm": 8.190808296203613,
      "learning_rate": 4.614492002727808e-06,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8321561366319656,
      "num_tokens": 116583993.0,
      "step": 96960
    },
    {
      "entropy": 1.84826198220253,
      "epoch": 0.300598515606897,
      "grad_norm": 9.064087867736816,
      "learning_rate": 4.61425406013909e-06,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8504088371992111,
      "num_tokens": 116596878.0,
      "step": 96970
    },
    {
      "entropy": 1.941984808444977,
      "epoch": 0.3006295147319467,
      "grad_norm": 4.682338237762451,
      "learning_rate": 4.614016154354533e-06,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8427264273166657,
      "num_tokens": 116608036.0,
      "step": 96980
    },
    {
      "entropy": 1.9229829460382462,
      "epoch": 0.3006605138569964,
      "grad_norm": 7.0781097412109375,
      "learning_rate": 4.6137782853646524e-06,
      "loss": 0.5264,
      "mean_token_accuracy": 0.839740289747715,
      "num_tokens": 116619904.0,
      "step": 96990
    },
    {
      "entropy": 1.8459256619215012,
      "epoch": 0.3006915129820461,
      "grad_norm": 7.523281097412109,
      "learning_rate": 4.613540453159963e-06,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8505584686994553,
      "num_tokens": 116632261.0,
      "step": 97000
    },
    {
      "entropy": 1.8035550713539124,
      "epoch": 0.3007225121070958,
      "grad_norm": 5.279726982116699,
      "learning_rate": 4.613302657730985e-06,
      "loss": 0.4401,
      "mean_token_accuracy": 0.854595598578453,
      "num_tokens": 116645315.0,
      "step": 97010
    },
    {
      "entropy": 1.907500149309635,
      "epoch": 0.3007535112321455,
      "grad_norm": 3.997288227081299,
      "learning_rate": 4.613064899068243e-06,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8380198463797569,
      "num_tokens": 116657205.0,
      "step": 97020
    },
    {
      "entropy": 1.815874347090721,
      "epoch": 0.30078451035719517,
      "grad_norm": 3.8312273025512695,
      "learning_rate": 4.612827177162262e-06,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8552998825907707,
      "num_tokens": 116669443.0,
      "step": 97030
    },
    {
      "entropy": 1.93814327865839,
      "epoch": 0.30081550948224484,
      "grad_norm": 9.422027587890625,
      "learning_rate": 4.612589492003573e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.840138903260231,
      "num_tokens": 116681521.0,
      "step": 97040
    },
    {
      "entropy": 1.9181523829698564,
      "epoch": 0.30084650860729456,
      "grad_norm": 7.659515380859375,
      "learning_rate": 4.6123518435827095e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8537562400102615,
      "num_tokens": 116692024.0,
      "step": 97050
    },
    {
      "entropy": 1.8777270019054413,
      "epoch": 0.30087750773234423,
      "grad_norm": 4.077333450317383,
      "learning_rate": 4.612114231890209e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8504292860627174,
      "num_tokens": 116703672.0,
      "step": 97060
    },
    {
      "entropy": 1.8734031990170479,
      "epoch": 0.30090850685739395,
      "grad_norm": 8.679883003234863,
      "learning_rate": 4.61187665691661e-06,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8424584448337555,
      "num_tokens": 116715061.0,
      "step": 97070
    },
    {
      "entropy": 1.919617336988449,
      "epoch": 0.3009395059824436,
      "grad_norm": 9.382501602172852,
      "learning_rate": 4.611639118652459e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8328076407313347,
      "num_tokens": 116726486.0,
      "step": 97080
    },
    {
      "entropy": 1.890953540802002,
      "epoch": 0.30097050510749335,
      "grad_norm": 10.713085174560547,
      "learning_rate": 4.611401617088301e-06,
      "loss": 0.563,
      "mean_token_accuracy": 0.8449400596320629,
      "num_tokens": 116738906.0,
      "step": 97090
    },
    {
      "entropy": 1.8357614412903787,
      "epoch": 0.301001504232543,
      "grad_norm": 8.783710479736328,
      "learning_rate": 4.611164152214689e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8468549698591232,
      "num_tokens": 116751196.0,
      "step": 97100
    },
    {
      "entropy": 1.9422936275601388,
      "epoch": 0.30103250335759274,
      "grad_norm": 8.501259803771973,
      "learning_rate": 4.6109267240221755e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8422497108578682,
      "num_tokens": 116763810.0,
      "step": 97110
    },
    {
      "entropy": 1.9155998945236206,
      "epoch": 0.3010635024826424,
      "grad_norm": 7.944989204406738,
      "learning_rate": 4.610689332501317e-06,
      "loss": 0.5321,
      "mean_token_accuracy": 0.823591648042202,
      "num_tokens": 116776024.0,
      "step": 97120
    },
    {
      "entropy": 1.8754037857055663,
      "epoch": 0.30109450160769213,
      "grad_norm": 7.605261325836182,
      "learning_rate": 4.610451977642677e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8444375693798065,
      "num_tokens": 116787890.0,
      "step": 97130
    },
    {
      "entropy": 1.8820706829428673,
      "epoch": 0.3011255007327418,
      "grad_norm": 4.352447509765625,
      "learning_rate": 4.610214659436818e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8387032672762871,
      "num_tokens": 116799997.0,
      "step": 97140
    },
    {
      "entropy": 1.8734632670879363,
      "epoch": 0.3011564998577915,
      "grad_norm": 7.570825576782227,
      "learning_rate": 4.609977377874307e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8414079815149307,
      "num_tokens": 116812400.0,
      "step": 97150
    },
    {
      "entropy": 1.9303068399429322,
      "epoch": 0.3011874989828412,
      "grad_norm": 3.602426528930664,
      "learning_rate": 4.609740132945716e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8503539338707924,
      "num_tokens": 116825076.0,
      "step": 97160
    },
    {
      "entropy": 1.8318970784544946,
      "epoch": 0.3012184981078909,
      "grad_norm": 9.406216621398926,
      "learning_rate": 4.609502924641619e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8480072125792504,
      "num_tokens": 116837724.0,
      "step": 97170
    },
    {
      "entropy": 1.9633089289069177,
      "epoch": 0.3012494972329406,
      "grad_norm": 8.81785774230957,
      "learning_rate": 4.609265752952596e-06,
      "loss": 0.543,
      "mean_token_accuracy": 0.8268476709723472,
      "num_tokens": 116849107.0,
      "step": 97180
    },
    {
      "entropy": 1.8834416687488555,
      "epoch": 0.3012804963579903,
      "grad_norm": 4.704411506652832,
      "learning_rate": 4.609028617869224e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8506078109145164,
      "num_tokens": 116861833.0,
      "step": 97190
    },
    {
      "entropy": 1.7407511696219444,
      "epoch": 0.30131149548304,
      "grad_norm": 2.5168025493621826,
      "learning_rate": 4.6087915193820916e-06,
      "loss": 0.3506,
      "mean_token_accuracy": 0.8639174044132233,
      "num_tokens": 116876452.0,
      "step": 97200
    },
    {
      "entropy": 1.8649092674255372,
      "epoch": 0.3013424946080897,
      "grad_norm": 3.976242780685425,
      "learning_rate": 4.608554457481785e-06,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8543577790260315,
      "num_tokens": 116888349.0,
      "step": 97210
    },
    {
      "entropy": 1.850515154004097,
      "epoch": 0.3013734937331394,
      "grad_norm": 4.014977931976318,
      "learning_rate": 4.608317432158896e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8551569744944573,
      "num_tokens": 116900337.0,
      "step": 97220
    },
    {
      "entropy": 1.904152835905552,
      "epoch": 0.3014044928581891,
      "grad_norm": 7.234251022338867,
      "learning_rate": 4.60808044340402e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8568873882293702,
      "num_tokens": 116912573.0,
      "step": 97230
    },
    {
      "entropy": 1.8178890123963356,
      "epoch": 0.30143549198323877,
      "grad_norm": 2.6580164432525635,
      "learning_rate": 4.607843491207752e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8546398520469666,
      "num_tokens": 116924977.0,
      "step": 97240
    },
    {
      "entropy": 1.9117593422532082,
      "epoch": 0.3014664911082885,
      "grad_norm": 13.091536521911621,
      "learning_rate": 4.607606575560697e-06,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8411953374743462,
      "num_tokens": 116936606.0,
      "step": 97250
    },
    {
      "entropy": 1.9153368800878525,
      "epoch": 0.30149749023333816,
      "grad_norm": 10.279766082763672,
      "learning_rate": 4.607369696453461e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8421198353171349,
      "num_tokens": 116948020.0,
      "step": 97260
    },
    {
      "entropy": 1.892761492729187,
      "epoch": 0.3015284893583879,
      "grad_norm": 8.145535469055176,
      "learning_rate": 4.6071328538766486e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8441890180110931,
      "num_tokens": 116959806.0,
      "step": 97270
    },
    {
      "entropy": 1.9218065708875656,
      "epoch": 0.30155948848343755,
      "grad_norm": 7.536378860473633,
      "learning_rate": 4.606896047820874e-06,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8465237602591514,
      "num_tokens": 116970254.0,
      "step": 97280
    },
    {
      "entropy": 1.8478126615285873,
      "epoch": 0.3015904876084872,
      "grad_norm": 8.085580825805664,
      "learning_rate": 4.60665927827675e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8454589918255806,
      "num_tokens": 116981953.0,
      "step": 97290
    },
    {
      "entropy": 1.7839539676904679,
      "epoch": 0.30162148673353695,
      "grad_norm": 4.7135491371154785,
      "learning_rate": 4.606422545234899e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8522520795464515,
      "num_tokens": 116994789.0,
      "step": 97300
    },
    {
      "entropy": 1.7900249511003494,
      "epoch": 0.3016524858585866,
      "grad_norm": 12.013949394226074,
      "learning_rate": 4.606185848685939e-06,
      "loss": 0.4124,
      "mean_token_accuracy": 0.8632920622825623,
      "num_tokens": 117007458.0,
      "step": 97310
    },
    {
      "entropy": 1.8314153790473937,
      "epoch": 0.30168348498363634,
      "grad_norm": 8.588581085205078,
      "learning_rate": 4.605949188620496e-06,
      "loss": 0.454,
      "mean_token_accuracy": 0.8470736399292946,
      "num_tokens": 117019927.0,
      "step": 97320
    },
    {
      "entropy": 1.8574935659766196,
      "epoch": 0.301714484108686,
      "grad_norm": 8.47958755493164,
      "learning_rate": 4.6057125650292e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8512878254055977,
      "num_tokens": 117031729.0,
      "step": 97330
    },
    {
      "entropy": 1.9198235914111137,
      "epoch": 0.30174548323373573,
      "grad_norm": 10.255192756652832,
      "learning_rate": 4.605475977902682e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8471052810549736,
      "num_tokens": 117042862.0,
      "step": 97340
    },
    {
      "entropy": 1.8835196584463119,
      "epoch": 0.3017764823587854,
      "grad_norm": 9.062085151672363,
      "learning_rate": 4.605239427231577e-06,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8494998052716255,
      "num_tokens": 117054126.0,
      "step": 97350
    },
    {
      "entropy": 1.9018598824739457,
      "epoch": 0.3018074814838351,
      "grad_norm": 8.304078102111816,
      "learning_rate": 4.6050029130065245e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8413213551044464,
      "num_tokens": 117065364.0,
      "step": 97360
    },
    {
      "entropy": 1.8747293829917908,
      "epoch": 0.3018384806088848,
      "grad_norm": 8.438359260559082,
      "learning_rate": 4.604766435218166e-06,
      "loss": 0.498,
      "mean_token_accuracy": 0.8443131268024444,
      "num_tokens": 117076938.0,
      "step": 97370
    },
    {
      "entropy": 1.8413864582777024,
      "epoch": 0.3018694797339345,
      "grad_norm": 8.747175216674805,
      "learning_rate": 4.604529993857147e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8414786517620086,
      "num_tokens": 117089574.0,
      "step": 97380
    },
    {
      "entropy": 1.9370706051588058,
      "epoch": 0.3019004788589842,
      "grad_norm": 7.200348854064941,
      "learning_rate": 4.604293588914116e-06,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8398382663726807,
      "num_tokens": 117100625.0,
      "step": 97390
    },
    {
      "entropy": 1.850702230632305,
      "epoch": 0.3019314779840339,
      "grad_norm": 8.624340057373047,
      "learning_rate": 4.604057220379726e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8501965671777725,
      "num_tokens": 117113376.0,
      "step": 97400
    },
    {
      "entropy": 1.8446269080042839,
      "epoch": 0.3019624771090836,
      "grad_norm": 3.699392318725586,
      "learning_rate": 4.603820888244632e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8584934890270233,
      "num_tokens": 117126682.0,
      "step": 97410
    },
    {
      "entropy": 1.943909691274166,
      "epoch": 0.3019934762341333,
      "grad_norm": 3.7331109046936035,
      "learning_rate": 4.603584592499492e-06,
      "loss": 0.569,
      "mean_token_accuracy": 0.8290564298629761,
      "num_tokens": 117138961.0,
      "step": 97420
    },
    {
      "entropy": 1.8698807314038277,
      "epoch": 0.302024475359183,
      "grad_norm": 8.145543098449707,
      "learning_rate": 4.603348333134969e-06,
      "loss": 0.466,
      "mean_token_accuracy": 0.8512676984071732,
      "num_tokens": 117150754.0,
      "step": 97430
    },
    {
      "entropy": 1.9144811987876893,
      "epoch": 0.3020554744842327,
      "grad_norm": 9.019363403320312,
      "learning_rate": 4.60311211014173e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8448151677846909,
      "num_tokens": 117162759.0,
      "step": 97440
    },
    {
      "entropy": 1.8737313896417618,
      "epoch": 0.30208647360928237,
      "grad_norm": 7.63023567199707,
      "learning_rate": 4.602875923510441e-06,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8467231750488281,
      "num_tokens": 117174665.0,
      "step": 97450
    },
    {
      "entropy": 1.8207977265119553,
      "epoch": 0.3021174727343321,
      "grad_norm": 8.674139976501465,
      "learning_rate": 4.602639773231776e-06,
      "loss": 0.471,
      "mean_token_accuracy": 0.8548212826251984,
      "num_tokens": 117187809.0,
      "step": 97460
    },
    {
      "entropy": 1.8152425453066825,
      "epoch": 0.30214847185938176,
      "grad_norm": 5.675132751464844,
      "learning_rate": 4.602403659296411e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8367105752229691,
      "num_tokens": 117201085.0,
      "step": 97470
    },
    {
      "entropy": 1.857864636182785,
      "epoch": 0.3021794709844315,
      "grad_norm": 8.102805137634277,
      "learning_rate": 4.602167581695023e-06,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8479077383875847,
      "num_tokens": 117213732.0,
      "step": 97480
    },
    {
      "entropy": 1.925594538450241,
      "epoch": 0.30221047010948116,
      "grad_norm": 7.970809459686279,
      "learning_rate": 4.601931540418297e-06,
      "loss": 0.523,
      "mean_token_accuracy": 0.8393017098307609,
      "num_tokens": 117224935.0,
      "step": 97490
    },
    {
      "entropy": 1.86695496737957,
      "epoch": 0.3022414692345309,
      "grad_norm": 9.16865348815918,
      "learning_rate": 4.601695535456917e-06,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8479771926999092,
      "num_tokens": 117236716.0,
      "step": 97500
    },
    {
      "entropy": 1.9163902431726456,
      "epoch": 0.30227246835958055,
      "grad_norm": 8.983051300048828,
      "learning_rate": 4.601459566801571e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8492978289723396,
      "num_tokens": 117247883.0,
      "step": 97510
    },
    {
      "entropy": 1.8202176943421364,
      "epoch": 0.3023034674846303,
      "grad_norm": 7.500359535217285,
      "learning_rate": 4.601223634442954e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8624223947525025,
      "num_tokens": 117261178.0,
      "step": 97520
    },
    {
      "entropy": 1.9173918321728707,
      "epoch": 0.30233446660967994,
      "grad_norm": 8.454766273498535,
      "learning_rate": 4.600987738371759e-06,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8437489971518517,
      "num_tokens": 117273426.0,
      "step": 97530
    },
    {
      "entropy": 1.77590219527483,
      "epoch": 0.3023654657347296,
      "grad_norm": 8.413315773010254,
      "learning_rate": 4.600751878578687e-06,
      "loss": 0.3971,
      "mean_token_accuracy": 0.8557587161660194,
      "num_tokens": 117286603.0,
      "step": 97540
    },
    {
      "entropy": 1.9176592335104943,
      "epoch": 0.30239646485977933,
      "grad_norm": 10.7260103225708,
      "learning_rate": 4.600516055054439e-06,
      "loss": 0.5096,
      "mean_token_accuracy": 0.8402794227004051,
      "num_tokens": 117297972.0,
      "step": 97550
    },
    {
      "entropy": 1.8973240569233893,
      "epoch": 0.302427463984829,
      "grad_norm": 4.1420087814331055,
      "learning_rate": 4.600280267789722e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8584436431527138,
      "num_tokens": 117309456.0,
      "step": 97560
    },
    {
      "entropy": 1.9296040430665016,
      "epoch": 0.30245846310987873,
      "grad_norm": 8.410199165344238,
      "learning_rate": 4.600044516775245e-06,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8443345412611961,
      "num_tokens": 117321272.0,
      "step": 97570
    },
    {
      "entropy": 1.9159424543380736,
      "epoch": 0.3024894622349284,
      "grad_norm": 8.414966583251953,
      "learning_rate": 4.5998088020017186e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8412731990218163,
      "num_tokens": 117333250.0,
      "step": 97580
    },
    {
      "entropy": 1.7769809067249298,
      "epoch": 0.3025204613599781,
      "grad_norm": 3.658212661743164,
      "learning_rate": 4.599573123459859e-06,
      "loss": 0.38,
      "mean_token_accuracy": 0.8634654805064201,
      "num_tokens": 117346443.0,
      "step": 97590
    },
    {
      "entropy": 1.8901727512478828,
      "epoch": 0.3025514604850278,
      "grad_norm": 9.50108814239502,
      "learning_rate": 4.599337481140387e-06,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8505553930997849,
      "num_tokens": 117358066.0,
      "step": 97600
    },
    {
      "entropy": 1.8625748187303544,
      "epoch": 0.3025824596100775,
      "grad_norm": 8.508552551269531,
      "learning_rate": 4.5991018750340235e-06,
      "loss": 0.4556,
      "mean_token_accuracy": 0.849023912847042,
      "num_tokens": 117369805.0,
      "step": 97610
    },
    {
      "entropy": 1.9069266065955162,
      "epoch": 0.3026134587351272,
      "grad_norm": 8.270326614379883,
      "learning_rate": 4.5988663051314944e-06,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8539838179945946,
      "num_tokens": 117381262.0,
      "step": 97620
    },
    {
      "entropy": 1.9128611549735068,
      "epoch": 0.3026444578601769,
      "grad_norm": 8.971906661987305,
      "learning_rate": 4.5986307714235286e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.832034258544445,
      "num_tokens": 117393179.0,
      "step": 97630
    },
    {
      "entropy": 1.8143539875745773,
      "epoch": 0.3026754569852266,
      "grad_norm": 7.190959453582764,
      "learning_rate": 4.5983952739008585e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8518998950719834,
      "num_tokens": 117406161.0,
      "step": 97640
    },
    {
      "entropy": 1.8843399420380593,
      "epoch": 0.3027064561102763,
      "grad_norm": 9.908082008361816,
      "learning_rate": 4.598159812554219e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8463353976607323,
      "num_tokens": 117418576.0,
      "step": 97650
    },
    {
      "entropy": 1.8831077113747596,
      "epoch": 0.30273745523532597,
      "grad_norm": 9.612486839294434,
      "learning_rate": 4.597924387374351e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8505209341645241,
      "num_tokens": 117430122.0,
      "step": 97660
    },
    {
      "entropy": 1.8462072387337685,
      "epoch": 0.3027684543603757,
      "grad_norm": 4.665594100952148,
      "learning_rate": 4.597688998351995e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8621034786105156,
      "num_tokens": 117442896.0,
      "step": 97670
    },
    {
      "entropy": 1.9408204436302186,
      "epoch": 0.30279945348542536,
      "grad_norm": 7.827560901641846,
      "learning_rate": 4.597453645477898e-06,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8394302636384964,
      "num_tokens": 117453977.0,
      "step": 97680
    },
    {
      "entropy": 1.8475888326764107,
      "epoch": 0.3028304526104751,
      "grad_norm": 8.166518211364746,
      "learning_rate": 4.597218328742807e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8580135375261306,
      "num_tokens": 117466325.0,
      "step": 97690
    },
    {
      "entropy": 1.8173963025212287,
      "epoch": 0.30286145173552476,
      "grad_norm": 4.001589775085449,
      "learning_rate": 4.596983048137475e-06,
      "loss": 0.4122,
      "mean_token_accuracy": 0.8486342668533325,
      "num_tokens": 117478782.0,
      "step": 97700
    },
    {
      "entropy": 1.8551449865102767,
      "epoch": 0.3028924508605745,
      "grad_norm": 8.81025218963623,
      "learning_rate": 4.596747803652658e-06,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8483422353863717,
      "num_tokens": 117491302.0,
      "step": 97710
    },
    {
      "entropy": 1.8700275629758836,
      "epoch": 0.30292344998562415,
      "grad_norm": 3.273538112640381,
      "learning_rate": 4.596512595279115e-06,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8363815456628799,
      "num_tokens": 117504377.0,
      "step": 97720
    },
    {
      "entropy": 1.9091096714138984,
      "epoch": 0.3029544491106739,
      "grad_norm": 4.058509349822998,
      "learning_rate": 4.5962774230076075e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8540859907865525,
      "num_tokens": 117515754.0,
      "step": 97730
    },
    {
      "entropy": 1.979654061794281,
      "epoch": 0.30298544823572354,
      "grad_norm": 8.739022254943848,
      "learning_rate": 4.596042286828902e-06,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8428146079182625,
      "num_tokens": 117526416.0,
      "step": 97740
    },
    {
      "entropy": 1.9291452512145042,
      "epoch": 0.30301644736077327,
      "grad_norm": 8.007489204406738,
      "learning_rate": 4.595807186733767e-06,
      "loss": 0.5136,
      "mean_token_accuracy": 0.8436457946896553,
      "num_tokens": 117538034.0,
      "step": 97750
    },
    {
      "entropy": 1.8014676854014398,
      "epoch": 0.30304744648582294,
      "grad_norm": 9.48009967803955,
      "learning_rate": 4.595572122712974e-06,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8604965686798096,
      "num_tokens": 117551372.0,
      "step": 97760
    },
    {
      "entropy": 1.8575839295983314,
      "epoch": 0.30307844561087266,
      "grad_norm": 2.4057457447052,
      "learning_rate": 4.595337094757297e-06,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8554554954171181,
      "num_tokens": 117563424.0,
      "step": 97770
    },
    {
      "entropy": 1.8977721393108369,
      "epoch": 0.30310944473592233,
      "grad_norm": 9.092135429382324,
      "learning_rate": 4.595102102857518e-06,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8410274058580398,
      "num_tokens": 117575055.0,
      "step": 97780
    },
    {
      "entropy": 1.89531751871109,
      "epoch": 0.303140443860972,
      "grad_norm": 8.363005638122559,
      "learning_rate": 4.594867147004416e-06,
      "loss": 0.484,
      "mean_token_accuracy": 0.8437793105840683,
      "num_tokens": 117586328.0,
      "step": 97790
    },
    {
      "entropy": 1.8966928854584695,
      "epoch": 0.3031714429860217,
      "grad_norm": 8.183049201965332,
      "learning_rate": 4.594632227188778e-06,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8496061801910401,
      "num_tokens": 117598053.0,
      "step": 97800
    },
    {
      "entropy": 1.7953755557537079,
      "epoch": 0.3032024421110714,
      "grad_norm": 7.209178924560547,
      "learning_rate": 4.594397343401393e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8673543766140938,
      "num_tokens": 117610778.0,
      "step": 97810
    },
    {
      "entropy": 1.8411307245492936,
      "epoch": 0.3032334412361211,
      "grad_norm": 4.28125,
      "learning_rate": 4.59416249563305e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8453304886817932,
      "num_tokens": 117623608.0,
      "step": 97820
    },
    {
      "entropy": 1.849568995833397,
      "epoch": 0.3032644403611708,
      "grad_norm": 7.9680023193359375,
      "learning_rate": 4.593927683874549e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8446632474660873,
      "num_tokens": 117635895.0,
      "step": 97830
    },
    {
      "entropy": 1.9411183446645737,
      "epoch": 0.3032954394862205,
      "grad_norm": 8.592232704162598,
      "learning_rate": 4.593692908116683e-06,
      "loss": 0.49,
      "mean_token_accuracy": 0.8454527780413628,
      "num_tokens": 117647042.0,
      "step": 97840
    },
    {
      "entropy": 1.9133759438991547,
      "epoch": 0.3033264386112702,
      "grad_norm": 8.33125114440918,
      "learning_rate": 4.593458168350257e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8406713515520096,
      "num_tokens": 117658493.0,
      "step": 97850
    },
    {
      "entropy": 1.9860854953527451,
      "epoch": 0.3033574377363199,
      "grad_norm": 8.772767066955566,
      "learning_rate": 4.593223464566075e-06,
      "loss": 0.5346,
      "mean_token_accuracy": 0.836230854690075,
      "num_tokens": 117669208.0,
      "step": 97860
    },
    {
      "entropy": 1.8566246941685676,
      "epoch": 0.30338843686136957,
      "grad_norm": 2.640052556991577,
      "learning_rate": 4.592988796754947e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8560033723711967,
      "num_tokens": 117681693.0,
      "step": 97870
    },
    {
      "entropy": 1.90928722769022,
      "epoch": 0.3034194359864193,
      "grad_norm": 7.907132148742676,
      "learning_rate": 4.592754164907683e-06,
      "loss": 0.5258,
      "mean_token_accuracy": 0.8417934641242028,
      "num_tokens": 117693406.0,
      "step": 97880
    },
    {
      "entropy": 1.9087349817156791,
      "epoch": 0.30345043511146896,
      "grad_norm": 11.36373519897461,
      "learning_rate": 4.592519569015098e-06,
      "loss": 0.4901,
      "mean_token_accuracy": 0.8501476049423218,
      "num_tokens": 117704501.0,
      "step": 97890
    },
    {
      "entropy": 1.8858125910162926,
      "epoch": 0.3034814342365187,
      "grad_norm": 7.878525733947754,
      "learning_rate": 4.592285009068011e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8410215243697167,
      "num_tokens": 117716498.0,
      "step": 97900
    },
    {
      "entropy": 1.9222358748316766,
      "epoch": 0.30351243336156836,
      "grad_norm": 9.170816421508789,
      "learning_rate": 4.592050485057241e-06,
      "loss": 0.4626,
      "mean_token_accuracy": 0.8559870198369026,
      "num_tokens": 117727660.0,
      "step": 97910
    },
    {
      "entropy": 1.9125974863767623,
      "epoch": 0.3035434324866181,
      "grad_norm": 8.71693229675293,
      "learning_rate": 4.591815996973617e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8463147193193435,
      "num_tokens": 117739528.0,
      "step": 97920
    },
    {
      "entropy": 1.8418530434370042,
      "epoch": 0.30357443161166775,
      "grad_norm": 4.724992752075195,
      "learning_rate": 4.591581544807964e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.849432897567749,
      "num_tokens": 117752050.0,
      "step": 97930
    },
    {
      "entropy": 1.8719761818647385,
      "epoch": 0.3036054307367175,
      "grad_norm": 8.673355102539062,
      "learning_rate": 4.591347128551114e-06,
      "loss": 0.4561,
      "mean_token_accuracy": 0.8506433501839638,
      "num_tokens": 117763777.0,
      "step": 97940
    },
    {
      "entropy": 1.834936611354351,
      "epoch": 0.30363642986176714,
      "grad_norm": 4.6238603591918945,
      "learning_rate": 4.5911127481939e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8557968467473984,
      "num_tokens": 117776099.0,
      "step": 97950
    },
    {
      "entropy": 1.7863100260496139,
      "epoch": 0.30366742898681687,
      "grad_norm": 4.946314811706543,
      "learning_rate": 4.590878403727164e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8499247878789902,
      "num_tokens": 117789509.0,
      "step": 97960
    },
    {
      "entropy": 1.915014560520649,
      "epoch": 0.30369842811186654,
      "grad_norm": 8.316082000732422,
      "learning_rate": 4.5906440951417435e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8483388528227807,
      "num_tokens": 117800926.0,
      "step": 97970
    },
    {
      "entropy": 1.9649482190608978,
      "epoch": 0.30372942723691626,
      "grad_norm": 10.604179382324219,
      "learning_rate": 4.590409822428485e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.8398748651146889,
      "num_tokens": 117811622.0,
      "step": 97980
    },
    {
      "entropy": 1.7894397795200347,
      "epoch": 0.30376042636196593,
      "grad_norm": 6.535918712615967,
      "learning_rate": 4.590175585578233e-06,
      "loss": 0.3897,
      "mean_token_accuracy": 0.8634741649031639,
      "num_tokens": 117824809.0,
      "step": 97990
    },
    {
      "entropy": 1.8452475354075433,
      "epoch": 0.30379142548701565,
      "grad_norm": 9.069293975830078,
      "learning_rate": 4.589941384581842e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8545988813042641,
      "num_tokens": 117837581.0,
      "step": 98000
    },
    {
      "entropy": 1.9193524435162543,
      "epoch": 0.3038224246120653,
      "grad_norm": 9.2352933883667,
      "learning_rate": 4.589707219430166e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8450753495097161,
      "num_tokens": 117849268.0,
      "step": 98010
    },
    {
      "entropy": 1.8406575858592986,
      "epoch": 0.30385342373711505,
      "grad_norm": 8.772045135498047,
      "learning_rate": 4.589473090114059e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8576031014323234,
      "num_tokens": 117861975.0,
      "step": 98020
    },
    {
      "entropy": 1.8483123630285263,
      "epoch": 0.3038844228621647,
      "grad_norm": 9.5717191696167,
      "learning_rate": 4.5892389966243866e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8474526911973953,
      "num_tokens": 117874143.0,
      "step": 98030
    },
    {
      "entropy": 1.9467709794640542,
      "epoch": 0.3039154219872144,
      "grad_norm": 8.297847747802734,
      "learning_rate": 4.589004938952009e-06,
      "loss": 0.488,
      "mean_token_accuracy": 0.8526682198047638,
      "num_tokens": 117885322.0,
      "step": 98040
    },
    {
      "entropy": 1.859838457405567,
      "epoch": 0.3039464211122641,
      "grad_norm": 8.134725570678711,
      "learning_rate": 4.588770917087794e-06,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8431189700961113,
      "num_tokens": 117897060.0,
      "step": 98050
    },
    {
      "entropy": 1.8738309249281884,
      "epoch": 0.3039774202373138,
      "grad_norm": 4.704370021820068,
      "learning_rate": 4.5885369310226145e-06,
      "loss": 0.4939,
      "mean_token_accuracy": 0.8435174241662026,
      "num_tokens": 117909430.0,
      "step": 98060
    },
    {
      "entropy": 1.8503778785467149,
      "epoch": 0.3040084193623635,
      "grad_norm": 3.1746442317962646,
      "learning_rate": 4.588302980747341e-06,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8513809859752655,
      "num_tokens": 117920805.0,
      "step": 98070
    },
    {
      "entropy": 1.8909773424267768,
      "epoch": 0.30403941848741317,
      "grad_norm": 8.956877708435059,
      "learning_rate": 4.588069066252854e-06,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8422663122415542,
      "num_tokens": 117932448.0,
      "step": 98080
    },
    {
      "entropy": 1.8859625205397605,
      "epoch": 0.3040704176124629,
      "grad_norm": 6.800996780395508,
      "learning_rate": 4.587835187530031e-06,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8531765311956405,
      "num_tokens": 117944171.0,
      "step": 98090
    },
    {
      "entropy": 1.9186548352241517,
      "epoch": 0.30410141673751256,
      "grad_norm": 8.095877647399902,
      "learning_rate": 4.587601344569756e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8412599250674248,
      "num_tokens": 117956478.0,
      "step": 98100
    },
    {
      "entropy": 1.8034547924995423,
      "epoch": 0.3041324158625623,
      "grad_norm": 4.433010578155518,
      "learning_rate": 4.587367537362918e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8495987445116043,
      "num_tokens": 117969524.0,
      "step": 98110
    },
    {
      "entropy": 1.8625740513205529,
      "epoch": 0.30416341498761196,
      "grad_norm": 8.75419807434082,
      "learning_rate": 4.587133765900404e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8563158795237541,
      "num_tokens": 117982012.0,
      "step": 98120
    },
    {
      "entropy": 1.8077227592468261,
      "epoch": 0.3041944141126617,
      "grad_norm": 8.769314765930176,
      "learning_rate": 4.586900030173109e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8591718584299087,
      "num_tokens": 117994852.0,
      "step": 98130
    },
    {
      "entropy": 1.8287582576274872,
      "epoch": 0.30422541323771135,
      "grad_norm": 5.158862590789795,
      "learning_rate": 4.58666633017193e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8468107968568802,
      "num_tokens": 118007827.0,
      "step": 98140
    },
    {
      "entropy": 1.917576177418232,
      "epoch": 0.3042564123627611,
      "grad_norm": 10.31728458404541,
      "learning_rate": 4.586432665887766e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8305649489164353,
      "num_tokens": 118018948.0,
      "step": 98150
    },
    {
      "entropy": 1.9343233779072762,
      "epoch": 0.30428741148781074,
      "grad_norm": 7.640726089477539,
      "learning_rate": 4.586199037311519e-06,
      "loss": 0.5013,
      "mean_token_accuracy": 0.849469755589962,
      "num_tokens": 118030240.0,
      "step": 98160
    },
    {
      "entropy": 1.874649804830551,
      "epoch": 0.30431841061286047,
      "grad_norm": 7.788882255554199,
      "learning_rate": 4.585965444434098e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8480836063623428,
      "num_tokens": 118042287.0,
      "step": 98170
    },
    {
      "entropy": 1.894361424446106,
      "epoch": 0.30434940973791014,
      "grad_norm": 5.997152328491211,
      "learning_rate": 4.58573188724641e-06,
      "loss": 0.533,
      "mean_token_accuracy": 0.8430122479796409,
      "num_tokens": 118054246.0,
      "step": 98180
    },
    {
      "entropy": 1.8837274074554444,
      "epoch": 0.30438040886295986,
      "grad_norm": 8.705531120300293,
      "learning_rate": 4.585498365739368e-06,
      "loss": 0.4851,
      "mean_token_accuracy": 0.8436176598072052,
      "num_tokens": 118066667.0,
      "step": 98190
    },
    {
      "entropy": 1.905727145075798,
      "epoch": 0.30441140798800953,
      "grad_norm": 8.105955123901367,
      "learning_rate": 4.585264879903889e-06,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8468491420149803,
      "num_tokens": 118078925.0,
      "step": 98200
    },
    {
      "entropy": 1.9071518421173095,
      "epoch": 0.30444240711305925,
      "grad_norm": 8.408157348632812,
      "learning_rate": 4.585031429730893e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8430832877755166,
      "num_tokens": 118090390.0,
      "step": 98210
    },
    {
      "entropy": 1.8832137271761895,
      "epoch": 0.3044734062381089,
      "grad_norm": 4.172749042510986,
      "learning_rate": 4.5847980152113015e-06,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8408018171787262,
      "num_tokens": 118103271.0,
      "step": 98220
    },
    {
      "entropy": 1.8779758632183075,
      "epoch": 0.30450440536315865,
      "grad_norm": 6.640566825866699,
      "learning_rate": 4.584564636336039e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8598590075969696,
      "num_tokens": 118115121.0,
      "step": 98230
    },
    {
      "entropy": 1.8136912897229194,
      "epoch": 0.3045354044882083,
      "grad_norm": 9.391630172729492,
      "learning_rate": 4.584331293096037e-06,
      "loss": 0.4272,
      "mean_token_accuracy": 0.8552101209759713,
      "num_tokens": 118127803.0,
      "step": 98240
    },
    {
      "entropy": 1.7988118350505828,
      "epoch": 0.30456640361325804,
      "grad_norm": 3.3279430866241455,
      "learning_rate": 4.584097985482225e-06,
      "loss": 0.3666,
      "mean_token_accuracy": 0.8751557558774948,
      "num_tokens": 118140663.0,
      "step": 98250
    },
    {
      "entropy": 1.8820227935910225,
      "epoch": 0.3045974027383077,
      "grad_norm": 12.045969009399414,
      "learning_rate": 4.583864713485541e-06,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8497835651040078,
      "num_tokens": 118152695.0,
      "step": 98260
    },
    {
      "entropy": 1.8159508243203164,
      "epoch": 0.30462840186335743,
      "grad_norm": 5.319043159484863,
      "learning_rate": 4.583631477096921e-06,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8409684166312218,
      "num_tokens": 118165387.0,
      "step": 98270
    },
    {
      "entropy": 1.8920594125986099,
      "epoch": 0.3046594009884071,
      "grad_norm": 8.387589454650879,
      "learning_rate": 4.583398276307309e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.8461384385824203,
      "num_tokens": 118177881.0,
      "step": 98280
    },
    {
      "entropy": 1.8834513157606125,
      "epoch": 0.30469040011345677,
      "grad_norm": 7.704468250274658,
      "learning_rate": 4.58316511110765e-06,
      "loss": 0.5188,
      "mean_token_accuracy": 0.84387187063694,
      "num_tokens": 118189579.0,
      "step": 98290
    },
    {
      "entropy": 1.9479789227247237,
      "epoch": 0.3047213992385065,
      "grad_norm": 9.058480262756348,
      "learning_rate": 4.582931981488891e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.822111751139164,
      "num_tokens": 118201583.0,
      "step": 98300
    },
    {
      "entropy": 1.914680229127407,
      "epoch": 0.30475239836355617,
      "grad_norm": 9.425095558166504,
      "learning_rate": 4.582698887441983e-06,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8463716924190521,
      "num_tokens": 118212830.0,
      "step": 98310
    },
    {
      "entropy": 1.9096634030342101,
      "epoch": 0.3047833974886059,
      "grad_norm": 4.8716912269592285,
      "learning_rate": 4.582465828957883e-06,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8446076348423958,
      "num_tokens": 118223765.0,
      "step": 98320
    },
    {
      "entropy": 1.8158196270465852,
      "epoch": 0.30481439661365556,
      "grad_norm": 7.788163185119629,
      "learning_rate": 4.582232806027548e-06,
      "loss": 0.4027,
      "mean_token_accuracy": 0.8658209949731827,
      "num_tokens": 118236493.0,
      "step": 98330
    },
    {
      "entropy": 1.8198370561003685,
      "epoch": 0.3048453957387053,
      "grad_norm": 3.936845541000366,
      "learning_rate": 4.581999818641939e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8498961672186851,
      "num_tokens": 118248905.0,
      "step": 98340
    },
    {
      "entropy": 1.8224836379289626,
      "epoch": 0.30487639486375495,
      "grad_norm": 2.4840080738067627,
      "learning_rate": 4.5817668667920205e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8523753926157951,
      "num_tokens": 118262429.0,
      "step": 98350
    },
    {
      "entropy": 1.776771107316017,
      "epoch": 0.3049073939888047,
      "grad_norm": 7.824039936065674,
      "learning_rate": 4.58153395046876e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8675816237926484,
      "num_tokens": 118275691.0,
      "step": 98360
    },
    {
      "entropy": 1.8841424211859703,
      "epoch": 0.30493839311385434,
      "grad_norm": 7.849323749542236,
      "learning_rate": 4.581301069663129e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8445948898792267,
      "num_tokens": 118288240.0,
      "step": 98370
    },
    {
      "entropy": 1.8760451026260854,
      "epoch": 0.30496939223890407,
      "grad_norm": 8.8077974319458,
      "learning_rate": 4.5810682243661e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8469317957758904,
      "num_tokens": 118301110.0,
      "step": 98380
    },
    {
      "entropy": 1.8852962955832482,
      "epoch": 0.30500039136395374,
      "grad_norm": 8.56628704071045,
      "learning_rate": 4.580835414568652e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.849009807407856,
      "num_tokens": 118312570.0,
      "step": 98390
    },
    {
      "entropy": 1.8616637766361237,
      "epoch": 0.30503139048900346,
      "grad_norm": 4.301907539367676,
      "learning_rate": 4.580602640261765e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.852383928000927,
      "num_tokens": 118324622.0,
      "step": 98400
    },
    {
      "entropy": 1.9427771091461181,
      "epoch": 0.30506238961405313,
      "grad_norm": 6.355135440826416,
      "learning_rate": 4.580369901436422e-06,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8413548216223716,
      "num_tokens": 118336164.0,
      "step": 98410
    },
    {
      "entropy": 1.9013810217380525,
      "epoch": 0.30509338873910286,
      "grad_norm": 8.685293197631836,
      "learning_rate": 4.580137198083611e-06,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8395018517971039,
      "num_tokens": 118348429.0,
      "step": 98420
    },
    {
      "entropy": 1.8083545833826065,
      "epoch": 0.3051243878641525,
      "grad_norm": 9.026556968688965,
      "learning_rate": 4.5799045301943205e-06,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8529153957962989,
      "num_tokens": 118361098.0,
      "step": 98430
    },
    {
      "entropy": 1.8322697907686234,
      "epoch": 0.30515538698920225,
      "grad_norm": 8.164215087890625,
      "learning_rate": 4.579671897759546e-06,
      "loss": 0.49,
      "mean_token_accuracy": 0.8390917211771012,
      "num_tokens": 118374073.0,
      "step": 98440
    },
    {
      "entropy": 1.8684598043560983,
      "epoch": 0.3051863861142519,
      "grad_norm": 7.570199489593506,
      "learning_rate": 4.579439300770282e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8506176099181175,
      "num_tokens": 118385894.0,
      "step": 98450
    },
    {
      "entropy": 1.959225982427597,
      "epoch": 0.30521738523930164,
      "grad_norm": 8.114801406860352,
      "learning_rate": 4.579206739217529e-06,
      "loss": 0.5804,
      "mean_token_accuracy": 0.8300017550587654,
      "num_tokens": 118397286.0,
      "step": 98460
    },
    {
      "entropy": 1.865318314731121,
      "epoch": 0.3052483843643513,
      "grad_norm": 9.157855033874512,
      "learning_rate": 4.57897421309229e-06,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8516108497977257,
      "num_tokens": 118408785.0,
      "step": 98470
    },
    {
      "entropy": 1.8846929490566253,
      "epoch": 0.30527938348940103,
      "grad_norm": 7.941446304321289,
      "learning_rate": 4.5787417223855705e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8428527131676674,
      "num_tokens": 118420758.0,
      "step": 98480
    },
    {
      "entropy": 1.8963923439383508,
      "epoch": 0.3053103826144507,
      "grad_norm": 9.210304260253906,
      "learning_rate": 4.57850926708838e-06,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8412079870700836,
      "num_tokens": 118432505.0,
      "step": 98490
    },
    {
      "entropy": 1.9127984009683132,
      "epoch": 0.30534138173950043,
      "grad_norm": 8.439544677734375,
      "learning_rate": 4.578276847191734e-06,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8416270837187767,
      "num_tokens": 118444368.0,
      "step": 98500
    },
    {
      "entropy": 1.8232401758432388,
      "epoch": 0.3053723808645501,
      "grad_norm": 8.125101089477539,
      "learning_rate": 4.578044462686643e-06,
      "loss": 0.4172,
      "mean_token_accuracy": 0.8568010330200195,
      "num_tokens": 118457274.0,
      "step": 98510
    },
    {
      "entropy": 1.988058003783226,
      "epoch": 0.30540337998959977,
      "grad_norm": 8.513818740844727,
      "learning_rate": 4.577812113564129e-06,
      "loss": 0.5254,
      "mean_token_accuracy": 0.8399447157979012,
      "num_tokens": 118468141.0,
      "step": 98520
    },
    {
      "entropy": 1.8777014121413231,
      "epoch": 0.3054343791146495,
      "grad_norm": 7.6989850997924805,
      "learning_rate": 4.577579799815213e-06,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8421044409275055,
      "num_tokens": 118479910.0,
      "step": 98530
    },
    {
      "entropy": 1.8567018955945969,
      "epoch": 0.30546537823969916,
      "grad_norm": 8.173233032226562,
      "learning_rate": 4.57734752143092e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8535317838191986,
      "num_tokens": 118491450.0,
      "step": 98540
    },
    {
      "entropy": 1.8805242463946343,
      "epoch": 0.3054963773647489,
      "grad_norm": 8.804410934448242,
      "learning_rate": 4.577115278402281e-06,
      "loss": 0.4536,
      "mean_token_accuracy": 0.8533170059323311,
      "num_tokens": 118502975.0,
      "step": 98550
    },
    {
      "entropy": 1.7814088612794876,
      "epoch": 0.30552737648979855,
      "grad_norm": 8.247437477111816,
      "learning_rate": 4.5768830707203236e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8531454712152481,
      "num_tokens": 118515769.0,
      "step": 98560
    },
    {
      "entropy": 1.8369492530822753,
      "epoch": 0.3055583756148483,
      "grad_norm": 7.819140911102295,
      "learning_rate": 4.576650898376085e-06,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8478553339838981,
      "num_tokens": 118528258.0,
      "step": 98570
    },
    {
      "entropy": 1.868079724907875,
      "epoch": 0.30558937473989795,
      "grad_norm": 4.0178446769714355,
      "learning_rate": 4.5764187613606045e-06,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8433499991893768,
      "num_tokens": 118540637.0,
      "step": 98580
    },
    {
      "entropy": 1.825770527124405,
      "epoch": 0.30562037386494767,
      "grad_norm": 6.859602451324463,
      "learning_rate": 4.57618665966492e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8496432945132255,
      "num_tokens": 118553277.0,
      "step": 98590
    },
    {
      "entropy": 1.8329296857118607,
      "epoch": 0.30565137298999734,
      "grad_norm": 7.438360214233398,
      "learning_rate": 4.575954593280079e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.858691118657589,
      "num_tokens": 118564885.0,
      "step": 98600
    },
    {
      "entropy": 1.8381449341773988,
      "epoch": 0.30568237211504706,
      "grad_norm": 4.074737071990967,
      "learning_rate": 4.575722562197127e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8469588398933411,
      "num_tokens": 118577848.0,
      "step": 98610
    },
    {
      "entropy": 1.85487762093544,
      "epoch": 0.30571337124009673,
      "grad_norm": 7.444735527038574,
      "learning_rate": 4.575490566407115e-06,
      "loss": 0.4466,
      "mean_token_accuracy": 0.849855688214302,
      "num_tokens": 118590154.0,
      "step": 98620
    },
    {
      "entropy": 1.8266935624182223,
      "epoch": 0.30574437036514646,
      "grad_norm": 4.6787896156311035,
      "learning_rate": 4.575258605901098e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8533053979277611,
      "num_tokens": 118603425.0,
      "step": 98630
    },
    {
      "entropy": 1.886932836472988,
      "epoch": 0.3057753694901961,
      "grad_norm": 7.9331889152526855,
      "learning_rate": 4.575026680670132e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8492220997810364,
      "num_tokens": 118614798.0,
      "step": 98640
    },
    {
      "entropy": 1.803956750035286,
      "epoch": 0.30580636861524585,
      "grad_norm": 7.170061111450195,
      "learning_rate": 4.5747947907052775e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8577531769871711,
      "num_tokens": 118628191.0,
      "step": 98650
    },
    {
      "entropy": 1.8016648098826409,
      "epoch": 0.3058373677402955,
      "grad_norm": 3.502978801727295,
      "learning_rate": 4.574562935997597e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8521159499883652,
      "num_tokens": 118641385.0,
      "step": 98660
    },
    {
      "entropy": 1.8296085387468337,
      "epoch": 0.30586836686534524,
      "grad_norm": 8.538934707641602,
      "learning_rate": 4.574331116538158e-06,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8565147161483765,
      "num_tokens": 118653485.0,
      "step": 98670
    },
    {
      "entropy": 1.8369739711284638,
      "epoch": 0.3058993659903949,
      "grad_norm": 7.255573749542236,
      "learning_rate": 4.574099332318032e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8531767651438713,
      "num_tokens": 118665318.0,
      "step": 98680
    },
    {
      "entropy": 1.8349349424242973,
      "epoch": 0.30593036511544464,
      "grad_norm": 8.855729103088379,
      "learning_rate": 4.573867583328289e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8501504331827163,
      "num_tokens": 118678218.0,
      "step": 98690
    },
    {
      "entropy": 1.8983908250927926,
      "epoch": 0.3059613642404943,
      "grad_norm": 8.671244621276855,
      "learning_rate": 4.573635869560006e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8487844780087471,
      "num_tokens": 118690095.0,
      "step": 98700
    },
    {
      "entropy": 1.8810410752892495,
      "epoch": 0.30599236336554403,
      "grad_norm": 8.668618202209473,
      "learning_rate": 4.573404191004263e-06,
      "loss": 0.482,
      "mean_token_accuracy": 0.8451575711369514,
      "num_tokens": 118701551.0,
      "step": 98710
    },
    {
      "entropy": 1.9009374052286148,
      "epoch": 0.3060233624905937,
      "grad_norm": 8.427909851074219,
      "learning_rate": 4.573172547652142e-06,
      "loss": 0.4786,
      "mean_token_accuracy": 0.839051516354084,
      "num_tokens": 118713008.0,
      "step": 98720
    },
    {
      "entropy": 1.9209471434354781,
      "epoch": 0.3060543616156434,
      "grad_norm": 7.763575077056885,
      "learning_rate": 4.572940939494728e-06,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8371775403618813,
      "num_tokens": 118724161.0,
      "step": 98730
    },
    {
      "entropy": 1.9421784669160842,
      "epoch": 0.3060853607406931,
      "grad_norm": 8.41716194152832,
      "learning_rate": 4.5727093665231095e-06,
      "loss": 0.498,
      "mean_token_accuracy": 0.8471307173371315,
      "num_tokens": 118734733.0,
      "step": 98740
    },
    {
      "entropy": 1.8512969225645066,
      "epoch": 0.3061163598657428,
      "grad_norm": 4.9068193435668945,
      "learning_rate": 4.57247782872838e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8417642742395401,
      "num_tokens": 118747239.0,
      "step": 98750
    },
    {
      "entropy": 1.8801366731524467,
      "epoch": 0.3061473589907925,
      "grad_norm": 8.114398956298828,
      "learning_rate": 4.5722463261016335e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8596544772386551,
      "num_tokens": 118759918.0,
      "step": 98760
    },
    {
      "entropy": 1.8776177152991296,
      "epoch": 0.30617835811584215,
      "grad_norm": 7.3587141036987305,
      "learning_rate": 4.572014858633968e-06,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8482073560357094,
      "num_tokens": 118771855.0,
      "step": 98770
    },
    {
      "entropy": 1.8818192645907401,
      "epoch": 0.3062093572408919,
      "grad_norm": 10.548823356628418,
      "learning_rate": 4.571783426316486e-06,
      "loss": 0.5322,
      "mean_token_accuracy": 0.835597887635231,
      "num_tokens": 118783654.0,
      "step": 98780
    },
    {
      "entropy": 1.8536282077431678,
      "epoch": 0.30624035636594155,
      "grad_norm": 3.757561206817627,
      "learning_rate": 4.571552029140291e-06,
      "loss": 0.4171,
      "mean_token_accuracy": 0.860873955488205,
      "num_tokens": 118795662.0,
      "step": 98790
    },
    {
      "entropy": 1.8578494921326638,
      "epoch": 0.30627135549099127,
      "grad_norm": 6.849057197570801,
      "learning_rate": 4.57132066709649e-06,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8355998247861862,
      "num_tokens": 118807279.0,
      "step": 98800
    },
    {
      "entropy": 1.9314350634813309,
      "epoch": 0.30630235461604094,
      "grad_norm": 9.63023853302002,
      "learning_rate": 4.571089340176196e-06,
      "loss": 0.5303,
      "mean_token_accuracy": 0.8435678333044052,
      "num_tokens": 118818128.0,
      "step": 98810
    },
    {
      "entropy": 1.816218902170658,
      "epoch": 0.30633335374109066,
      "grad_norm": 8.004508972167969,
      "learning_rate": 4.570858048370521e-06,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8539455533027649,
      "num_tokens": 118830413.0,
      "step": 98820
    },
    {
      "entropy": 1.9110864594578743,
      "epoch": 0.30636435286614033,
      "grad_norm": 8.797080039978027,
      "learning_rate": 4.570626791670582e-06,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8348402082920074,
      "num_tokens": 118841694.0,
      "step": 98830
    },
    {
      "entropy": 1.8853062570095063,
      "epoch": 0.30639535199119006,
      "grad_norm": 8.307490348815918,
      "learning_rate": 4.570395570067499e-06,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8420171350240707,
      "num_tokens": 118853649.0,
      "step": 98840
    },
    {
      "entropy": 1.902986891567707,
      "epoch": 0.3064263511162397,
      "grad_norm": 6.329873561859131,
      "learning_rate": 4.5701643835523984e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8435382351279259,
      "num_tokens": 118864431.0,
      "step": 98850
    },
    {
      "entropy": 1.9434664219617843,
      "epoch": 0.30645735024128945,
      "grad_norm": 7.9458184242248535,
      "learning_rate": 4.569933232116404e-06,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8374830722808838,
      "num_tokens": 118875327.0,
      "step": 98860
    },
    {
      "entropy": 1.824971318244934,
      "epoch": 0.3064883493663391,
      "grad_norm": 8.772302627563477,
      "learning_rate": 4.569702115750646e-06,
      "loss": 0.417,
      "mean_token_accuracy": 0.8616791382431984,
      "num_tokens": 118888102.0,
      "step": 98870
    },
    {
      "entropy": 1.924868457019329,
      "epoch": 0.30651934849138884,
      "grad_norm": 7.891080379486084,
      "learning_rate": 4.569471034446258e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8509886354207993,
      "num_tokens": 118898982.0,
      "step": 98880
    },
    {
      "entropy": 1.8691664278507232,
      "epoch": 0.3065503476164385,
      "grad_norm": 8.10325813293457,
      "learning_rate": 4.5692399881943754e-06,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8400322616100311,
      "num_tokens": 118910630.0,
      "step": 98890
    },
    {
      "entropy": 1.8352098122239113,
      "epoch": 0.30658134674148824,
      "grad_norm": 3.6725313663482666,
      "learning_rate": 4.569008976986136e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8578842118382454,
      "num_tokens": 118923666.0,
      "step": 98900
    },
    {
      "entropy": 1.852015070617199,
      "epoch": 0.3066123458665379,
      "grad_norm": 7.963640213012695,
      "learning_rate": 4.568778000812685e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8513859167695046,
      "num_tokens": 118935379.0,
      "step": 98910
    },
    {
      "entropy": 1.8301201090216637,
      "epoch": 0.30664334499158763,
      "grad_norm": 7.35858678817749,
      "learning_rate": 4.568547059665164e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8498409286141395,
      "num_tokens": 118948414.0,
      "step": 98920
    },
    {
      "entropy": 1.8332775115966797,
      "epoch": 0.3066743441166373,
      "grad_norm": 4.402047157287598,
      "learning_rate": 4.568316153534725e-06,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8492154240608215,
      "num_tokens": 118961750.0,
      "step": 98930
    },
    {
      "entropy": 1.912602500617504,
      "epoch": 0.306705343241687,
      "grad_norm": 8.157501220703125,
      "learning_rate": 4.568085282412518e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8328171208500862,
      "num_tokens": 118973215.0,
      "step": 98940
    },
    {
      "entropy": 1.8730465933680533,
      "epoch": 0.3067363423667367,
      "grad_norm": 10.181466102600098,
      "learning_rate": 4.567854446289697e-06,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8437885701656341,
      "num_tokens": 118985524.0,
      "step": 98950
    },
    {
      "entropy": 1.902740554511547,
      "epoch": 0.3067673414917864,
      "grad_norm": 9.36436939239502,
      "learning_rate": 4.567623645157422e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8489101201295852,
      "num_tokens": 118997046.0,
      "step": 98960
    },
    {
      "entropy": 1.8747300267219544,
      "epoch": 0.3067983406168361,
      "grad_norm": 9.957647323608398,
      "learning_rate": 4.567392879006852e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8495268151164055,
      "num_tokens": 119009030.0,
      "step": 98970
    },
    {
      "entropy": 1.8686933800578118,
      "epoch": 0.3068293397418858,
      "grad_norm": 7.555231094360352,
      "learning_rate": 4.567162147829152e-06,
      "loss": 0.491,
      "mean_token_accuracy": 0.8467110440135002,
      "num_tokens": 119020843.0,
      "step": 98980
    },
    {
      "entropy": 1.884840413928032,
      "epoch": 0.3068603388669355,
      "grad_norm": 7.894239902496338,
      "learning_rate": 4.56693145161549e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8425130993127823,
      "num_tokens": 119032739.0,
      "step": 98990
    },
    {
      "entropy": 1.846720139682293,
      "epoch": 0.3068913379919852,
      "grad_norm": 7.1045403480529785,
      "learning_rate": 4.566700790357034e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8530875831842423,
      "num_tokens": 119044858.0,
      "step": 99000
    },
    {
      "entropy": 1.9004247322678567,
      "epoch": 0.30692233711703487,
      "grad_norm": 3.9099464416503906,
      "learning_rate": 4.56647016404496e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8432266771793365,
      "num_tokens": 119056720.0,
      "step": 99010
    },
    {
      "entropy": 1.845305010676384,
      "epoch": 0.30695333624208454,
      "grad_norm": 4.365594387054443,
      "learning_rate": 4.566239572670445e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8529621243476868,
      "num_tokens": 119069481.0,
      "step": 99020
    },
    {
      "entropy": 1.8452321007847785,
      "epoch": 0.30698433536713426,
      "grad_norm": 7.914214611053467,
      "learning_rate": 4.566009016224666e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.8480882182717323,
      "num_tokens": 119081620.0,
      "step": 99030
    },
    {
      "entropy": 1.8590731248259544,
      "epoch": 0.30701533449218393,
      "grad_norm": 8.533273696899414,
      "learning_rate": 4.565778494698808e-06,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8423286929726601,
      "num_tokens": 119093680.0,
      "step": 99040
    },
    {
      "entropy": 1.8688816666603087,
      "epoch": 0.30704633361723366,
      "grad_norm": 8.64100456237793,
      "learning_rate": 4.5655480080840556e-06,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8599061399698258,
      "num_tokens": 119105568.0,
      "step": 99050
    },
    {
      "entropy": 1.853632289171219,
      "epoch": 0.3070773327422833,
      "grad_norm": 2.459723949432373,
      "learning_rate": 4.565317556371598e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8528544023633003,
      "num_tokens": 119118403.0,
      "step": 99060
    },
    {
      "entropy": 1.867573080956936,
      "epoch": 0.30710833186733305,
      "grad_norm": 7.473991394042969,
      "learning_rate": 4.56508713955263e-06,
      "loss": 0.477,
      "mean_token_accuracy": 0.844258151948452,
      "num_tokens": 119130445.0,
      "step": 99070
    },
    {
      "entropy": 1.8869913056492806,
      "epoch": 0.3071393309923827,
      "grad_norm": 6.70831298828125,
      "learning_rate": 4.564856757618344e-06,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8497973084449768,
      "num_tokens": 119142338.0,
      "step": 99080
    },
    {
      "entropy": 1.8372093975543975,
      "epoch": 0.30717033011743244,
      "grad_norm": 7.9739603996276855,
      "learning_rate": 4.564626410559939e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8506271958351135,
      "num_tokens": 119154394.0,
      "step": 99090
    },
    {
      "entropy": 1.8323589369654656,
      "epoch": 0.3072013292424821,
      "grad_norm": 8.829228401184082,
      "learning_rate": 4.564396098368618e-06,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8452668026089668,
      "num_tokens": 119167066.0,
      "step": 99100
    },
    {
      "entropy": 1.9608383178710938,
      "epoch": 0.30723232836753184,
      "grad_norm": 8.926813125610352,
      "learning_rate": 4.564165821035583e-06,
      "loss": 0.5453,
      "mean_token_accuracy": 0.8367260545492172,
      "num_tokens": 119178017.0,
      "step": 99110
    },
    {
      "entropy": 1.861896750330925,
      "epoch": 0.3072633274925815,
      "grad_norm": 3.713798761367798,
      "learning_rate": 4.563935578552043e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8421308517456054,
      "num_tokens": 119190396.0,
      "step": 99120
    },
    {
      "entropy": 1.9013278767466546,
      "epoch": 0.30729432661763123,
      "grad_norm": 9.141477584838867,
      "learning_rate": 4.563705370909211e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8393938109278679,
      "num_tokens": 119201792.0,
      "step": 99130
    },
    {
      "entropy": 1.850807547569275,
      "epoch": 0.3073253257426809,
      "grad_norm": 4.37838888168335,
      "learning_rate": 4.563475198098299e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8489650651812554,
      "num_tokens": 119214868.0,
      "step": 99140
    },
    {
      "entropy": 1.7602989837527274,
      "epoch": 0.3073563248677306,
      "grad_norm": 3.604959726333618,
      "learning_rate": 4.563245060110523e-06,
      "loss": 0.412,
      "mean_token_accuracy": 0.8620150581002235,
      "num_tokens": 119228299.0,
      "step": 99150
    },
    {
      "entropy": 1.949777290225029,
      "epoch": 0.3073873239927803,
      "grad_norm": 7.350498199462891,
      "learning_rate": 4.563014956937104e-06,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8472014635801315,
      "num_tokens": 119239986.0,
      "step": 99160
    },
    {
      "entropy": 1.9088080897927284,
      "epoch": 0.30741832311783,
      "grad_norm": 8.599087715148926,
      "learning_rate": 4.562784888569266e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8541075602173805,
      "num_tokens": 119251421.0,
      "step": 99170
    },
    {
      "entropy": 1.9487151458859444,
      "epoch": 0.3074493222428797,
      "grad_norm": 9.30274772644043,
      "learning_rate": 4.562554854998235e-06,
      "loss": 0.4915,
      "mean_token_accuracy": 0.842944149672985,
      "num_tokens": 119263018.0,
      "step": 99180
    },
    {
      "entropy": 1.8647134892642498,
      "epoch": 0.3074803213679294,
      "grad_norm": 7.472745418548584,
      "learning_rate": 4.56232485621524e-06,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8561579942703247,
      "num_tokens": 119275984.0,
      "step": 99190
    },
    {
      "entropy": 1.8933315068483352,
      "epoch": 0.3075113204929791,
      "grad_norm": 8.288209915161133,
      "learning_rate": 4.5620948922115156e-06,
      "loss": 0.5475,
      "mean_token_accuracy": 0.831151905655861,
      "num_tokens": 119287771.0,
      "step": 99200
    },
    {
      "entropy": 1.8554987102746963,
      "epoch": 0.3075423196180288,
      "grad_norm": 7.526989459991455,
      "learning_rate": 4.561864962978294e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8532639876008034,
      "num_tokens": 119299903.0,
      "step": 99210
    },
    {
      "entropy": 1.9998096972703934,
      "epoch": 0.30757331874307847,
      "grad_norm": 9.368232727050781,
      "learning_rate": 4.5616350685068165e-06,
      "loss": 0.567,
      "mean_token_accuracy": 0.8291105717420578,
      "num_tokens": 119310444.0,
      "step": 99220
    },
    {
      "entropy": 1.898825192451477,
      "epoch": 0.3076043178681282,
      "grad_norm": 8.540651321411133,
      "learning_rate": 4.561405208788324e-06,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8367143541574478,
      "num_tokens": 119321635.0,
      "step": 99230
    },
    {
      "entropy": 1.8685836613178253,
      "epoch": 0.30763531699317787,
      "grad_norm": 3.111175775527954,
      "learning_rate": 4.561175383814061e-06,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8581192702054977,
      "num_tokens": 119333768.0,
      "step": 99240
    },
    {
      "entropy": 1.8917466133832932,
      "epoch": 0.3076663161182276,
      "grad_norm": 8.266393661499023,
      "learning_rate": 4.560945593575276e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8560093253850937,
      "num_tokens": 119345845.0,
      "step": 99250
    },
    {
      "entropy": 1.9194523006677628,
      "epoch": 0.30769731524327726,
      "grad_norm": 7.8154683113098145,
      "learning_rate": 4.560715838063221e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8392749205231667,
      "num_tokens": 119356959.0,
      "step": 99260
    },
    {
      "entropy": 1.9294108331203461,
      "epoch": 0.3077283143683269,
      "grad_norm": 7.9482831954956055,
      "learning_rate": 4.560486117269149e-06,
      "loss": 0.5774,
      "mean_token_accuracy": 0.8344770103693009,
      "num_tokens": 119368443.0,
      "step": 99270
    },
    {
      "entropy": 1.8910026296973228,
      "epoch": 0.30775931349337665,
      "grad_norm": 6.961696624755859,
      "learning_rate": 4.560256431184316e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8482179164886474,
      "num_tokens": 119380584.0,
      "step": 99280
    },
    {
      "entropy": 1.9641872704029084,
      "epoch": 0.3077903126184263,
      "grad_norm": 9.529417991638184,
      "learning_rate": 4.5600267797999856e-06,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8355140164494514,
      "num_tokens": 119390739.0,
      "step": 99290
    },
    {
      "entropy": 1.8502513483166694,
      "epoch": 0.30782131174347604,
      "grad_norm": 7.629122734069824,
      "learning_rate": 4.55979716310742e-06,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8559545025229454,
      "num_tokens": 119403174.0,
      "step": 99300
    },
    {
      "entropy": 1.9880924820899963,
      "epoch": 0.3078523108685257,
      "grad_norm": 11.568595886230469,
      "learning_rate": 4.5595675810978835e-06,
      "loss": 0.5379,
      "mean_token_accuracy": 0.8477470189332962,
      "num_tokens": 119414065.0,
      "step": 99310
    },
    {
      "entropy": 1.9389469519257545,
      "epoch": 0.30788330999357544,
      "grad_norm": 8.663850784301758,
      "learning_rate": 4.559338033762647e-06,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8486195042729378,
      "num_tokens": 119425228.0,
      "step": 99320
    },
    {
      "entropy": 1.8266063407063484,
      "epoch": 0.3079143091186251,
      "grad_norm": 5.284822940826416,
      "learning_rate": 4.559108521092985e-06,
      "loss": 0.3787,
      "mean_token_accuracy": 0.861154480278492,
      "num_tokens": 119439104.0,
      "step": 99330
    },
    {
      "entropy": 1.8920496240258218,
      "epoch": 0.30794530824367483,
      "grad_norm": 3.9276669025421143,
      "learning_rate": 4.558879043080171e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8588141456246376,
      "num_tokens": 119451373.0,
      "step": 99340
    },
    {
      "entropy": 1.951807254552841,
      "epoch": 0.3079763073687245,
      "grad_norm": 7.063690662384033,
      "learning_rate": 4.5586495997154835e-06,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8351276248693467,
      "num_tokens": 119462888.0,
      "step": 99350
    },
    {
      "entropy": 2.002240237593651,
      "epoch": 0.3080073064937742,
      "grad_norm": 9.491320610046387,
      "learning_rate": 4.558420190990207e-06,
      "loss": 0.5519,
      "mean_token_accuracy": 0.8310097977519035,
      "num_tokens": 119473869.0,
      "step": 99360
    },
    {
      "entropy": 1.847242882847786,
      "epoch": 0.3080383056188239,
      "grad_norm": 8.938675880432129,
      "learning_rate": 4.558190816895623e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8449984312057495,
      "num_tokens": 119486256.0,
      "step": 99370
    },
    {
      "entropy": 1.865819439291954,
      "epoch": 0.3080693047438736,
      "grad_norm": 7.645913600921631,
      "learning_rate": 4.557961477423024e-06,
      "loss": 0.4424,
      "mean_token_accuracy": 0.848369000852108,
      "num_tokens": 119497901.0,
      "step": 99380
    },
    {
      "entropy": 1.8621415674686432,
      "epoch": 0.3081003038689233,
      "grad_norm": 9.088942527770996,
      "learning_rate": 4.557732172563696e-06,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8500281676650048,
      "num_tokens": 119510003.0,
      "step": 99390
    },
    {
      "entropy": 1.8306550726294517,
      "epoch": 0.308131302993973,
      "grad_norm": 7.942266941070557,
      "learning_rate": 4.557502902308936e-06,
      "loss": 0.3983,
      "mean_token_accuracy": 0.8491405546665192,
      "num_tokens": 119522659.0,
      "step": 99400
    },
    {
      "entropy": 1.9008427813649178,
      "epoch": 0.3081623021190227,
      "grad_norm": 10.027981758117676,
      "learning_rate": 4.557273666650041e-06,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8394996523857117,
      "num_tokens": 119534862.0,
      "step": 99410
    },
    {
      "entropy": 1.9339268684387207,
      "epoch": 0.3081933012440724,
      "grad_norm": 8.332462310791016,
      "learning_rate": 4.5570444655783105e-06,
      "loss": 0.5281,
      "mean_token_accuracy": 0.848136767745018,
      "num_tokens": 119546767.0,
      "step": 99420
    },
    {
      "entropy": 1.8939686760306358,
      "epoch": 0.3082243003691221,
      "grad_norm": 7.6620259284973145,
      "learning_rate": 4.556815299085049e-06,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8582808762788773,
      "num_tokens": 119558146.0,
      "step": 99430
    },
    {
      "entropy": 1.8960684776306151,
      "epoch": 0.3082552994941718,
      "grad_norm": 8.92011547088623,
      "learning_rate": 4.556586167161562e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8485264018177986,
      "num_tokens": 119569433.0,
      "step": 99440
    },
    {
      "entropy": 1.8812731936573983,
      "epoch": 0.30828629861922147,
      "grad_norm": 7.918720722198486,
      "learning_rate": 4.556357069799159e-06,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8485799968242645,
      "num_tokens": 119581623.0,
      "step": 99450
    },
    {
      "entropy": 1.9013389393687248,
      "epoch": 0.3083172977442712,
      "grad_norm": 8.778814315795898,
      "learning_rate": 4.556128006989152e-06,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8451300621032715,
      "num_tokens": 119593213.0,
      "step": 99460
    },
    {
      "entropy": 1.8785056233406068,
      "epoch": 0.30834829686932086,
      "grad_norm": 7.801638126373291,
      "learning_rate": 4.555898978722858e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8590483710169792,
      "num_tokens": 119605344.0,
      "step": 99470
    },
    {
      "entropy": 1.9365007862448693,
      "epoch": 0.3083792959943706,
      "grad_norm": 10.32783317565918,
      "learning_rate": 4.555669984991595e-06,
      "loss": 0.5556,
      "mean_token_accuracy": 0.830289502441883,
      "num_tokens": 119617211.0,
      "step": 99480
    },
    {
      "entropy": 1.9110652074217795,
      "epoch": 0.30841029511942025,
      "grad_norm": 7.38109016418457,
      "learning_rate": 4.555441025786685e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8413866236805916,
      "num_tokens": 119628112.0,
      "step": 99490
    },
    {
      "entropy": 1.9205258354544639,
      "epoch": 0.30844129424447,
      "grad_norm": 7.779130458831787,
      "learning_rate": 4.555212101099452e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8441153079271316,
      "num_tokens": 119639907.0,
      "step": 99500
    },
    {
      "entropy": 1.9123625665903092,
      "epoch": 0.30847229336951965,
      "grad_norm": 3.761373281478882,
      "learning_rate": 4.554983210921223e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8475768864154816,
      "num_tokens": 119651966.0,
      "step": 99510
    },
    {
      "entropy": 1.8980589851737022,
      "epoch": 0.3085032924945693,
      "grad_norm": 9.282678604125977,
      "learning_rate": 4.55475435524333e-06,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8474918335676194,
      "num_tokens": 119663224.0,
      "step": 99520
    },
    {
      "entropy": 1.807744611799717,
      "epoch": 0.30853429161961904,
      "grad_norm": 8.637643814086914,
      "learning_rate": 4.554525534057108e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.8565024748444557,
      "num_tokens": 119676021.0,
      "step": 99530
    },
    {
      "entropy": 1.936571803689003,
      "epoch": 0.3085652907446687,
      "grad_norm": 8.615496635437012,
      "learning_rate": 4.554296747353892e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8555189058184623,
      "num_tokens": 119686726.0,
      "step": 99540
    },
    {
      "entropy": 1.869454000890255,
      "epoch": 0.30859628986971843,
      "grad_norm": 8.078155517578125,
      "learning_rate": 4.554067995125023e-06,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8461382001638412,
      "num_tokens": 119698220.0,
      "step": 99550
    },
    {
      "entropy": 1.8459780648350717,
      "epoch": 0.3086272889947681,
      "grad_norm": 4.338291168212891,
      "learning_rate": 4.5538392773618436e-06,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8427159383893013,
      "num_tokens": 119710621.0,
      "step": 99560
    },
    {
      "entropy": 1.8610286951065063,
      "epoch": 0.3086582881198178,
      "grad_norm": 8.782474517822266,
      "learning_rate": 4.553610594055699e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8488268256187439,
      "num_tokens": 119722541.0,
      "step": 99570
    },
    {
      "entropy": 1.9070811703801156,
      "epoch": 0.3086892872448675,
      "grad_norm": 7.238033771514893,
      "learning_rate": 4.553381945197941e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.836867593228817,
      "num_tokens": 119733908.0,
      "step": 99580
    },
    {
      "entropy": 1.9172224178910255,
      "epoch": 0.3087202863699172,
      "grad_norm": 8.506465911865234,
      "learning_rate": 4.553153330779919e-06,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8402721121907234,
      "num_tokens": 119745237.0,
      "step": 99590
    },
    {
      "entropy": 1.8876484125852584,
      "epoch": 0.3087512854949669,
      "grad_norm": 4.413023471832275,
      "learning_rate": 4.552924750792989e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.846372652053833,
      "num_tokens": 119757146.0,
      "step": 99600
    },
    {
      "entropy": 1.8622125744819642,
      "epoch": 0.3087822846200166,
      "grad_norm": 8.20417308807373,
      "learning_rate": 4.552696205228509e-06,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8524843230843544,
      "num_tokens": 119769534.0,
      "step": 99610
    },
    {
      "entropy": 1.847460262477398,
      "epoch": 0.3088132837450663,
      "grad_norm": 7.062432289123535,
      "learning_rate": 4.552467694077842e-06,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8552860572934151,
      "num_tokens": 119781883.0,
      "step": 99620
    },
    {
      "entropy": 1.8279512047767639,
      "epoch": 0.308844282870116,
      "grad_norm": 3.6404471397399902,
      "learning_rate": 4.552239217332351e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8477238088846206,
      "num_tokens": 119795146.0,
      "step": 99630
    },
    {
      "entropy": 1.8767915710806846,
      "epoch": 0.3088752819951657,
      "grad_norm": 10.762784004211426,
      "learning_rate": 4.552010774983402e-06,
      "loss": 0.466,
      "mean_token_accuracy": 0.8552503928542137,
      "num_tokens": 119806803.0,
      "step": 99640
    },
    {
      "entropy": 1.8293926134705543,
      "epoch": 0.3089062811202154,
      "grad_norm": 7.491893291473389,
      "learning_rate": 4.551782367022367e-06,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8511452227830887,
      "num_tokens": 119819906.0,
      "step": 99650
    },
    {
      "entropy": 1.8923638671636582,
      "epoch": 0.30893728024526507,
      "grad_norm": 9.077742576599121,
      "learning_rate": 4.55155399344062e-06,
      "loss": 0.474,
      "mean_token_accuracy": 0.8456351578235626,
      "num_tokens": 119831777.0,
      "step": 99660
    },
    {
      "entropy": 1.9003022998571395,
      "epoch": 0.3089682793703148,
      "grad_norm": 8.246248245239258,
      "learning_rate": 4.551325654229535e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8550438165664673,
      "num_tokens": 119843799.0,
      "step": 99670
    },
    {
      "entropy": 1.8228402808308601,
      "epoch": 0.30899927849536446,
      "grad_norm": 5.1642746925354,
      "learning_rate": 4.551097349380495e-06,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8491297498345375,
      "num_tokens": 119856606.0,
      "step": 99680
    },
    {
      "entropy": 1.96132450401783,
      "epoch": 0.3090302776204142,
      "grad_norm": 9.276740074157715,
      "learning_rate": 4.550869078884878e-06,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8399526447057724,
      "num_tokens": 119867481.0,
      "step": 99690
    },
    {
      "entropy": 1.8656087532639503,
      "epoch": 0.30906127674546385,
      "grad_norm": 8.782885551452637,
      "learning_rate": 4.550640842734073e-06,
      "loss": 0.4616,
      "mean_token_accuracy": 0.8477922558784485,
      "num_tokens": 119879627.0,
      "step": 99700
    },
    {
      "entropy": 1.8200937718153,
      "epoch": 0.3090922758705136,
      "grad_norm": 9.573867797851562,
      "learning_rate": 4.550412640919468e-06,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8348228052258492,
      "num_tokens": 119892028.0,
      "step": 99710
    },
    {
      "entropy": 1.8446323826909066,
      "epoch": 0.30912327499556325,
      "grad_norm": 8.252911567687988,
      "learning_rate": 4.550184473432453e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8526568725705147,
      "num_tokens": 119904371.0,
      "step": 99720
    },
    {
      "entropy": 1.9776091009378434,
      "epoch": 0.30915427412061297,
      "grad_norm": 9.516276359558105,
      "learning_rate": 4.5499563402644234e-06,
      "loss": 0.5689,
      "mean_token_accuracy": 0.8284485876560211,
      "num_tokens": 119915090.0,
      "step": 99730
    },
    {
      "entropy": 1.9320985347032547,
      "epoch": 0.30918527324566264,
      "grad_norm": 9.039697647094727,
      "learning_rate": 4.5497282414067775e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.850306898355484,
      "num_tokens": 119926461.0,
      "step": 99740
    },
    {
      "entropy": 1.9415932521224022,
      "epoch": 0.30921627237071236,
      "grad_norm": 9.981196403503418,
      "learning_rate": 4.549500176850916e-06,
      "loss": 0.5159,
      "mean_token_accuracy": 0.848623238503933,
      "num_tokens": 119937402.0,
      "step": 99750
    },
    {
      "entropy": 1.8851657703518867,
      "epoch": 0.30924727149576203,
      "grad_norm": 6.050877571105957,
      "learning_rate": 4.549272146588241e-06,
      "loss": 0.525,
      "mean_token_accuracy": 0.8291679427027703,
      "num_tokens": 119948670.0,
      "step": 99760
    },
    {
      "entropy": 1.8636073276400567,
      "epoch": 0.3092782706208117,
      "grad_norm": 9.428133964538574,
      "learning_rate": 4.54904415061016e-06,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8505959510803223,
      "num_tokens": 119961140.0,
      "step": 99770
    },
    {
      "entropy": 1.8380408152937888,
      "epoch": 0.3093092697458614,
      "grad_norm": 8.414907455444336,
      "learning_rate": 4.548816188908081e-06,
      "loss": 0.501,
      "mean_token_accuracy": 0.839817276597023,
      "num_tokens": 119973426.0,
      "step": 99780
    },
    {
      "entropy": 1.8116330251097679,
      "epoch": 0.3093402688709111,
      "grad_norm": 8.415764808654785,
      "learning_rate": 4.548588261473421e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8535687699913979,
      "num_tokens": 119986183.0,
      "step": 99790
    },
    {
      "entropy": 1.8176366582512855,
      "epoch": 0.3093712679959608,
      "grad_norm": 9.07093620300293,
      "learning_rate": 4.548360368297591e-06,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8550222635269165,
      "num_tokens": 119998588.0,
      "step": 99800
    },
    {
      "entropy": 1.9179295137524606,
      "epoch": 0.3094022671210105,
      "grad_norm": 8.444205284118652,
      "learning_rate": 4.548132509372013e-06,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8391743138432503,
      "num_tokens": 120010012.0,
      "step": 99810
    },
    {
      "entropy": 1.8938566118478775,
      "epoch": 0.3094332662460602,
      "grad_norm": 8.31008529663086,
      "learning_rate": 4.5479046846881064e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8451041653752327,
      "num_tokens": 120022944.0,
      "step": 99820
    },
    {
      "entropy": 1.906507858633995,
      "epoch": 0.3094642653711099,
      "grad_norm": 8.747406959533691,
      "learning_rate": 4.547676894237297e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8478483647108078,
      "num_tokens": 120034662.0,
      "step": 99830
    },
    {
      "entropy": 1.8798761084675788,
      "epoch": 0.3094952644961596,
      "grad_norm": 9.963387489318848,
      "learning_rate": 4.547449138011013e-06,
      "loss": 0.466,
      "mean_token_accuracy": 0.8498962029814721,
      "num_tokens": 120047172.0,
      "step": 99840
    },
    {
      "entropy": 1.8998943746089936,
      "epoch": 0.3095262636212093,
      "grad_norm": 8.54676342010498,
      "learning_rate": 4.5472214160006844e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8516019433736801,
      "num_tokens": 120058507.0,
      "step": 99850
    },
    {
      "entropy": 1.8910057291388511,
      "epoch": 0.309557262746259,
      "grad_norm": 9.665603637695312,
      "learning_rate": 4.546993728197744e-06,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8422492250800133,
      "num_tokens": 120070406.0,
      "step": 99860
    },
    {
      "entropy": 1.9007139950990677,
      "epoch": 0.30958826187130867,
      "grad_norm": 8.319001197814941,
      "learning_rate": 4.546766074593631e-06,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8561020061373711,
      "num_tokens": 120082456.0,
      "step": 99870
    },
    {
      "entropy": 1.9839264631271363,
      "epoch": 0.3096192609963584,
      "grad_norm": 7.754944324493408,
      "learning_rate": 4.546538455179782e-06,
      "loss": 0.5336,
      "mean_token_accuracy": 0.836320398747921,
      "num_tokens": 120093393.0,
      "step": 99880
    },
    {
      "entropy": 1.827572251856327,
      "epoch": 0.30965026012140806,
      "grad_norm": 6.521658897399902,
      "learning_rate": 4.546310869947643e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8617355406284333,
      "num_tokens": 120106497.0,
      "step": 99890
    },
    {
      "entropy": 1.952468115091324,
      "epoch": 0.3096812592464578,
      "grad_norm": 7.609374523162842,
      "learning_rate": 4.546083318888656e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8570076480507851,
      "num_tokens": 120117150.0,
      "step": 99900
    },
    {
      "entropy": 1.9119912654161453,
      "epoch": 0.30971225837150745,
      "grad_norm": 8.479022979736328,
      "learning_rate": 4.5458558019942736e-06,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8490529209375381,
      "num_tokens": 120128766.0,
      "step": 99910
    },
    {
      "entropy": 1.8472444072365761,
      "epoch": 0.3097432574965572,
      "grad_norm": 8.135096549987793,
      "learning_rate": 4.5456283192559455e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8514307633042335,
      "num_tokens": 120140846.0,
      "step": 99920
    },
    {
      "entropy": 1.8320681855082512,
      "epoch": 0.30977425662160685,
      "grad_norm": 7.373263835906982,
      "learning_rate": 4.5454008706651255e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8532561391592026,
      "num_tokens": 120152671.0,
      "step": 99930
    },
    {
      "entropy": 1.8469135209918022,
      "epoch": 0.30980525574665657,
      "grad_norm": 6.601968288421631,
      "learning_rate": 4.545173456213272e-06,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8549439549446106,
      "num_tokens": 120165189.0,
      "step": 99940
    },
    {
      "entropy": 1.8578027948737144,
      "epoch": 0.30983625487170624,
      "grad_norm": 9.094026565551758,
      "learning_rate": 4.5449460758918485e-06,
      "loss": 0.4063,
      "mean_token_accuracy": 0.8560818761587143,
      "num_tokens": 120177445.0,
      "step": 99950
    },
    {
      "entropy": 1.8183515168726445,
      "epoch": 0.30986725399675596,
      "grad_norm": 8.197356224060059,
      "learning_rate": 4.544718729692315e-06,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8482865273952485,
      "num_tokens": 120190572.0,
      "step": 99960
    },
    {
      "entropy": 1.8944060549139976,
      "epoch": 0.30989825312180563,
      "grad_norm": 8.696087837219238,
      "learning_rate": 4.544491417606139e-06,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8414214491844177,
      "num_tokens": 120202087.0,
      "step": 99970
    },
    {
      "entropy": 1.7967160180211068,
      "epoch": 0.30992925224685536,
      "grad_norm": 8.540863990783691,
      "learning_rate": 4.544264139624791e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8545686945319175,
      "num_tokens": 120214607.0,
      "step": 99980
    },
    {
      "entropy": 1.814838644862175,
      "epoch": 0.309960251371905,
      "grad_norm": 6.886814594268799,
      "learning_rate": 4.544036895739743e-06,
      "loss": 0.42,
      "mean_token_accuracy": 0.8569683223962784,
      "num_tokens": 120227295.0,
      "step": 99990
    },
    {
      "entropy": 1.8500340178608894,
      "epoch": 0.30999125049695475,
      "grad_norm": 4.226168632507324,
      "learning_rate": 4.5438096859424714e-06,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8484739258885383,
      "num_tokens": 120240549.0,
      "step": 100000
    },
    {
      "entropy": 1.7947644501924516,
      "epoch": 0.3100222496220044,
      "grad_norm": 4.175483703613281,
      "learning_rate": 4.543582510224454e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8510774120688438,
      "num_tokens": 120253357.0,
      "step": 100010
    },
    {
      "entropy": 1.8611189171671867,
      "epoch": 0.3100532487470541,
      "grad_norm": 11.313339233398438,
      "learning_rate": 4.543355368577173e-06,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8596706509590148,
      "num_tokens": 120265081.0,
      "step": 100020
    },
    {
      "entropy": 1.8015659287571908,
      "epoch": 0.3100842478721038,
      "grad_norm": 8.804397583007812,
      "learning_rate": 4.543128260992112e-06,
      "loss": 0.4095,
      "mean_token_accuracy": 0.8543227300047874,
      "num_tokens": 120278300.0,
      "step": 100030
    },
    {
      "entropy": 1.8594049900770186,
      "epoch": 0.3101152469971535,
      "grad_norm": 9.17402458190918,
      "learning_rate": 4.54290118746076e-06,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8479228600859642,
      "num_tokens": 120289822.0,
      "step": 100040
    },
    {
      "entropy": 1.808043359220028,
      "epoch": 0.3101462461222032,
      "grad_norm": 7.376701831817627,
      "learning_rate": 4.542674147974606e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8577449038624764,
      "num_tokens": 120302594.0,
      "step": 100050
    },
    {
      "entropy": 1.9016132436692714,
      "epoch": 0.3101772452472529,
      "grad_norm": 8.157571792602539,
      "learning_rate": 4.5424471425251435e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8419206649065017,
      "num_tokens": 120314205.0,
      "step": 100060
    },
    {
      "entropy": 1.8275883719325066,
      "epoch": 0.3102082443723026,
      "grad_norm": 8.187117576599121,
      "learning_rate": 4.542220171103871e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8554164975881576,
      "num_tokens": 120327120.0,
      "step": 100070
    },
    {
      "entropy": 1.8765954077243805,
      "epoch": 0.31023924349735227,
      "grad_norm": 2.51336932182312,
      "learning_rate": 4.541993233702286e-06,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8441577076911926,
      "num_tokens": 120338971.0,
      "step": 100080
    },
    {
      "entropy": 1.8348333820700646,
      "epoch": 0.310270242622402,
      "grad_norm": 9.080048561096191,
      "learning_rate": 4.541766330311893e-06,
      "loss": 0.4552,
      "mean_token_accuracy": 0.8454143956303597,
      "num_tokens": 120351609.0,
      "step": 100090
    },
    {
      "entropy": 1.8745562806725502,
      "epoch": 0.31030124174745166,
      "grad_norm": 7.322804927825928,
      "learning_rate": 4.541539460924194e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8521630093455315,
      "num_tokens": 120363483.0,
      "step": 100100
    },
    {
      "entropy": 1.899325506389141,
      "epoch": 0.3103322408725014,
      "grad_norm": 9.89995002746582,
      "learning_rate": 4.541312625530701e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8446615591645241,
      "num_tokens": 120374672.0,
      "step": 100110
    },
    {
      "entropy": 1.8418832674622536,
      "epoch": 0.31036323999755105,
      "grad_norm": 7.776636123657227,
      "learning_rate": 4.541085824122922e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8611134603619576,
      "num_tokens": 120386805.0,
      "step": 100120
    },
    {
      "entropy": 1.908632791042328,
      "epoch": 0.3103942391226008,
      "grad_norm": 9.52039909362793,
      "learning_rate": 4.540859056692375e-06,
      "loss": 0.5108,
      "mean_token_accuracy": 0.8384132295846939,
      "num_tokens": 120398539.0,
      "step": 100130
    },
    {
      "entropy": 1.9329130694270134,
      "epoch": 0.31042523824765045,
      "grad_norm": 8.478260040283203,
      "learning_rate": 4.540632323230573e-06,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8484991729259491,
      "num_tokens": 120409458.0,
      "step": 100140
    },
    {
      "entropy": 1.9148776784539223,
      "epoch": 0.3104562373727002,
      "grad_norm": 8.679959297180176,
      "learning_rate": 4.54040562372904e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.8426013261079788,
      "num_tokens": 120420579.0,
      "step": 100150
    },
    {
      "entropy": 1.8648797929286958,
      "epoch": 0.31048723649774984,
      "grad_norm": 4.328078269958496,
      "learning_rate": 4.5401789581792985e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.846242044866085,
      "num_tokens": 120432678.0,
      "step": 100160
    },
    {
      "entropy": 1.8050495654344558,
      "epoch": 0.31051823562279957,
      "grad_norm": 9.263623237609863,
      "learning_rate": 4.539952326572873e-06,
      "loss": 0.4235,
      "mean_token_accuracy": 0.8572616815567017,
      "num_tokens": 120446167.0,
      "step": 100170
    },
    {
      "entropy": 2.000722700357437,
      "epoch": 0.31054923474784923,
      "grad_norm": 7.7509355545043945,
      "learning_rate": 4.539725728901292e-06,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8357238873839379,
      "num_tokens": 120457079.0,
      "step": 100180
    },
    {
      "entropy": 1.8446622014045715,
      "epoch": 0.31058023387289896,
      "grad_norm": 7.534979820251465,
      "learning_rate": 4.539499165156091e-06,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8397464141249656,
      "num_tokens": 120469823.0,
      "step": 100190
    },
    {
      "entropy": 1.8747181564569473,
      "epoch": 0.3106112329979486,
      "grad_norm": 3.7008886337280273,
      "learning_rate": 4.539272635328803e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.846717146039009,
      "num_tokens": 120482088.0,
      "step": 100200
    },
    {
      "entropy": 1.8778936117887497,
      "epoch": 0.31064223212299835,
      "grad_norm": 8.01706314086914,
      "learning_rate": 4.539046139410965e-06,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8487760618329048,
      "num_tokens": 120494848.0,
      "step": 100210
    },
    {
      "entropy": 1.8788290441036224,
      "epoch": 0.310673231248048,
      "grad_norm": 7.9294915199279785,
      "learning_rate": 4.53881967739412e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8471907436847687,
      "num_tokens": 120506578.0,
      "step": 100220
    },
    {
      "entropy": 1.8919280216097831,
      "epoch": 0.31070423037309775,
      "grad_norm": 4.634057521820068,
      "learning_rate": 4.53859324926981e-06,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8390332102775574,
      "num_tokens": 120518962.0,
      "step": 100230
    },
    {
      "entropy": 1.73563092648983,
      "epoch": 0.3107352294981474,
      "grad_norm": 6.966121196746826,
      "learning_rate": 4.538366855029584e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8594012647867203,
      "num_tokens": 120532862.0,
      "step": 100240
    },
    {
      "entropy": 1.8883020922541618,
      "epoch": 0.3107662286231971,
      "grad_norm": 2.8376030921936035,
      "learning_rate": 4.53814049466499e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8497428327798844,
      "num_tokens": 120544852.0,
      "step": 100250
    },
    {
      "entropy": 1.980030670762062,
      "epoch": 0.3107972277482468,
      "grad_norm": 9.871118545532227,
      "learning_rate": 4.537914168167582e-06,
      "loss": 0.517,
      "mean_token_accuracy": 0.8421619832515717,
      "num_tokens": 120555617.0,
      "step": 100260
    },
    {
      "entropy": 1.9399761855602264,
      "epoch": 0.3108282268732965,
      "grad_norm": 8.313036918640137,
      "learning_rate": 4.5376878755289136e-06,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8462784796953201,
      "num_tokens": 120567039.0,
      "step": 100270
    },
    {
      "entropy": 1.8688796371221543,
      "epoch": 0.3108592259983462,
      "grad_norm": 8.893397331237793,
      "learning_rate": 4.537461616740546e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8576091736555099,
      "num_tokens": 120579341.0,
      "step": 100280
    },
    {
      "entropy": 1.902301235496998,
      "epoch": 0.31089022512339587,
      "grad_norm": 8.338861465454102,
      "learning_rate": 4.53723539179404e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8621632814407348,
      "num_tokens": 120590733.0,
      "step": 100290
    },
    {
      "entropy": 1.9393490612506867,
      "epoch": 0.3109212242484456,
      "grad_norm": 7.925196170806885,
      "learning_rate": 4.53700920068096e-06,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8450248718261719,
      "num_tokens": 120601534.0,
      "step": 100300
    },
    {
      "entropy": 1.8962682634592056,
      "epoch": 0.31095222337349526,
      "grad_norm": 7.99591064453125,
      "learning_rate": 4.536783043392873e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8482768073678016,
      "num_tokens": 120612997.0,
      "step": 100310
    },
    {
      "entropy": 1.9093605667352676,
      "epoch": 0.310983222498545,
      "grad_norm": 8.260348320007324,
      "learning_rate": 4.536556919921349e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8399380281567573,
      "num_tokens": 120624613.0,
      "step": 100320
    },
    {
      "entropy": 1.8416539490222932,
      "epoch": 0.31101422162359466,
      "grad_norm": 4.14152193069458,
      "learning_rate": 4.536330830257964e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8592947632074356,
      "num_tokens": 120637123.0,
      "step": 100330
    },
    {
      "entropy": 1.9081314831972123,
      "epoch": 0.3110452207486444,
      "grad_norm": 10.202445030212402,
      "learning_rate": 4.536104774394291e-06,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8314228966832161,
      "num_tokens": 120648950.0,
      "step": 100340
    },
    {
      "entropy": 1.8932959333062171,
      "epoch": 0.31107621987369405,
      "grad_norm": 9.009939193725586,
      "learning_rate": 4.5358787523219115e-06,
      "loss": 0.4466,
      "mean_token_accuracy": 0.8498683959245682,
      "num_tokens": 120660896.0,
      "step": 100350
    },
    {
      "entropy": 1.8493494719266892,
      "epoch": 0.3111072189987438,
      "grad_norm": 3.8888235092163086,
      "learning_rate": 4.535652764032407e-06,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8555730670690537,
      "num_tokens": 120673291.0,
      "step": 100360
    },
    {
      "entropy": 1.8772372379899025,
      "epoch": 0.31113821812379344,
      "grad_norm": 8.41147232055664,
      "learning_rate": 4.535426809517363e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.844378887116909,
      "num_tokens": 120685691.0,
      "step": 100370
    },
    {
      "entropy": 1.86507987678051,
      "epoch": 0.31116921724884317,
      "grad_norm": 8.958087921142578,
      "learning_rate": 4.535200888768366e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8503421053290368,
      "num_tokens": 120697935.0,
      "step": 100380
    },
    {
      "entropy": 1.849946916103363,
      "epoch": 0.31120021637389284,
      "grad_norm": 7.368825435638428,
      "learning_rate": 4.534975001777008e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8594206288456917,
      "num_tokens": 120709752.0,
      "step": 100390
    },
    {
      "entropy": 1.859765648841858,
      "epoch": 0.31123121549894256,
      "grad_norm": 7.649475574493408,
      "learning_rate": 4.5347491485348835e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8555939674377442,
      "num_tokens": 120721721.0,
      "step": 100400
    },
    {
      "entropy": 1.877078004181385,
      "epoch": 0.31126221462399223,
      "grad_norm": 8.297224998474121,
      "learning_rate": 4.534523329033589e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8504548445343971,
      "num_tokens": 120732890.0,
      "step": 100410
    },
    {
      "entropy": 1.8714374005794525,
      "epoch": 0.31129321374904195,
      "grad_norm": 4.207152366638184,
      "learning_rate": 4.534297543264725e-06,
      "loss": 0.5621,
      "mean_token_accuracy": 0.8266729831695556,
      "num_tokens": 120745682.0,
      "step": 100420
    },
    {
      "entropy": 1.8621924549341202,
      "epoch": 0.3113242128740916,
      "grad_norm": 8.956070899963379,
      "learning_rate": 4.534071791219892e-06,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8492157995700836,
      "num_tokens": 120758011.0,
      "step": 100430
    },
    {
      "entropy": 1.9551179885864258,
      "epoch": 0.31135521199914135,
      "grad_norm": 8.54804515838623,
      "learning_rate": 4.533846072890697e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.83407664000988,
      "num_tokens": 120768890.0,
      "step": 100440
    },
    {
      "entropy": 1.8986855819821358,
      "epoch": 0.311386211124191,
      "grad_norm": 8.727336883544922,
      "learning_rate": 4.533620388268749e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8469956681132317,
      "num_tokens": 120780472.0,
      "step": 100450
    },
    {
      "entropy": 1.8609646454453468,
      "epoch": 0.31141721024924074,
      "grad_norm": 7.309536457061768,
      "learning_rate": 4.533394737345659e-06,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8551263570785522,
      "num_tokens": 120792295.0,
      "step": 100460
    },
    {
      "entropy": 1.8796255081892013,
      "epoch": 0.3114482093742904,
      "grad_norm": 10.224238395690918,
      "learning_rate": 4.5331691201130415e-06,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8438882201910018,
      "num_tokens": 120804242.0,
      "step": 100470
    },
    {
      "entropy": 1.907512204349041,
      "epoch": 0.31147920849934013,
      "grad_norm": 10.103514671325684,
      "learning_rate": 4.532943536562514e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8437295973300933,
      "num_tokens": 120815319.0,
      "step": 100480
    },
    {
      "entropy": 1.8957229569554328,
      "epoch": 0.3115102076243898,
      "grad_norm": 5.280575752258301,
      "learning_rate": 4.5327179866856965e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8484357133507728,
      "num_tokens": 120827094.0,
      "step": 100490
    },
    {
      "entropy": 1.856425480544567,
      "epoch": 0.31154120674943947,
      "grad_norm": 8.965953826904297,
      "learning_rate": 4.532492470474212e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8441311463713645,
      "num_tokens": 120839247.0,
      "step": 100500
    },
    {
      "entropy": 1.8147829815745353,
      "epoch": 0.3115722058744892,
      "grad_norm": 9.263935089111328,
      "learning_rate": 4.532266987919687e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8509012356400489,
      "num_tokens": 120852066.0,
      "step": 100510
    },
    {
      "entropy": 1.9309598222374915,
      "epoch": 0.31160320499953886,
      "grad_norm": 7.9816975593566895,
      "learning_rate": 4.53204153901375e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8446980267763138,
      "num_tokens": 120863149.0,
      "step": 100520
    },
    {
      "entropy": 1.8563687920570373,
      "epoch": 0.3116342041245886,
      "grad_norm": 8.456686973571777,
      "learning_rate": 4.531816123748033e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8485777124762535,
      "num_tokens": 120875622.0,
      "step": 100530
    },
    {
      "entropy": 1.9105036780238152,
      "epoch": 0.31166520324963826,
      "grad_norm": 8.747064590454102,
      "learning_rate": 4.531590742114171e-06,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8524237051606178,
      "num_tokens": 120886828.0,
      "step": 100540
    },
    {
      "entropy": 1.9014281533658504,
      "epoch": 0.311696202374688,
      "grad_norm": 10.515109062194824,
      "learning_rate": 4.531365394103802e-06,
      "loss": 0.491,
      "mean_token_accuracy": 0.8461366832256317,
      "num_tokens": 120898889.0,
      "step": 100550
    },
    {
      "entropy": 1.7405464336276055,
      "epoch": 0.31172720149973765,
      "grad_norm": 4.015621662139893,
      "learning_rate": 4.531140079708566e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8626249134540558,
      "num_tokens": 120913012.0,
      "step": 100560
    },
    {
      "entropy": 1.875110612809658,
      "epoch": 0.3117582006247874,
      "grad_norm": 9.846309661865234,
      "learning_rate": 4.530914798920107e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8495632246136665,
      "num_tokens": 120924428.0,
      "step": 100570
    },
    {
      "entropy": 1.7907123163342475,
      "epoch": 0.31178919974983704,
      "grad_norm": 8.938268661499023,
      "learning_rate": 4.530689551730072e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8574684172868728,
      "num_tokens": 120937451.0,
      "step": 100580
    },
    {
      "entropy": 1.9089242190122604,
      "epoch": 0.31182019887488677,
      "grad_norm": 10.029963493347168,
      "learning_rate": 4.5304643381301094e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8454339489340782,
      "num_tokens": 120948860.0,
      "step": 100590
    },
    {
      "entropy": 1.9048386812210083,
      "epoch": 0.31185119799993644,
      "grad_norm": 8.10409927368164,
      "learning_rate": 4.530239158111872e-06,
      "loss": 0.4551,
      "mean_token_accuracy": 0.8487098172307015,
      "num_tokens": 120960368.0,
      "step": 100600
    },
    {
      "entropy": 1.7758634328842162,
      "epoch": 0.31188219712498616,
      "grad_norm": 3.4512832164764404,
      "learning_rate": 4.530014011667015e-06,
      "loss": 0.379,
      "mean_token_accuracy": 0.8484735369682312,
      "num_tokens": 120973621.0,
      "step": 100610
    },
    {
      "entropy": 1.9328245520591736,
      "epoch": 0.31191319625003583,
      "grad_norm": 9.077108383178711,
      "learning_rate": 4.5297888987871956e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8441737592220306,
      "num_tokens": 120984223.0,
      "step": 100620
    },
    {
      "entropy": 1.8256781995296478,
      "epoch": 0.31194419537508555,
      "grad_norm": 7.908056735992432,
      "learning_rate": 4.529563819464075e-06,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8539151340723038,
      "num_tokens": 120997133.0,
      "step": 100630
    },
    {
      "entropy": 1.8427986733615398,
      "epoch": 0.3119751945001352,
      "grad_norm": 9.019706726074219,
      "learning_rate": 4.529338773689319e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8452917739748955,
      "num_tokens": 121009965.0,
      "step": 100640
    },
    {
      "entropy": 1.912809681892395,
      "epoch": 0.31200619362518495,
      "grad_norm": 8.718803405761719,
      "learning_rate": 4.529113761454591e-06,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8374342992901802,
      "num_tokens": 121021538.0,
      "step": 100650
    },
    {
      "entropy": 1.8587751030921935,
      "epoch": 0.3120371927502346,
      "grad_norm": 8.5054931640625,
      "learning_rate": 4.528888782751565e-06,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8497788742184639,
      "num_tokens": 121033842.0,
      "step": 100660
    },
    {
      "entropy": 1.8983284369111062,
      "epoch": 0.31206819187528434,
      "grad_norm": 8.823607444763184,
      "learning_rate": 4.52866383757191e-06,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8428234368562698,
      "num_tokens": 121045774.0,
      "step": 100670
    },
    {
      "entropy": 1.9349606305360794,
      "epoch": 0.312099191000334,
      "grad_norm": 8.035542488098145,
      "learning_rate": 4.528438925907303e-06,
      "loss": 0.5277,
      "mean_token_accuracy": 0.8406255826354027,
      "num_tokens": 121057292.0,
      "step": 100680
    },
    {
      "entropy": 1.8452520370483398,
      "epoch": 0.31213019012538373,
      "grad_norm": 9.03587818145752,
      "learning_rate": 4.528214047749422e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8514409214258194,
      "num_tokens": 121070931.0,
      "step": 100690
    },
    {
      "entropy": 1.7869605004787446,
      "epoch": 0.3121611892504334,
      "grad_norm": 10.495682716369629,
      "learning_rate": 4.5279892030899485e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8594531282782555,
      "num_tokens": 121083578.0,
      "step": 100700
    },
    {
      "entropy": 1.8690698593854904,
      "epoch": 0.3121921883754831,
      "grad_norm": 8.482170104980469,
      "learning_rate": 4.527764391920566e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8500708416104317,
      "num_tokens": 121095247.0,
      "step": 100710
    },
    {
      "entropy": 1.9164055377244948,
      "epoch": 0.3122231875005328,
      "grad_norm": 8.222436904907227,
      "learning_rate": 4.527539614232962e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8398829951882363,
      "num_tokens": 121107205.0,
      "step": 100720
    },
    {
      "entropy": 1.8966273233294486,
      "epoch": 0.3122541866255825,
      "grad_norm": 10.007433891296387,
      "learning_rate": 4.527314870018826e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8499047040939331,
      "num_tokens": 121118189.0,
      "step": 100730
    },
    {
      "entropy": 1.9177552998065948,
      "epoch": 0.3122851857506322,
      "grad_norm": 9.334546089172363,
      "learning_rate": 4.527090159269853e-06,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8446200251579284,
      "num_tokens": 121129883.0,
      "step": 100740
    },
    {
      "entropy": 1.951483702659607,
      "epoch": 0.31231618487568186,
      "grad_norm": 9.152246475219727,
      "learning_rate": 4.5268654819777355e-06,
      "loss": 0.5432,
      "mean_token_accuracy": 0.8399815455079078,
      "num_tokens": 121140945.0,
      "step": 100750
    },
    {
      "entropy": 1.843865168094635,
      "epoch": 0.3123471840007316,
      "grad_norm": 6.65300178527832,
      "learning_rate": 4.5266408381341735e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.846330837905407,
      "num_tokens": 121153921.0,
      "step": 100760
    },
    {
      "entropy": 1.9063566982746125,
      "epoch": 0.31237818312578125,
      "grad_norm": 7.834247589111328,
      "learning_rate": 4.526416227730868e-06,
      "loss": 0.549,
      "mean_token_accuracy": 0.8455073088407516,
      "num_tokens": 121165017.0,
      "step": 100770
    },
    {
      "entropy": 1.8952314227819442,
      "epoch": 0.312409182250831,
      "grad_norm": 8.009584426879883,
      "learning_rate": 4.526191650759525e-06,
      "loss": 0.4638,
      "mean_token_accuracy": 0.846976974606514,
      "num_tokens": 121176827.0,
      "step": 100780
    },
    {
      "entropy": 1.9277762562036513,
      "epoch": 0.31244018137588064,
      "grad_norm": 8.662957191467285,
      "learning_rate": 4.52596710721185e-06,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8444670990109444,
      "num_tokens": 121187950.0,
      "step": 100790
    },
    {
      "entropy": 1.943731963634491,
      "epoch": 0.31247118050093037,
      "grad_norm": 7.965757846832275,
      "learning_rate": 4.525742597079554e-06,
      "loss": 0.529,
      "mean_token_accuracy": 0.8364956364035606,
      "num_tokens": 121199219.0,
      "step": 100800
    },
    {
      "entropy": 1.9287627547979356,
      "epoch": 0.31250217962598004,
      "grad_norm": 6.632516384124756,
      "learning_rate": 4.52551812035435e-06,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8509150311350823,
      "num_tokens": 121210268.0,
      "step": 100810
    },
    {
      "entropy": 1.9134871885180473,
      "epoch": 0.31253317875102976,
      "grad_norm": 4.123570919036865,
      "learning_rate": 4.525293677027954e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.8477164819836617,
      "num_tokens": 121222196.0,
      "step": 100820
    },
    {
      "entropy": 1.916480553150177,
      "epoch": 0.31256417787607943,
      "grad_norm": 10.33231258392334,
      "learning_rate": 4.525069267092083e-06,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8356222257018089,
      "num_tokens": 121233821.0,
      "step": 100830
    },
    {
      "entropy": 1.8253281712532043,
      "epoch": 0.31259517700112915,
      "grad_norm": 8.607940673828125,
      "learning_rate": 4.52484489053846e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8448203921318054,
      "num_tokens": 121246155.0,
      "step": 100840
    },
    {
      "entropy": 1.9012310445308684,
      "epoch": 0.3126261761261788,
      "grad_norm": 7.7749247550964355,
      "learning_rate": 4.524620547358811e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8430942222476006,
      "num_tokens": 121258155.0,
      "step": 100850
    },
    {
      "entropy": 1.9564151272177697,
      "epoch": 0.31265717525122855,
      "grad_norm": 8.548589706420898,
      "learning_rate": 4.52439623754486e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8461882799863816,
      "num_tokens": 121269354.0,
      "step": 100860
    },
    {
      "entropy": 1.8889728412032127,
      "epoch": 0.3126881743762782,
      "grad_norm": 5.758060455322266,
      "learning_rate": 4.524171961088339e-06,
      "loss": 0.4714,
      "mean_token_accuracy": 0.857310351729393,
      "num_tokens": 121281304.0,
      "step": 100870
    },
    {
      "entropy": 1.930538135766983,
      "epoch": 0.31271917350132794,
      "grad_norm": 8.554981231689453,
      "learning_rate": 4.523947717980982e-06,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8413084149360657,
      "num_tokens": 121293664.0,
      "step": 100880
    },
    {
      "entropy": 1.950375673174858,
      "epoch": 0.3127501726263776,
      "grad_norm": 8.186065673828125,
      "learning_rate": 4.5237235082145235e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.849441209435463,
      "num_tokens": 121304745.0,
      "step": 100890
    },
    {
      "entropy": 1.9061787739396094,
      "epoch": 0.31278117175142733,
      "grad_norm": 3.3713457584381104,
      "learning_rate": 4.523499331780703e-06,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8370168238878251,
      "num_tokens": 121316310.0,
      "step": 100900
    },
    {
      "entropy": 1.830712193250656,
      "epoch": 0.312812170876477,
      "grad_norm": 4.071151256561279,
      "learning_rate": 4.5232751886712615e-06,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8477034792304039,
      "num_tokens": 121328915.0,
      "step": 100910
    },
    {
      "entropy": 1.9675110548734664,
      "epoch": 0.3128431700015267,
      "grad_norm": 8.07413101196289,
      "learning_rate": 4.523051078877946e-06,
      "loss": 0.5416,
      "mean_token_accuracy": 0.8389467343688011,
      "num_tokens": 121339986.0,
      "step": 100920
    },
    {
      "entropy": 1.8929841727018357,
      "epoch": 0.3128741691265764,
      "grad_norm": 8.73824691772461,
      "learning_rate": 4.5228270023925e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8490571796894073,
      "num_tokens": 121352037.0,
      "step": 100930
    },
    {
      "entropy": 1.8892315909266473,
      "epoch": 0.3129051682516261,
      "grad_norm": 4.492147922515869,
      "learning_rate": 4.522602959206678e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8552650704979896,
      "num_tokens": 121364144.0,
      "step": 100940
    },
    {
      "entropy": 1.923583671450615,
      "epoch": 0.3129361673766758,
      "grad_norm": 10.050043106079102,
      "learning_rate": 4.52237894931223e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.844260835647583,
      "num_tokens": 121375808.0,
      "step": 100950
    },
    {
      "entropy": 1.9433955758810044,
      "epoch": 0.3129671665017255,
      "grad_norm": 9.592854499816895,
      "learning_rate": 4.522154972700912e-06,
      "loss": 0.5002,
      "mean_token_accuracy": 0.8481174424290657,
      "num_tokens": 121387011.0,
      "step": 100960
    },
    {
      "entropy": 1.847121460735798,
      "epoch": 0.3129981656267752,
      "grad_norm": 8.493393898010254,
      "learning_rate": 4.5219310293644856e-06,
      "loss": 0.4003,
      "mean_token_accuracy": 0.8667575031518936,
      "num_tokens": 121399217.0,
      "step": 100970
    },
    {
      "entropy": 1.895681183040142,
      "epoch": 0.3130291647518249,
      "grad_norm": 8.458040237426758,
      "learning_rate": 4.52170711929471e-06,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8476063832640648,
      "num_tokens": 121411380.0,
      "step": 100980
    },
    {
      "entropy": 1.8396486029028893,
      "epoch": 0.3130601638768746,
      "grad_norm": 8.87826156616211,
      "learning_rate": 4.521483242483351e-06,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8522458449006081,
      "num_tokens": 121424542.0,
      "step": 100990
    },
    {
      "entropy": 1.8425541028380394,
      "epoch": 0.31309116300192424,
      "grad_norm": 8.997958183288574,
      "learning_rate": 4.521259398922175e-06,
      "loss": 0.4341,
      "mean_token_accuracy": 0.8560503959655762,
      "num_tokens": 121437235.0,
      "step": 101000
    },
    {
      "entropy": 1.857857908308506,
      "epoch": 0.31312216212697397,
      "grad_norm": 9.751863479614258,
      "learning_rate": 4.521035588602953e-06,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8419872790575027,
      "num_tokens": 121449522.0,
      "step": 101010
    },
    {
      "entropy": 1.858136995136738,
      "epoch": 0.31315316125202364,
      "grad_norm": 3.528123617172241,
      "learning_rate": 4.520811811517458e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8612496837973594,
      "num_tokens": 121462199.0,
      "step": 101020
    },
    {
      "entropy": 1.907412128150463,
      "epoch": 0.31318416037707336,
      "grad_norm": 6.3340744972229,
      "learning_rate": 4.520588067657467e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8482323855161666,
      "num_tokens": 121474247.0,
      "step": 101030
    },
    {
      "entropy": 1.8920990169048308,
      "epoch": 0.31321515950212303,
      "grad_norm": 7.975953578948975,
      "learning_rate": 4.520364357014758e-06,
      "loss": 0.4311,
      "mean_token_accuracy": 0.8457095667719841,
      "num_tokens": 121486893.0,
      "step": 101040
    },
    {
      "entropy": 1.8772882983088492,
      "epoch": 0.31324615862717275,
      "grad_norm": 8.792813301086426,
      "learning_rate": 4.520140679581111e-06,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8604818403720855,
      "num_tokens": 121499181.0,
      "step": 101050
    },
    {
      "entropy": 1.8207586660981179,
      "epoch": 0.3132771577522224,
      "grad_norm": 7.4785590171813965,
      "learning_rate": 4.519917035348314e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8638281837105751,
      "num_tokens": 121511510.0,
      "step": 101060
    },
    {
      "entropy": 1.847452338039875,
      "epoch": 0.31330815687727215,
      "grad_norm": 4.0800886154174805,
      "learning_rate": 4.519693424308152e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8405447885394096,
      "num_tokens": 121524110.0,
      "step": 101070
    },
    {
      "entropy": 1.9054955273866654,
      "epoch": 0.3133391560023218,
      "grad_norm": 5.289824485778809,
      "learning_rate": 4.519469846452415e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8457769706845284,
      "num_tokens": 121535699.0,
      "step": 101080
    },
    {
      "entropy": 1.915067094564438,
      "epoch": 0.31337015512737154,
      "grad_norm": 8.302640914916992,
      "learning_rate": 4.519246301772896e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8345979511737823,
      "num_tokens": 121547506.0,
      "step": 101090
    },
    {
      "entropy": 1.8765003278851509,
      "epoch": 0.3134011542524212,
      "grad_norm": 9.370650291442871,
      "learning_rate": 4.519022790261393e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8564626336097717,
      "num_tokens": 121560190.0,
      "step": 101100
    },
    {
      "entropy": 1.6999532222747802,
      "epoch": 0.31343215337747093,
      "grad_norm": 3.211577892303467,
      "learning_rate": 4.5187993119097045e-06,
      "loss": 0.3213,
      "mean_token_accuracy": 0.8762291237711907,
      "num_tokens": 121574646.0,
      "step": 101110
    },
    {
      "entropy": 1.8197277091443538,
      "epoch": 0.3134631525025206,
      "grad_norm": 7.400459289550781,
      "learning_rate": 4.5185758667096295e-06,
      "loss": 0.3932,
      "mean_token_accuracy": 0.8629156649112701,
      "num_tokens": 121588419.0,
      "step": 101120
    },
    {
      "entropy": 1.9415106505155564,
      "epoch": 0.3134941516275703,
      "grad_norm": 9.660387992858887,
      "learning_rate": 4.518352454652974e-06,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8417216569185257,
      "num_tokens": 121599422.0,
      "step": 101130
    },
    {
      "entropy": 1.9480364575982094,
      "epoch": 0.31352515075262,
      "grad_norm": 9.710609436035156,
      "learning_rate": 4.518129075731546e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8369411841034889,
      "num_tokens": 121610421.0,
      "step": 101140
    },
    {
      "entropy": 1.9201453104615211,
      "epoch": 0.3135561498776697,
      "grad_norm": 3.905876874923706,
      "learning_rate": 4.517905729937153e-06,
      "loss": 0.5488,
      "mean_token_accuracy": 0.8450462609529495,
      "num_tokens": 121622409.0,
      "step": 101150
    },
    {
      "entropy": 1.9243328019976615,
      "epoch": 0.3135871490027194,
      "grad_norm": 8.177422523498535,
      "learning_rate": 4.517682417261611e-06,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8475325971841812,
      "num_tokens": 121634012.0,
      "step": 101160
    },
    {
      "entropy": 1.8541806608438491,
      "epoch": 0.3136181481277691,
      "grad_norm": 9.121359825134277,
      "learning_rate": 4.517459137696734e-06,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8584785044193268,
      "num_tokens": 121646345.0,
      "step": 101170
    },
    {
      "entropy": 1.8980771958827973,
      "epoch": 0.3136491472528188,
      "grad_norm": 8.328401565551758,
      "learning_rate": 4.517235891234341e-06,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8456721261143685,
      "num_tokens": 121657741.0,
      "step": 101180
    },
    {
      "entropy": 1.8372810363769532,
      "epoch": 0.3136801463778685,
      "grad_norm": 9.05524730682373,
      "learning_rate": 4.517012677866254e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.857391194999218,
      "num_tokens": 121670981.0,
      "step": 101190
    },
    {
      "entropy": 1.8419764667749405,
      "epoch": 0.3137111455029182,
      "grad_norm": 3.9297235012054443,
      "learning_rate": 4.516789497584297e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8536780118942261,
      "num_tokens": 121682826.0,
      "step": 101200
    },
    {
      "entropy": 1.8321501910686493,
      "epoch": 0.3137421446279679,
      "grad_norm": 4.388725757598877,
      "learning_rate": 4.516566350380297e-06,
      "loss": 0.4487,
      "mean_token_accuracy": 0.852819861471653,
      "num_tokens": 121695953.0,
      "step": 101210
    },
    {
      "entropy": 1.8966490373015403,
      "epoch": 0.31377314375301757,
      "grad_norm": 4.4960737228393555,
      "learning_rate": 4.5163432362460825e-06,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8485043302178383,
      "num_tokens": 121707846.0,
      "step": 101220
    },
    {
      "entropy": 1.8889821529388429,
      "epoch": 0.3138041428780673,
      "grad_norm": 9.122937202453613,
      "learning_rate": 4.516120155173487e-06,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8423930436372757,
      "num_tokens": 121719869.0,
      "step": 101230
    },
    {
      "entropy": 1.832286487519741,
      "epoch": 0.31383514200311696,
      "grad_norm": 3.832523822784424,
      "learning_rate": 4.515897107154348e-06,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8503172323107719,
      "num_tokens": 121732133.0,
      "step": 101240
    },
    {
      "entropy": 1.8415155798196792,
      "epoch": 0.31386614112816663,
      "grad_norm": 3.9735727310180664,
      "learning_rate": 4.515674092180501e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.848311547935009,
      "num_tokens": 121744373.0,
      "step": 101250
    },
    {
      "entropy": 1.9100284710526467,
      "epoch": 0.31389714025321636,
      "grad_norm": 8.413751602172852,
      "learning_rate": 4.51545111024379e-06,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8440921515226364,
      "num_tokens": 121755112.0,
      "step": 101260
    },
    {
      "entropy": 1.8759055025875568,
      "epoch": 0.313928139378266,
      "grad_norm": 9.331445693969727,
      "learning_rate": 4.515228161336056e-06,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8457464009523392,
      "num_tokens": 121767583.0,
      "step": 101270
    },
    {
      "entropy": 1.8825054869055748,
      "epoch": 0.31395913850331575,
      "grad_norm": 3.7135350704193115,
      "learning_rate": 4.515005245449148e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8491608336567879,
      "num_tokens": 121779121.0,
      "step": 101280
    },
    {
      "entropy": 1.8358449339866638,
      "epoch": 0.3139901376283654,
      "grad_norm": 8.02281379699707,
      "learning_rate": 4.514782362574916e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8574068158864975,
      "num_tokens": 121791200.0,
      "step": 101290
    },
    {
      "entropy": 1.7923202894628047,
      "epoch": 0.31402113675341514,
      "grad_norm": 8.330782890319824,
      "learning_rate": 4.514559512705209e-06,
      "loss": 0.4104,
      "mean_token_accuracy": 0.8662643000483513,
      "num_tokens": 121804297.0,
      "step": 101300
    },
    {
      "entropy": 1.9169503509998322,
      "epoch": 0.3140521358784648,
      "grad_norm": 8.749537467956543,
      "learning_rate": 4.514336695831886e-06,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8455056488513947,
      "num_tokens": 121815612.0,
      "step": 101310
    },
    {
      "entropy": 1.7872198984026908,
      "epoch": 0.31408313500351454,
      "grad_norm": 7.533987522125244,
      "learning_rate": 4.514113911946806e-06,
      "loss": 0.4268,
      "mean_token_accuracy": 0.858782111108303,
      "num_tokens": 121828475.0,
      "step": 101320
    },
    {
      "entropy": 1.9148533344268799,
      "epoch": 0.3141141341285642,
      "grad_norm": 8.236104011535645,
      "learning_rate": 4.5138911610418245e-06,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8438421174883842,
      "num_tokens": 121838831.0,
      "step": 101330
    },
    {
      "entropy": 1.9065877795219421,
      "epoch": 0.31414513325361393,
      "grad_norm": 8.377083778381348,
      "learning_rate": 4.51366844310881e-06,
      "loss": 0.4838,
      "mean_token_accuracy": 0.8459710478782654,
      "num_tokens": 121850367.0,
      "step": 101340
    },
    {
      "entropy": 1.9168651878833771,
      "epoch": 0.3141761323786636,
      "grad_norm": 8.766077995300293,
      "learning_rate": 4.513445758139627e-06,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8526653304696084,
      "num_tokens": 121860852.0,
      "step": 101350
    },
    {
      "entropy": 1.8872756123542787,
      "epoch": 0.3142071315037133,
      "grad_norm": 8.51346492767334,
      "learning_rate": 4.513223106126145e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8485476955771446,
      "num_tokens": 121871508.0,
      "step": 101360
    },
    {
      "entropy": 1.8321407943964005,
      "epoch": 0.314238130628763,
      "grad_norm": 7.802825450897217,
      "learning_rate": 4.513000487060237e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8585144847631454,
      "num_tokens": 121883191.0,
      "step": 101370
    },
    {
      "entropy": 1.8902756407856942,
      "epoch": 0.3142691297538127,
      "grad_norm": 8.383158683776855,
      "learning_rate": 4.5127779009337785e-06,
      "loss": 0.5165,
      "mean_token_accuracy": 0.840720933675766,
      "num_tokens": 121894503.0,
      "step": 101380
    },
    {
      "entropy": 1.9538305789232253,
      "epoch": 0.3143001288788624,
      "grad_norm": 10.174178123474121,
      "learning_rate": 4.5125553477386455e-06,
      "loss": 0.5445,
      "mean_token_accuracy": 0.8378095537424087,
      "num_tokens": 121905568.0,
      "step": 101390
    },
    {
      "entropy": 1.8209902182221414,
      "epoch": 0.3143311280039121,
      "grad_norm": 9.949512481689453,
      "learning_rate": 4.512332827466718e-06,
      "loss": 0.466,
      "mean_token_accuracy": 0.8366820439696312,
      "num_tokens": 121918020.0,
      "step": 101400
    },
    {
      "entropy": 1.8949487626552581,
      "epoch": 0.3143621271289618,
      "grad_norm": 7.5901360511779785,
      "learning_rate": 4.5121103401098816e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8482216715812683,
      "num_tokens": 121929522.0,
      "step": 101410
    },
    {
      "entropy": 1.805781890451908,
      "epoch": 0.3143931262540115,
      "grad_norm": 7.198166370391846,
      "learning_rate": 4.5118878856600216e-06,
      "loss": 0.4295,
      "mean_token_accuracy": 0.8508959487080574,
      "num_tokens": 121941861.0,
      "step": 101420
    },
    {
      "entropy": 1.9467833280563354,
      "epoch": 0.31442412537906117,
      "grad_norm": 7.574606895446777,
      "learning_rate": 4.511665464109026e-06,
      "loss": 0.531,
      "mean_token_accuracy": 0.8413808315992355,
      "num_tokens": 121952629.0,
      "step": 101430
    },
    {
      "entropy": 1.8594104155898095,
      "epoch": 0.3144551245041109,
      "grad_norm": 8.646973609924316,
      "learning_rate": 4.511443075448789e-06,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8487472698092461,
      "num_tokens": 121964076.0,
      "step": 101440
    },
    {
      "entropy": 1.8867372497916222,
      "epoch": 0.31448612362916056,
      "grad_norm": 9.636038780212402,
      "learning_rate": 4.511220719671201e-06,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8549345463514328,
      "num_tokens": 121975596.0,
      "step": 101450
    },
    {
      "entropy": 1.8463930860161781,
      "epoch": 0.3145171227542103,
      "grad_norm": 4.379640579223633,
      "learning_rate": 4.510998396768163e-06,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8542065605521202,
      "num_tokens": 121988727.0,
      "step": 101460
    },
    {
      "entropy": 1.7745927199721336,
      "epoch": 0.31454812187925996,
      "grad_norm": 8.495243072509766,
      "learning_rate": 4.510776106731574e-06,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8536753922700882,
      "num_tokens": 122001880.0,
      "step": 101470
    },
    {
      "entropy": 1.857958073914051,
      "epoch": 0.3145791210043097,
      "grad_norm": 8.16500473022461,
      "learning_rate": 4.510553849553338e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8449077561497689,
      "num_tokens": 122013484.0,
      "step": 101480
    },
    {
      "entropy": 1.894733229279518,
      "epoch": 0.31461012012935935,
      "grad_norm": 8.241775512695312,
      "learning_rate": 4.5103316252253596e-06,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8418209850788116,
      "num_tokens": 122024930.0,
      "step": 101490
    },
    {
      "entropy": 1.8933887377381324,
      "epoch": 0.314641119254409,
      "grad_norm": 6.502079963684082,
      "learning_rate": 4.510109433739546e-06,
      "loss": 0.5135,
      "mean_token_accuracy": 0.8463633015751839,
      "num_tokens": 122036290.0,
      "step": 101500
    },
    {
      "entropy": 1.7902062192559243,
      "epoch": 0.31467211837945874,
      "grad_norm": 7.463756084442139,
      "learning_rate": 4.5098872750878105e-06,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8511275082826615,
      "num_tokens": 122048789.0,
      "step": 101510
    },
    {
      "entropy": 1.8736846506595612,
      "epoch": 0.3147031175045084,
      "grad_norm": 8.760028839111328,
      "learning_rate": 4.509665149262067e-06,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8344025030732155,
      "num_tokens": 122060583.0,
      "step": 101520
    },
    {
      "entropy": 1.863266195356846,
      "epoch": 0.31473411662955814,
      "grad_norm": 8.545906066894531,
      "learning_rate": 4.509443056254233e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8548782065510749,
      "num_tokens": 122072828.0,
      "step": 101530
    },
    {
      "entropy": 1.8797732189297676,
      "epoch": 0.3147651157546078,
      "grad_norm": 7.313852787017822,
      "learning_rate": 4.509220996056225e-06,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8434521064162255,
      "num_tokens": 122084579.0,
      "step": 101540
    },
    {
      "entropy": 1.823886838555336,
      "epoch": 0.31479611487965753,
      "grad_norm": 5.079513072967529,
      "learning_rate": 4.508998968659968e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8580957293510437,
      "num_tokens": 122097466.0,
      "step": 101550
    },
    {
      "entropy": 1.8132152885198594,
      "epoch": 0.3148271140047072,
      "grad_norm": 10.782577514648438,
      "learning_rate": 4.508776974057388e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.838719479739666,
      "num_tokens": 122109870.0,
      "step": 101560
    },
    {
      "entropy": 1.8953714028000832,
      "epoch": 0.3148581131297569,
      "grad_norm": 8.248565673828125,
      "learning_rate": 4.508555012240411e-06,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8457905784249306,
      "num_tokens": 122121740.0,
      "step": 101570
    },
    {
      "entropy": 1.8286928310990334,
      "epoch": 0.3148891122548066,
      "grad_norm": 9.213640213012695,
      "learning_rate": 4.50833308320097e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8573044881224632,
      "num_tokens": 122133522.0,
      "step": 101580
    },
    {
      "entropy": 1.8590153113007546,
      "epoch": 0.3149201113798563,
      "grad_norm": 5.629488945007324,
      "learning_rate": 4.508111186930996e-06,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8378587946295738,
      "num_tokens": 122145368.0,
      "step": 101590
    },
    {
      "entropy": 1.8506905019283295,
      "epoch": 0.314951110504906,
      "grad_norm": 7.348735332489014,
      "learning_rate": 4.507889323422427e-06,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8469028800725937,
      "num_tokens": 122157779.0,
      "step": 101600
    },
    {
      "entropy": 1.8291833832859994,
      "epoch": 0.3149821096299557,
      "grad_norm": 2.6711909770965576,
      "learning_rate": 4.507667492667202e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8578299075365067,
      "num_tokens": 122170157.0,
      "step": 101610
    },
    {
      "entropy": 1.9278810530900956,
      "epoch": 0.3150131087550054,
      "grad_norm": 6.753495693206787,
      "learning_rate": 4.507445694657263e-06,
      "loss": 0.5123,
      "mean_token_accuracy": 0.8417115181684494,
      "num_tokens": 122181568.0,
      "step": 101620
    },
    {
      "entropy": 1.8500063866376877,
      "epoch": 0.3150441078800551,
      "grad_norm": 9.120920181274414,
      "learning_rate": 4.507223929384555e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.8450711652636528,
      "num_tokens": 122193790.0,
      "step": 101630
    },
    {
      "entropy": 1.8878964528441429,
      "epoch": 0.31507510700510477,
      "grad_norm": 7.948580265045166,
      "learning_rate": 4.507002196841023e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8558190748095512,
      "num_tokens": 122205835.0,
      "step": 101640
    },
    {
      "entropy": 1.870565366744995,
      "epoch": 0.3151061061301545,
      "grad_norm": 9.229273796081543,
      "learning_rate": 4.506780497018622e-06,
      "loss": 0.4816,
      "mean_token_accuracy": 0.8505363553762436,
      "num_tokens": 122217561.0,
      "step": 101650
    },
    {
      "entropy": 1.8509984895586968,
      "epoch": 0.31513710525520416,
      "grad_norm": 9.244194984436035,
      "learning_rate": 4.506558829909301e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8476543664932251,
      "num_tokens": 122230248.0,
      "step": 101660
    },
    {
      "entropy": 1.7851860001683235,
      "epoch": 0.3151681043802539,
      "grad_norm": 8.21959114074707,
      "learning_rate": 4.506337195505015e-06,
      "loss": 0.3845,
      "mean_token_accuracy": 0.8607925057411194,
      "num_tokens": 122243301.0,
      "step": 101670
    },
    {
      "entropy": 1.843540082871914,
      "epoch": 0.31519910350530356,
      "grad_norm": 8.222967147827148,
      "learning_rate": 4.506115593797727e-06,
      "loss": 0.4182,
      "mean_token_accuracy": 0.8591127768158913,
      "num_tokens": 122255771.0,
      "step": 101680
    },
    {
      "entropy": 1.8111511290073394,
      "epoch": 0.3152301026303533,
      "grad_norm": 4.0410075187683105,
      "learning_rate": 4.5058940247793955e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8595302075147628,
      "num_tokens": 122268473.0,
      "step": 101690
    },
    {
      "entropy": 1.849461354315281,
      "epoch": 0.31526110175540295,
      "grad_norm": 7.890182971954346,
      "learning_rate": 4.505672488441985e-06,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8466949865221978,
      "num_tokens": 122280554.0,
      "step": 101700
    },
    {
      "entropy": 1.8682439640164374,
      "epoch": 0.3152921008804527,
      "grad_norm": 3.920172691345215,
      "learning_rate": 4.505450984777461e-06,
      "loss": 0.4249,
      "mean_token_accuracy": 0.8509997203946114,
      "num_tokens": 122293129.0,
      "step": 101710
    },
    {
      "entropy": 1.8806020855903625,
      "epoch": 0.31532310000550234,
      "grad_norm": 9.121551513671875,
      "learning_rate": 4.505229513777795e-06,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8400007635354996,
      "num_tokens": 122304732.0,
      "step": 101720
    },
    {
      "entropy": 1.902934204041958,
      "epoch": 0.315354099130552,
      "grad_norm": 8.14372730255127,
      "learning_rate": 4.50500807543496e-06,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8440614491701126,
      "num_tokens": 122315872.0,
      "step": 101730
    },
    {
      "entropy": 1.8729471534490585,
      "epoch": 0.31538509825560174,
      "grad_norm": 7.441555023193359,
      "learning_rate": 4.50478666974093e-06,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8485934257507324,
      "num_tokens": 122327035.0,
      "step": 101740
    },
    {
      "entropy": 1.9011955708265305,
      "epoch": 0.3154160973806514,
      "grad_norm": 8.446456909179688,
      "learning_rate": 4.50456529668768e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8409095421433449,
      "num_tokens": 122338445.0,
      "step": 101750
    },
    {
      "entropy": 1.7948980048298835,
      "epoch": 0.31544709650570113,
      "grad_norm": 10.517394065856934,
      "learning_rate": 4.5043439562671966e-06,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8519067704677582,
      "num_tokens": 122351396.0,
      "step": 101760
    },
    {
      "entropy": 1.7647177547216415,
      "epoch": 0.3154780956307508,
      "grad_norm": 7.822119235992432,
      "learning_rate": 4.504122648471458e-06,
      "loss": 0.3804,
      "mean_token_accuracy": 0.8645659565925599,
      "num_tokens": 122364593.0,
      "step": 101770
    },
    {
      "entropy": 1.912806712090969,
      "epoch": 0.3155090947558005,
      "grad_norm": 8.647019386291504,
      "learning_rate": 4.503901373292454e-06,
      "loss": 0.5724,
      "mean_token_accuracy": 0.8309219375252723,
      "num_tokens": 122376209.0,
      "step": 101780
    },
    {
      "entropy": 1.8549558535218238,
      "epoch": 0.3155400938808502,
      "grad_norm": 7.488748073577881,
      "learning_rate": 4.503680130722171e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8498001918196678,
      "num_tokens": 122388129.0,
      "step": 101790
    },
    {
      "entropy": 1.9467090874910356,
      "epoch": 0.3155710930058999,
      "grad_norm": 7.977104663848877,
      "learning_rate": 4.5034589207526026e-06,
      "loss": 0.5692,
      "mean_token_accuracy": 0.8243438869714736,
      "num_tokens": 122399489.0,
      "step": 101800
    },
    {
      "entropy": 1.8753120481967926,
      "epoch": 0.3156020921309496,
      "grad_norm": 7.6365275382995605,
      "learning_rate": 4.503237743375743e-06,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8401636779308319,
      "num_tokens": 122411610.0,
      "step": 101810
    },
    {
      "entropy": 1.8790741473436356,
      "epoch": 0.3156330912559993,
      "grad_norm": 9.488393783569336,
      "learning_rate": 4.503016598583588e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8573433116078377,
      "num_tokens": 122423284.0,
      "step": 101820
    },
    {
      "entropy": 1.9299686640501021,
      "epoch": 0.315664090381049,
      "grad_norm": 7.945889472961426,
      "learning_rate": 4.502795486368138e-06,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8483487293124199,
      "num_tokens": 122434547.0,
      "step": 101830
    },
    {
      "entropy": 1.778617675602436,
      "epoch": 0.3156950895060987,
      "grad_norm": 9.670980453491211,
      "learning_rate": 4.502574406721396e-06,
      "loss": 0.4062,
      "mean_token_accuracy": 0.8523766458034515,
      "num_tokens": 122447628.0,
      "step": 101840
    },
    {
      "entropy": 1.7951662212610244,
      "epoch": 0.31572608863114837,
      "grad_norm": 8.881525039672852,
      "learning_rate": 4.502353359635368e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.861037427186966,
      "num_tokens": 122460464.0,
      "step": 101850
    },
    {
      "entropy": 1.8687596887350082,
      "epoch": 0.3157570877561981,
      "grad_norm": 7.735236644744873,
      "learning_rate": 4.502132345102062e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.853579594194889,
      "num_tokens": 122471873.0,
      "step": 101860
    },
    {
      "entropy": 1.8538660779595375,
      "epoch": 0.31578808688124776,
      "grad_norm": 9.091965675354004,
      "learning_rate": 4.501911363113488e-06,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8424102887511253,
      "num_tokens": 122483758.0,
      "step": 101870
    },
    {
      "entropy": 1.867423267662525,
      "epoch": 0.3158190860062975,
      "grad_norm": 7.476759433746338,
      "learning_rate": 4.50169041366166e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.855827559530735,
      "num_tokens": 122495774.0,
      "step": 101880
    },
    {
      "entropy": 1.9684102550148963,
      "epoch": 0.31585008513134716,
      "grad_norm": 9.030182838439941,
      "learning_rate": 4.501469496738595e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8357870906591416,
      "num_tokens": 122507503.0,
      "step": 101890
    },
    {
      "entropy": 1.8453612983226777,
      "epoch": 0.3158810842563969,
      "grad_norm": 3.91621732711792,
      "learning_rate": 4.501248612336311e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8493709444999695,
      "num_tokens": 122519916.0,
      "step": 101900
    },
    {
      "entropy": 1.8780709460377694,
      "epoch": 0.31591208338144655,
      "grad_norm": 4.055104732513428,
      "learning_rate": 4.501027760446832e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8418286308646202,
      "num_tokens": 122532278.0,
      "step": 101910
    },
    {
      "entropy": 1.8270500496029853,
      "epoch": 0.3159430825064963,
      "grad_norm": 8.200374603271484,
      "learning_rate": 4.500806941062181e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8556545317173004,
      "num_tokens": 122544680.0,
      "step": 101920
    },
    {
      "entropy": 1.8527694910764694,
      "epoch": 0.31597408163154594,
      "grad_norm": 7.4447526931762695,
      "learning_rate": 4.500586154174386e-06,
      "loss": 0.4066,
      "mean_token_accuracy": 0.8587725609540939,
      "num_tokens": 122557057.0,
      "step": 101930
    },
    {
      "entropy": 1.8287393301725388,
      "epoch": 0.31600508075659567,
      "grad_norm": 8.269747734069824,
      "learning_rate": 4.500365399775477e-06,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8537361025810242,
      "num_tokens": 122570020.0,
      "step": 101940
    },
    {
      "entropy": 1.9338515534996987,
      "epoch": 0.31603607988164534,
      "grad_norm": 4.0301194190979,
      "learning_rate": 4.500144677857487e-06,
      "loss": 0.508,
      "mean_token_accuracy": 0.8461434602737427,
      "num_tokens": 122580842.0,
      "step": 101950
    },
    {
      "entropy": 1.8989423006772994,
      "epoch": 0.31606707900669506,
      "grad_norm": 9.705557823181152,
      "learning_rate": 4.499923988412451e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.844089737534523,
      "num_tokens": 122592807.0,
      "step": 101960
    },
    {
      "entropy": 1.7944003835320472,
      "epoch": 0.31609807813174473,
      "grad_norm": 8.264009475708008,
      "learning_rate": 4.4997033314324076e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.853903503715992,
      "num_tokens": 122606158.0,
      "step": 101970
    },
    {
      "entropy": 1.9085728481411934,
      "epoch": 0.3161290772567944,
      "grad_norm": 4.879970073699951,
      "learning_rate": 4.499482706909398e-06,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8358641535043716,
      "num_tokens": 122618189.0,
      "step": 101980
    },
    {
      "entropy": 1.9110256865620614,
      "epoch": 0.3161600763818441,
      "grad_norm": 11.392416954040527,
      "learning_rate": 4.4992621148354666e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8411575809121132,
      "num_tokens": 122630458.0,
      "step": 101990
    },
    {
      "entropy": 1.8680866047739983,
      "epoch": 0.3161910755068938,
      "grad_norm": 8.535940170288086,
      "learning_rate": 4.499041555202658e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8537841528654099,
      "num_tokens": 122642497.0,
      "step": 102000
    },
    {
      "entropy": 1.9318107053637505,
      "epoch": 0.3162220746319435,
      "grad_norm": 7.174762725830078,
      "learning_rate": 4.498821028003023e-06,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8535374835133552,
      "num_tokens": 122654391.0,
      "step": 102010
    },
    {
      "entropy": 1.8076833948493003,
      "epoch": 0.3162530737569932,
      "grad_norm": 8.442280769348145,
      "learning_rate": 4.498600533228614e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.8697671309113503,
      "num_tokens": 122667692.0,
      "step": 102020
    },
    {
      "entropy": 1.9444027364253997,
      "epoch": 0.3162840728820429,
      "grad_norm": 8.13665771484375,
      "learning_rate": 4.498380070871485e-06,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8343800812959671,
      "num_tokens": 122679172.0,
      "step": 102030
    },
    {
      "entropy": 1.8926381140947341,
      "epoch": 0.3163150720070926,
      "grad_norm": 4.520792484283447,
      "learning_rate": 4.498159640923693e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8546594932675362,
      "num_tokens": 122691105.0,
      "step": 102040
    },
    {
      "entropy": 1.9276057749986648,
      "epoch": 0.3163460711321423,
      "grad_norm": 9.296843528747559,
      "learning_rate": 4.497939243377298e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8524391040205955,
      "num_tokens": 122702895.0,
      "step": 102050
    },
    {
      "entropy": 1.9363893195986748,
      "epoch": 0.316377070257192,
      "grad_norm": 6.961995601654053,
      "learning_rate": 4.497718878224365e-06,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8385184094309807,
      "num_tokens": 122714385.0,
      "step": 102060
    },
    {
      "entropy": 1.9360941782593728,
      "epoch": 0.3164080693822417,
      "grad_norm": 8.984434127807617,
      "learning_rate": 4.497498545456957e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8492272660136223,
      "num_tokens": 122725974.0,
      "step": 102070
    },
    {
      "entropy": 1.88591488301754,
      "epoch": 0.31643906850729137,
      "grad_norm": 9.188456535339355,
      "learning_rate": 4.497278245067143e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8486263528466225,
      "num_tokens": 122737041.0,
      "step": 102080
    },
    {
      "entropy": 1.9170858517289162,
      "epoch": 0.3164700676323411,
      "grad_norm": 8.99303913116455,
      "learning_rate": 4.497057977046996e-06,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8467420935630798,
      "num_tokens": 122748309.0,
      "step": 102090
    },
    {
      "entropy": 1.950069211423397,
      "epoch": 0.31650106675739076,
      "grad_norm": 8.753754615783691,
      "learning_rate": 4.4968377413885885e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.849167463183403,
      "num_tokens": 122759466.0,
      "step": 102100
    },
    {
      "entropy": 1.9460075750947,
      "epoch": 0.3165320658824405,
      "grad_norm": 8.198152542114258,
      "learning_rate": 4.496617538083995e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.847400327026844,
      "num_tokens": 122770569.0,
      "step": 102110
    },
    {
      "entropy": 1.9162710309028625,
      "epoch": 0.31656306500749015,
      "grad_norm": 8.986804008483887,
      "learning_rate": 4.496397367125297e-06,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8529955089092255,
      "num_tokens": 122782469.0,
      "step": 102120
    },
    {
      "entropy": 1.9209685817360878,
      "epoch": 0.3165940641325399,
      "grad_norm": 9.081794738769531,
      "learning_rate": 4.496177228504574e-06,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8467201054096222,
      "num_tokens": 122794654.0,
      "step": 102130
    },
    {
      "entropy": 1.976951664686203,
      "epoch": 0.31662506325758955,
      "grad_norm": 7.217496395111084,
      "learning_rate": 4.495957122213915e-06,
      "loss": 0.5074,
      "mean_token_accuracy": 0.852759413421154,
      "num_tokens": 122805399.0,
      "step": 102140
    },
    {
      "entropy": 1.7019195973873138,
      "epoch": 0.31665606238263927,
      "grad_norm": 8.643301963806152,
      "learning_rate": 4.495737048245404e-06,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8538576439023018,
      "num_tokens": 122819420.0,
      "step": 102150
    },
    {
      "entropy": 1.8965346455574035,
      "epoch": 0.31668706150768894,
      "grad_norm": 8.723462104797363,
      "learning_rate": 4.495517006591132e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8469094544649124,
      "num_tokens": 122831159.0,
      "step": 102160
    },
    {
      "entropy": 1.8087441340088843,
      "epoch": 0.31671806063273866,
      "grad_norm": 7.827014923095703,
      "learning_rate": 4.495296997243191e-06,
      "loss": 0.4458,
      "mean_token_accuracy": 0.8492914751172066,
      "num_tokens": 122844637.0,
      "step": 102170
    },
    {
      "entropy": 1.8186364084482194,
      "epoch": 0.31674905975778833,
      "grad_norm": 8.309804916381836,
      "learning_rate": 4.495077020193676e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8481403559446334,
      "num_tokens": 122857719.0,
      "step": 102180
    },
    {
      "entropy": 1.924719001352787,
      "epoch": 0.31678005888283806,
      "grad_norm": 8.645857810974121,
      "learning_rate": 4.494857075434688e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8571926310658455,
      "num_tokens": 122869193.0,
      "step": 102190
    },
    {
      "entropy": 1.9453157484531403,
      "epoch": 0.3168110580078877,
      "grad_norm": 8.812455177307129,
      "learning_rate": 4.494637162958325e-06,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8361015647649765,
      "num_tokens": 122879724.0,
      "step": 102200
    },
    {
      "entropy": 1.8693825080990791,
      "epoch": 0.31684205713293745,
      "grad_norm": 10.073441505432129,
      "learning_rate": 4.494417282756691e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8441578835248947,
      "num_tokens": 122892200.0,
      "step": 102210
    },
    {
      "entropy": 1.8365435376763344,
      "epoch": 0.3168730562579871,
      "grad_norm": 3.941197633743286,
      "learning_rate": 4.494197434821895e-06,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8510231733322143,
      "num_tokens": 122904637.0,
      "step": 102220
    },
    {
      "entropy": 1.8284949362277985,
      "epoch": 0.3169040553830368,
      "grad_norm": 7.839810371398926,
      "learning_rate": 4.493977619146042e-06,
      "loss": 0.4285,
      "mean_token_accuracy": 0.8462954074144363,
      "num_tokens": 122917337.0,
      "step": 102230
    },
    {
      "entropy": 1.8114649429917336,
      "epoch": 0.3169350545080865,
      "grad_norm": 7.352813243865967,
      "learning_rate": 4.493757835721245e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8442055761814118,
      "num_tokens": 122930610.0,
      "step": 102240
    },
    {
      "entropy": 1.9509634003043175,
      "epoch": 0.3169660536331362,
      "grad_norm": 8.08341121673584,
      "learning_rate": 4.49353808453962e-06,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8404506236314774,
      "num_tokens": 122941825.0,
      "step": 102250
    },
    {
      "entropy": 1.9052265360951424,
      "epoch": 0.3169970527581859,
      "grad_norm": 8.447335243225098,
      "learning_rate": 4.4933183655932825e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8546632289886474,
      "num_tokens": 122953086.0,
      "step": 102260
    },
    {
      "entropy": 1.8697161242365836,
      "epoch": 0.3170280518832356,
      "grad_norm": 7.899079322814941,
      "learning_rate": 4.493098678874353e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8598993286490441,
      "num_tokens": 122965302.0,
      "step": 102270
    },
    {
      "entropy": 1.9763395830988884,
      "epoch": 0.3170590510082853,
      "grad_norm": 7.561324119567871,
      "learning_rate": 4.4928790243749535e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8493476256728172,
      "num_tokens": 122976430.0,
      "step": 102280
    },
    {
      "entropy": 1.9137044921517372,
      "epoch": 0.31709005013333497,
      "grad_norm": 8.271001815795898,
      "learning_rate": 4.49265940208721e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8426155328750611,
      "num_tokens": 122987244.0,
      "step": 102290
    },
    {
      "entropy": 1.8825251743197442,
      "epoch": 0.3171210492583847,
      "grad_norm": 9.43757152557373,
      "learning_rate": 4.4924398120032505e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.851975978910923,
      "num_tokens": 122998662.0,
      "step": 102300
    },
    {
      "entropy": 1.8874466240406036,
      "epoch": 0.31715204838343436,
      "grad_norm": 8.041158676147461,
      "learning_rate": 4.492220254115204e-06,
      "loss": 0.496,
      "mean_token_accuracy": 0.8433241844177246,
      "num_tokens": 123010762.0,
      "step": 102310
    },
    {
      "entropy": 1.9816171437501908,
      "epoch": 0.3171830475084841,
      "grad_norm": 9.10623836517334,
      "learning_rate": 4.492000728415204e-06,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8387098371982574,
      "num_tokens": 123021337.0,
      "step": 102320
    },
    {
      "entropy": 1.9616368293762207,
      "epoch": 0.31721404663353375,
      "grad_norm": 9.265618324279785,
      "learning_rate": 4.491781234895389e-06,
      "loss": 0.5216,
      "mean_token_accuracy": 0.8416030526161193,
      "num_tokens": 123032883.0,
      "step": 102330
    },
    {
      "entropy": 1.8963989228010179,
      "epoch": 0.3172450457585835,
      "grad_norm": 8.671114921569824,
      "learning_rate": 4.4915617735478936e-06,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8574667409062385,
      "num_tokens": 123044915.0,
      "step": 102340
    },
    {
      "entropy": 1.8731760695576667,
      "epoch": 0.31727604488363315,
      "grad_norm": 7.5588297843933105,
      "learning_rate": 4.491342344364861e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8465674906969071,
      "num_tokens": 123056529.0,
      "step": 102350
    },
    {
      "entropy": 1.9712413370609283,
      "epoch": 0.31730704400868287,
      "grad_norm": 8.025903701782227,
      "learning_rate": 4.491122947338437e-06,
      "loss": 0.5524,
      "mean_token_accuracy": 0.8367073446512222,
      "num_tokens": 123067329.0,
      "step": 102360
    },
    {
      "entropy": 1.849116511642933,
      "epoch": 0.31733804313373254,
      "grad_norm": 7.625123023986816,
      "learning_rate": 4.490903582460766e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.847317686676979,
      "num_tokens": 123080562.0,
      "step": 102370
    },
    {
      "entropy": 1.980668443441391,
      "epoch": 0.31736904225878226,
      "grad_norm": 7.4852705001831055,
      "learning_rate": 4.490684249724e-06,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8438632428646088,
      "num_tokens": 123092615.0,
      "step": 102380
    },
    {
      "entropy": 1.9390667513012887,
      "epoch": 0.31740004138383193,
      "grad_norm": 8.869697570800781,
      "learning_rate": 4.4904649491202866e-06,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8526061177253723,
      "num_tokens": 123104152.0,
      "step": 102390
    },
    {
      "entropy": 1.866631529480219,
      "epoch": 0.31743104050888166,
      "grad_norm": 7.319535255432129,
      "learning_rate": 4.490245680641784e-06,
      "loss": 0.4087,
      "mean_token_accuracy": 0.8547472521662712,
      "num_tokens": 123116432.0,
      "step": 102400
    },
    {
      "entropy": 1.9240225657820702,
      "epoch": 0.3174620396339313,
      "grad_norm": 7.42906379699707,
      "learning_rate": 4.490026444280649e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8577197745442391,
      "num_tokens": 123128281.0,
      "step": 102410
    },
    {
      "entropy": 1.8641801089048387,
      "epoch": 0.31749303875898105,
      "grad_norm": 8.197351455688477,
      "learning_rate": 4.48980724002904e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8553032830357552,
      "num_tokens": 123140851.0,
      "step": 102420
    },
    {
      "entropy": 1.8485140323638916,
      "epoch": 0.3175240378840307,
      "grad_norm": 11.487910270690918,
      "learning_rate": 4.489588067879123e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.8603411689400673,
      "num_tokens": 123153326.0,
      "step": 102430
    },
    {
      "entropy": 1.93117448836565,
      "epoch": 0.31755503700908044,
      "grad_norm": 9.506771087646484,
      "learning_rate": 4.489368927823061e-06,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8498334422707557,
      "num_tokens": 123165165.0,
      "step": 102440
    },
    {
      "entropy": 1.958940689265728,
      "epoch": 0.3175860361341301,
      "grad_norm": 7.967811107635498,
      "learning_rate": 4.489149819853024e-06,
      "loss": 0.4962,
      "mean_token_accuracy": 0.847984354197979,
      "num_tokens": 123176618.0,
      "step": 102450
    },
    {
      "entropy": 1.9103962182998657,
      "epoch": 0.31761703525917984,
      "grad_norm": 8.108328819274902,
      "learning_rate": 4.4889307439611805e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8437720760703087,
      "num_tokens": 123188100.0,
      "step": 102460
    },
    {
      "entropy": 1.916041937470436,
      "epoch": 0.3176480343842295,
      "grad_norm": 8.520224571228027,
      "learning_rate": 4.488711700139705e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8483468234539032,
      "num_tokens": 123199271.0,
      "step": 102470
    },
    {
      "entropy": 1.8879058375954627,
      "epoch": 0.3176790335092792,
      "grad_norm": 6.797135829925537,
      "learning_rate": 4.488492688380775e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8475539967417717,
      "num_tokens": 123211068.0,
      "step": 102480
    },
    {
      "entropy": 1.9228038251399995,
      "epoch": 0.3177100326343289,
      "grad_norm": 8.574908256530762,
      "learning_rate": 4.488273708676567e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8467731967568397,
      "num_tokens": 123222860.0,
      "step": 102490
    },
    {
      "entropy": 1.9041634023189544,
      "epoch": 0.31774103175937857,
      "grad_norm": 3.588517427444458,
      "learning_rate": 4.488054761019265e-06,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8531323969364166,
      "num_tokens": 123234902.0,
      "step": 102500
    },
    {
      "entropy": 1.880505882203579,
      "epoch": 0.3177720308844283,
      "grad_norm": 7.573834419250488,
      "learning_rate": 4.487835845401051e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8521980285644531,
      "num_tokens": 123247483.0,
      "step": 102510
    },
    {
      "entropy": 1.9097674801945685,
      "epoch": 0.31780303000947796,
      "grad_norm": 8.0872163772583,
      "learning_rate": 4.487616961814113e-06,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8391719311475754,
      "num_tokens": 123259632.0,
      "step": 102520
    },
    {
      "entropy": 1.9394026383757592,
      "epoch": 0.3178340291345277,
      "grad_norm": 7.864894390106201,
      "learning_rate": 4.48739811025064e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8454336494207382,
      "num_tokens": 123270948.0,
      "step": 102530
    },
    {
      "entropy": 1.8619687780737877,
      "epoch": 0.31786502825957735,
      "grad_norm": 3.913823127746582,
      "learning_rate": 4.487179290702825e-06,
      "loss": 0.465,
      "mean_token_accuracy": 0.8360814332962037,
      "num_tokens": 123283604.0,
      "step": 102540
    },
    {
      "entropy": 1.934596875309944,
      "epoch": 0.3178960273846271,
      "grad_norm": 8.981165885925293,
      "learning_rate": 4.486960503162861e-06,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8490733042359352,
      "num_tokens": 123294899.0,
      "step": 102550
    },
    {
      "entropy": 1.9189150497317313,
      "epoch": 0.31792702650967675,
      "grad_norm": 7.386570453643799,
      "learning_rate": 4.4867417476229475e-06,
      "loss": 0.496,
      "mean_token_accuracy": 0.8401806384325028,
      "num_tokens": 123306606.0,
      "step": 102560
    },
    {
      "entropy": 1.746788166463375,
      "epoch": 0.31795802563472647,
      "grad_norm": 8.746540069580078,
      "learning_rate": 4.486523024075284e-06,
      "loss": 0.3844,
      "mean_token_accuracy": 0.8678254216909409,
      "num_tokens": 123319627.0,
      "step": 102570
    },
    {
      "entropy": 1.8892858445644378,
      "epoch": 0.31798902475977614,
      "grad_norm": 9.26871109008789,
      "learning_rate": 4.486304332512073e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8421612724661827,
      "num_tokens": 123331981.0,
      "step": 102580
    },
    {
      "entropy": 1.847158958017826,
      "epoch": 0.31802002388482586,
      "grad_norm": 8.670111656188965,
      "learning_rate": 4.48608567292552e-06,
      "loss": 0.4146,
      "mean_token_accuracy": 0.8566896244883537,
      "num_tokens": 123344136.0,
      "step": 102590
    },
    {
      "entropy": 1.839782963693142,
      "epoch": 0.31805102300987553,
      "grad_norm": 3.6876060962677,
      "learning_rate": 4.485867045307833e-06,
      "loss": 0.435,
      "mean_token_accuracy": 0.8545563265681266,
      "num_tokens": 123356337.0,
      "step": 102600
    },
    {
      "entropy": 1.8948510199785233,
      "epoch": 0.31808202213492526,
      "grad_norm": 8.7213773727417,
      "learning_rate": 4.485648449651225e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8595595166087151,
      "num_tokens": 123367945.0,
      "step": 102610
    },
    {
      "entropy": 1.8524553552269936,
      "epoch": 0.3181130212599749,
      "grad_norm": 8.925561904907227,
      "learning_rate": 4.485429885947906e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8614932551980019,
      "num_tokens": 123380029.0,
      "step": 102620
    },
    {
      "entropy": 1.8003907322883606,
      "epoch": 0.31814402038502465,
      "grad_norm": 7.866354465484619,
      "learning_rate": 4.485211354190095e-06,
      "loss": 0.3988,
      "mean_token_accuracy": 0.8583444371819496,
      "num_tokens": 123392498.0,
      "step": 102630
    },
    {
      "entropy": 1.8503638431429863,
      "epoch": 0.3181750195100743,
      "grad_norm": 9.890091896057129,
      "learning_rate": 4.4849928543700085e-06,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8414298683404923,
      "num_tokens": 123405679.0,
      "step": 102640
    },
    {
      "entropy": 1.8876867666840553,
      "epoch": 0.31820601863512404,
      "grad_norm": 4.612701416015625,
      "learning_rate": 4.4847743864798694e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8404382139444351,
      "num_tokens": 123417818.0,
      "step": 102650
    },
    {
      "entropy": 1.8589577794075012,
      "epoch": 0.3182370177601737,
      "grad_norm": 8.45789909362793,
      "learning_rate": 4.4845559505119026e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8587053641676903,
      "num_tokens": 123430054.0,
      "step": 102660
    },
    {
      "entropy": 1.8961587071418762,
      "epoch": 0.31826801688522344,
      "grad_norm": 7.61850118637085,
      "learning_rate": 4.484337546458332e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8480044439435005,
      "num_tokens": 123442303.0,
      "step": 102670
    },
    {
      "entropy": 1.8941489323973655,
      "epoch": 0.3182990160102731,
      "grad_norm": 7.859434604644775,
      "learning_rate": 4.484119174311389e-06,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8440102905035018,
      "num_tokens": 123454240.0,
      "step": 102680
    },
    {
      "entropy": 1.8841777712106704,
      "epoch": 0.31833001513532283,
      "grad_norm": 3.9713032245635986,
      "learning_rate": 4.483900834063305e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.8437553867697716,
      "num_tokens": 123465668.0,
      "step": 102690
    },
    {
      "entropy": 1.8325778350234032,
      "epoch": 0.3183610142603725,
      "grad_norm": 7.278379917144775,
      "learning_rate": 4.483682525706316e-06,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8592413857579231,
      "num_tokens": 123478018.0,
      "step": 102700
    },
    {
      "entropy": 1.868157622218132,
      "epoch": 0.3183920133854222,
      "grad_norm": 4.052938461303711,
      "learning_rate": 4.483464249232657e-06,
      "loss": 0.4497,
      "mean_token_accuracy": 0.8553399801254272,
      "num_tokens": 123490636.0,
      "step": 102710
    },
    {
      "entropy": 1.8868106931447983,
      "epoch": 0.3184230125104719,
      "grad_norm": 7.89746618270874,
      "learning_rate": 4.483246004634569e-06,
      "loss": 0.4555,
      "mean_token_accuracy": 0.8540831059217453,
      "num_tokens": 123502078.0,
      "step": 102720
    },
    {
      "entropy": 1.9166793465614318,
      "epoch": 0.31845401163552156,
      "grad_norm": 8.033954620361328,
      "learning_rate": 4.4830277919042956e-06,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8464692577719688,
      "num_tokens": 123513541.0,
      "step": 102730
    },
    {
      "entropy": 1.8216803684830665,
      "epoch": 0.3184850107605713,
      "grad_norm": 7.996546745300293,
      "learning_rate": 4.482809611034082e-06,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8364669859409333,
      "num_tokens": 123526355.0,
      "step": 102740
    },
    {
      "entropy": 1.8570113122463225,
      "epoch": 0.31851600988562095,
      "grad_norm": 7.581510066986084,
      "learning_rate": 4.482591462016174e-06,
      "loss": 0.429,
      "mean_token_accuracy": 0.8555145755410194,
      "num_tokens": 123538211.0,
      "step": 102750
    },
    {
      "entropy": 1.812189969420433,
      "epoch": 0.3185470090106707,
      "grad_norm": 9.858771324157715,
      "learning_rate": 4.482373344842824e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8614937171339989,
      "num_tokens": 123550270.0,
      "step": 102760
    },
    {
      "entropy": 1.8564748376607896,
      "epoch": 0.31857800813572035,
      "grad_norm": 6.510000228881836,
      "learning_rate": 4.482155259506284e-06,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8492413744330406,
      "num_tokens": 123561940.0,
      "step": 102770
    },
    {
      "entropy": 1.9163195461034774,
      "epoch": 0.31860900726077007,
      "grad_norm": 7.715924263000488,
      "learning_rate": 4.4819372059988115e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8402003988623619,
      "num_tokens": 123573352.0,
      "step": 102780
    },
    {
      "entropy": 1.851955994963646,
      "epoch": 0.31864000638581974,
      "grad_norm": 8.36665153503418,
      "learning_rate": 4.4817191843126635e-06,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8535103261470794,
      "num_tokens": 123584588.0,
      "step": 102790
    },
    {
      "entropy": 1.9031839981675147,
      "epoch": 0.31867100551086947,
      "grad_norm": 8.757781982421875,
      "learning_rate": 4.4815011944401015e-06,
      "loss": 0.469,
      "mean_token_accuracy": 0.8555735006928444,
      "num_tokens": 123596035.0,
      "step": 102800
    },
    {
      "entropy": 1.8560205325484276,
      "epoch": 0.31870200463591913,
      "grad_norm": 8.463615417480469,
      "learning_rate": 4.4812832363733894e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8499815389513969,
      "num_tokens": 123607275.0,
      "step": 102810
    },
    {
      "entropy": 1.848600834608078,
      "epoch": 0.31873300376096886,
      "grad_norm": 7.636923313140869,
      "learning_rate": 4.481065310104793e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8392280101776123,
      "num_tokens": 123619498.0,
      "step": 102820
    },
    {
      "entropy": 1.9192082822322845,
      "epoch": 0.3187640028860185,
      "grad_norm": 7.930289268493652,
      "learning_rate": 4.480847415626582e-06,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8376297026872634,
      "num_tokens": 123631251.0,
      "step": 102830
    },
    {
      "entropy": 1.897643305361271,
      "epoch": 0.31879500201106825,
      "grad_norm": 8.173277854919434,
      "learning_rate": 4.480629552931028e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8563371822237968,
      "num_tokens": 123642062.0,
      "step": 102840
    },
    {
      "entropy": 1.88473329693079,
      "epoch": 0.3188260011361179,
      "grad_norm": 3.9585182666778564,
      "learning_rate": 4.480411722010404e-06,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8535011231899261,
      "num_tokens": 123652956.0,
      "step": 102850
    },
    {
      "entropy": 1.8646925508975982,
      "epoch": 0.31885700026116764,
      "grad_norm": 9.49219799041748,
      "learning_rate": 4.4801939228569895e-06,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8506576329469681,
      "num_tokens": 123665406.0,
      "step": 102860
    },
    {
      "entropy": 1.8895646423101424,
      "epoch": 0.3188879993862173,
      "grad_norm": 8.293597221374512,
      "learning_rate": 4.4799761554630605e-06,
      "loss": 0.5429,
      "mean_token_accuracy": 0.8343157604336738,
      "num_tokens": 123677366.0,
      "step": 102870
    },
    {
      "entropy": 1.8557443410158156,
      "epoch": 0.31891899851126704,
      "grad_norm": 3.470597505569458,
      "learning_rate": 4.479758419820902e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8372866466641427,
      "num_tokens": 123689278.0,
      "step": 102880
    },
    {
      "entropy": 1.8852709233760834,
      "epoch": 0.3189499976363167,
      "grad_norm": 8.924466133117676,
      "learning_rate": 4.479540715922798e-06,
      "loss": 0.488,
      "mean_token_accuracy": 0.8414081260561943,
      "num_tokens": 123701095.0,
      "step": 102890
    },
    {
      "entropy": 1.8459098264575005,
      "epoch": 0.31898099676136643,
      "grad_norm": 9.664191246032715,
      "learning_rate": 4.479323043761035e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8502274408936501,
      "num_tokens": 123713345.0,
      "step": 102900
    },
    {
      "entropy": 1.8116124227643013,
      "epoch": 0.3190119958864161,
      "grad_norm": 8.264068603515625,
      "learning_rate": 4.479105403327904e-06,
      "loss": 0.4638,
      "mean_token_accuracy": 0.8466699972748757,
      "num_tokens": 123725722.0,
      "step": 102910
    },
    {
      "entropy": 1.8975732818245887,
      "epoch": 0.3190429950114658,
      "grad_norm": 8.010908126831055,
      "learning_rate": 4.478887794615696e-06,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8428432583808899,
      "num_tokens": 123736798.0,
      "step": 102920
    },
    {
      "entropy": 1.836000031232834,
      "epoch": 0.3190739941365155,
      "grad_norm": 10.32380485534668,
      "learning_rate": 4.4786702176167084e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8488903164863586,
      "num_tokens": 123748662.0,
      "step": 102930
    },
    {
      "entropy": 1.8443504258990289,
      "epoch": 0.3191049932615652,
      "grad_norm": 7.753139019012451,
      "learning_rate": 4.478452672323238e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8580980643630027,
      "num_tokens": 123761662.0,
      "step": 102940
    },
    {
      "entropy": 1.8338020756840705,
      "epoch": 0.3191359923866149,
      "grad_norm": 9.386250495910645,
      "learning_rate": 4.4782351587275865e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.844118581712246,
      "num_tokens": 123773805.0,
      "step": 102950
    },
    {
      "entropy": 1.8451603963971137,
      "epoch": 0.3191669915116646,
      "grad_norm": 8.358642578125,
      "learning_rate": 4.478017676822054e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8548496574163437,
      "num_tokens": 123786092.0,
      "step": 102960
    },
    {
      "entropy": 1.9037387266755104,
      "epoch": 0.3191979906367143,
      "grad_norm": 10.784769058227539,
      "learning_rate": 4.47780022659895e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8458883672952652,
      "num_tokens": 123798134.0,
      "step": 102970
    },
    {
      "entropy": 1.9118751406669616,
      "epoch": 0.31922898976176395,
      "grad_norm": 8.400568962097168,
      "learning_rate": 4.47758280805058e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8501716613769531,
      "num_tokens": 123809256.0,
      "step": 102980
    },
    {
      "entropy": 1.9268585234880446,
      "epoch": 0.3192599888868137,
      "grad_norm": 9.43397045135498,
      "learning_rate": 4.477365421169256e-06,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8423597529530525,
      "num_tokens": 123819919.0,
      "step": 102990
    },
    {
      "entropy": 1.7849035665392876,
      "epoch": 0.31929098801186334,
      "grad_norm": 8.134185791015625,
      "learning_rate": 4.477148065947293e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8558384835720062,
      "num_tokens": 123832624.0,
      "step": 103000
    },
    {
      "entropy": 1.8854929059743881,
      "epoch": 0.31932198713691307,
      "grad_norm": 7.981983661651611,
      "learning_rate": 4.476930742377004e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8419500634074211,
      "num_tokens": 123843833.0,
      "step": 103010
    },
    {
      "entropy": 1.910216248035431,
      "epoch": 0.31935298626196273,
      "grad_norm": 8.01833438873291,
      "learning_rate": 4.47671345045071e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8421685129404068,
      "num_tokens": 123854736.0,
      "step": 103020
    },
    {
      "entropy": 1.869119329750538,
      "epoch": 0.31938398538701246,
      "grad_norm": 9.197733879089355,
      "learning_rate": 4.4764961901607315e-06,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8363765180110931,
      "num_tokens": 123866473.0,
      "step": 103030
    },
    {
      "entropy": 1.9347961366176605,
      "epoch": 0.31941498451206213,
      "grad_norm": 7.009976863861084,
      "learning_rate": 4.476278961499394e-06,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8408634766936303,
      "num_tokens": 123877914.0,
      "step": 103040
    },
    {
      "entropy": 1.765951743721962,
      "epoch": 0.31944598363711185,
      "grad_norm": 6.8173933029174805,
      "learning_rate": 4.4760617644590216e-06,
      "loss": 0.3763,
      "mean_token_accuracy": 0.8628764286637306,
      "num_tokens": 123891728.0,
      "step": 103050
    },
    {
      "entropy": 1.8665249332785607,
      "epoch": 0.3194769827621615,
      "grad_norm": 7.756616592407227,
      "learning_rate": 4.475844599031945e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8461050614714622,
      "num_tokens": 123904519.0,
      "step": 103060
    },
    {
      "entropy": 1.9601498633623122,
      "epoch": 0.31950798188721125,
      "grad_norm": 9.403843879699707,
      "learning_rate": 4.475627465210497e-06,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8456299170851708,
      "num_tokens": 123915934.0,
      "step": 103070
    },
    {
      "entropy": 1.872973631322384,
      "epoch": 0.3195389810122609,
      "grad_norm": 8.8914155960083,
      "learning_rate": 4.475410362987011e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.837957626581192,
      "num_tokens": 123928155.0,
      "step": 103080
    },
    {
      "entropy": 1.909055621922016,
      "epoch": 0.31956998013731064,
      "grad_norm": 7.7289509773254395,
      "learning_rate": 4.475193292353822e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.848746582865715,
      "num_tokens": 123939530.0,
      "step": 103090
    },
    {
      "entropy": 1.9184820994734764,
      "epoch": 0.3196009792623603,
      "grad_norm": 8.20902156829834,
      "learning_rate": 4.474976253303274e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.843792799115181,
      "num_tokens": 123950860.0,
      "step": 103100
    },
    {
      "entropy": 1.8916692659258842,
      "epoch": 0.31963197838741003,
      "grad_norm": 11.049240112304688,
      "learning_rate": 4.4747592458277056e-06,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8373655050992965,
      "num_tokens": 123962960.0,
      "step": 103110
    },
    {
      "entropy": 1.9099874019622802,
      "epoch": 0.3196629775124597,
      "grad_norm": 7.249629497528076,
      "learning_rate": 4.474542269919464e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8369929790496826,
      "num_tokens": 123974112.0,
      "step": 103120
    },
    {
      "entropy": 1.8539279848337173,
      "epoch": 0.3196939766375094,
      "grad_norm": 8.467607498168945,
      "learning_rate": 4.474325325570893e-06,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8553314670920372,
      "num_tokens": 123986201.0,
      "step": 103130
    },
    {
      "entropy": 1.9297155156731605,
      "epoch": 0.3197249757625591,
      "grad_norm": 11.687642097473145,
      "learning_rate": 4.474108412774347e-06,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8387605383992195,
      "num_tokens": 123997260.0,
      "step": 103140
    },
    {
      "entropy": 1.8051437750458716,
      "epoch": 0.3197559748876088,
      "grad_norm": 7.733287334442139,
      "learning_rate": 4.473891531522177e-06,
      "loss": 0.3998,
      "mean_token_accuracy": 0.8608478337526322,
      "num_tokens": 124009391.0,
      "step": 103150
    },
    {
      "entropy": 1.7467816695570946,
      "epoch": 0.3197869740126585,
      "grad_norm": 2.378082036972046,
      "learning_rate": 4.473674681806737e-06,
      "loss": 0.3921,
      "mean_token_accuracy": 0.869289082288742,
      "num_tokens": 124022647.0,
      "step": 103160
    },
    {
      "entropy": 1.876607745885849,
      "epoch": 0.3198179731377082,
      "grad_norm": 9.72205924987793,
      "learning_rate": 4.473457863620386e-06,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8447935312986374,
      "num_tokens": 124034377.0,
      "step": 103170
    },
    {
      "entropy": 1.8972087427973747,
      "epoch": 0.3198489722627579,
      "grad_norm": 7.99284553527832,
      "learning_rate": 4.473241076955484e-06,
      "loss": 0.4895,
      "mean_token_accuracy": 0.847452299296856,
      "num_tokens": 124045591.0,
      "step": 103180
    },
    {
      "entropy": 1.8817680388689042,
      "epoch": 0.3198799713878076,
      "grad_norm": 8.3921480178833,
      "learning_rate": 4.473024321804395e-06,
      "loss": 0.4908,
      "mean_token_accuracy": 0.8453219920396805,
      "num_tokens": 124056874.0,
      "step": 103190
    },
    {
      "entropy": 1.845687435567379,
      "epoch": 0.3199109705128573,
      "grad_norm": 8.912002563476562,
      "learning_rate": 4.472807598159483e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.845789223909378,
      "num_tokens": 124068631.0,
      "step": 103200
    },
    {
      "entropy": 1.7905006155371666,
      "epoch": 0.319941969637907,
      "grad_norm": 9.591071128845215,
      "learning_rate": 4.472590906013117e-06,
      "loss": 0.4584,
      "mean_token_accuracy": 0.8493612468242645,
      "num_tokens": 124081286.0,
      "step": 103210
    },
    {
      "entropy": 1.9089469194412232,
      "epoch": 0.31997296876295667,
      "grad_norm": 7.169071197509766,
      "learning_rate": 4.4723742453576675e-06,
      "loss": 0.502,
      "mean_token_accuracy": 0.8382734417915344,
      "num_tokens": 124092376.0,
      "step": 103220
    },
    {
      "entropy": 1.8337425097823143,
      "epoch": 0.32000396788800634,
      "grad_norm": 8.232246398925781,
      "learning_rate": 4.472157616185508e-06,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8578509896993637,
      "num_tokens": 124104861.0,
      "step": 103230
    },
    {
      "entropy": 1.9013035476207734,
      "epoch": 0.32003496701305606,
      "grad_norm": 8.12004280090332,
      "learning_rate": 4.471941018489015e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8492400407791137,
      "num_tokens": 124116168.0,
      "step": 103240
    },
    {
      "entropy": 1.8799728155136108,
      "epoch": 0.32006596613810573,
      "grad_norm": 7.341602325439453,
      "learning_rate": 4.471724452260566e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8415581986308098,
      "num_tokens": 124127690.0,
      "step": 103250
    },
    {
      "entropy": 1.8120905488729477,
      "epoch": 0.32009696526315545,
      "grad_norm": 9.098899841308594,
      "learning_rate": 4.471507917492542e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8570401355624199,
      "num_tokens": 124140339.0,
      "step": 103260
    },
    {
      "entropy": 1.8739472836256028,
      "epoch": 0.3201279643882051,
      "grad_norm": 9.830232620239258,
      "learning_rate": 4.471291414177328e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8453883707523346,
      "num_tokens": 124151956.0,
      "step": 103270
    },
    {
      "entropy": 1.84071436971426,
      "epoch": 0.32015896351325485,
      "grad_norm": 8.632100105285645,
      "learning_rate": 4.47107494230731e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8345205932855606,
      "num_tokens": 124165412.0,
      "step": 103280
    },
    {
      "entropy": 1.8590612232685089,
      "epoch": 0.3201899626383045,
      "grad_norm": 4.4268975257873535,
      "learning_rate": 4.470858501874875e-06,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8525667667388916,
      "num_tokens": 124177344.0,
      "step": 103290
    },
    {
      "entropy": 1.887212759256363,
      "epoch": 0.32022096176335424,
      "grad_norm": 7.400831699371338,
      "learning_rate": 4.470642092872416e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8470708400011062,
      "num_tokens": 124188119.0,
      "step": 103300
    },
    {
      "entropy": 1.8698201969265937,
      "epoch": 0.3202519608884039,
      "grad_norm": 9.03722858428955,
      "learning_rate": 4.470425715292328e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8450605183839798,
      "num_tokens": 124199654.0,
      "step": 103310
    },
    {
      "entropy": 1.9616728246212005,
      "epoch": 0.32028296001345363,
      "grad_norm": 7.2699503898620605,
      "learning_rate": 4.4702093691270045e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8387251660227776,
      "num_tokens": 124210832.0,
      "step": 103320
    },
    {
      "entropy": 1.888064543902874,
      "epoch": 0.3203139591385033,
      "grad_norm": 9.468146324157715,
      "learning_rate": 4.469993054368849e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8365824237465859,
      "num_tokens": 124223111.0,
      "step": 103330
    },
    {
      "entropy": 1.8957604005932809,
      "epoch": 0.320344958263553,
      "grad_norm": 4.3390302658081055,
      "learning_rate": 4.469776771010258e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8410006389021873,
      "num_tokens": 124234226.0,
      "step": 103340
    },
    {
      "entropy": 1.9063668623566628,
      "epoch": 0.3203759573886027,
      "grad_norm": 8.515616416931152,
      "learning_rate": 4.46956051904364e-06,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8457826554775238,
      "num_tokens": 124245737.0,
      "step": 103350
    },
    {
      "entropy": 1.8583616137504577,
      "epoch": 0.3204069565136524,
      "grad_norm": 8.301240921020508,
      "learning_rate": 4.469344298461399e-06,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8528946593403817,
      "num_tokens": 124258363.0,
      "step": 103360
    },
    {
      "entropy": 1.8560489758849144,
      "epoch": 0.3204379556387021,
      "grad_norm": 7.1716532707214355,
      "learning_rate": 4.4691281092559474e-06,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8615340426564216,
      "num_tokens": 124270052.0,
      "step": 103370
    },
    {
      "entropy": 1.762592874467373,
      "epoch": 0.3204689547637518,
      "grad_norm": 3.3581652641296387,
      "learning_rate": 4.468911951419696e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8641716584563255,
      "num_tokens": 124283377.0,
      "step": 103380
    },
    {
      "entropy": 1.8988282978534698,
      "epoch": 0.3204999538888015,
      "grad_norm": 6.756303310394287,
      "learning_rate": 4.468695824945058e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8476831942796708,
      "num_tokens": 124294752.0,
      "step": 103390
    },
    {
      "entropy": 1.846705262362957,
      "epoch": 0.3205309530138512,
      "grad_norm": 3.833975315093994,
      "learning_rate": 4.46847972982445e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8555555865168571,
      "num_tokens": 124306566.0,
      "step": 103400
    },
    {
      "entropy": 1.8886190831661225,
      "epoch": 0.3205619521389009,
      "grad_norm": 7.964195728302002,
      "learning_rate": 4.4682636660502945e-06,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8455360010266304,
      "num_tokens": 124318282.0,
      "step": 103410
    },
    {
      "entropy": 1.9076228067278862,
      "epoch": 0.3205929512639506,
      "grad_norm": 7.338057994842529,
      "learning_rate": 4.468047633615013e-06,
      "loss": 0.4968,
      "mean_token_accuracy": 0.8368429213762283,
      "num_tokens": 124329917.0,
      "step": 103420
    },
    {
      "entropy": 1.9275338113307954,
      "epoch": 0.32062395038900027,
      "grad_norm": 10.642918586730957,
      "learning_rate": 4.46783163251103e-06,
      "loss": 0.5179,
      "mean_token_accuracy": 0.8408781468868256,
      "num_tokens": 124341468.0,
      "step": 103430
    },
    {
      "entropy": 1.967161425948143,
      "epoch": 0.32065494951405,
      "grad_norm": 7.011558532714844,
      "learning_rate": 4.467615662730772e-06,
      "loss": 0.535,
      "mean_token_accuracy": 0.8357431918382645,
      "num_tokens": 124352683.0,
      "step": 103440
    },
    {
      "entropy": 1.8656065806746482,
      "epoch": 0.32068594863909966,
      "grad_norm": 7.173338413238525,
      "learning_rate": 4.467399724266671e-06,
      "loss": 0.4432,
      "mean_token_accuracy": 0.8562691912055016,
      "num_tokens": 124364944.0,
      "step": 103450
    },
    {
      "entropy": 1.7879148676991463,
      "epoch": 0.32071694776414933,
      "grad_norm": 8.152206420898438,
      "learning_rate": 4.467183817111157e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8556606069207191,
      "num_tokens": 124378575.0,
      "step": 103460
    },
    {
      "entropy": 1.9068648159503936,
      "epoch": 0.32074794688919905,
      "grad_norm": 10.881712913513184,
      "learning_rate": 4.466967941256668e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8376988887786865,
      "num_tokens": 124390206.0,
      "step": 103470
    },
    {
      "entropy": 1.9137423366308213,
      "epoch": 0.3207789460142487,
      "grad_norm": 7.468229293823242,
      "learning_rate": 4.4667520966956385e-06,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8356907978653908,
      "num_tokens": 124401455.0,
      "step": 103480
    },
    {
      "entropy": 1.9200310036540031,
      "epoch": 0.32080994513929845,
      "grad_norm": 6.459151744842529,
      "learning_rate": 4.466536283420511e-06,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8393821954727173,
      "num_tokens": 124412634.0,
      "step": 103490
    },
    {
      "entropy": 1.9078100383281709,
      "epoch": 0.3208409442643481,
      "grad_norm": 6.7956013679504395,
      "learning_rate": 4.466320501423726e-06,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8611381649971008,
      "num_tokens": 124423788.0,
      "step": 103500
    },
    {
      "entropy": 1.9619006097316742,
      "epoch": 0.32087194338939784,
      "grad_norm": 9.467850685119629,
      "learning_rate": 4.466104750697733e-06,
      "loss": 0.5486,
      "mean_token_accuracy": 0.8296520456671714,
      "num_tokens": 124434439.0,
      "step": 103510
    },
    {
      "entropy": 1.8933661192655564,
      "epoch": 0.3209029425144475,
      "grad_norm": 4.713155746459961,
      "learning_rate": 4.465889031234975e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8469473794102669,
      "num_tokens": 124446667.0,
      "step": 103520
    },
    {
      "entropy": 1.853841246664524,
      "epoch": 0.32093394163949723,
      "grad_norm": 8.347504615783691,
      "learning_rate": 4.465673343027906e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.844866256415844,
      "num_tokens": 124459684.0,
      "step": 103530
    },
    {
      "entropy": 1.8124510049819946,
      "epoch": 0.3209649407645469,
      "grad_norm": 4.092881202697754,
      "learning_rate": 4.465457686068977e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8626367405056954,
      "num_tokens": 124472022.0,
      "step": 103540
    },
    {
      "entropy": 1.8816181853413583,
      "epoch": 0.3209959398895966,
      "grad_norm": 7.426637172698975,
      "learning_rate": 4.465242060350643e-06,
      "loss": 0.4722,
      "mean_token_accuracy": 0.844549834728241,
      "num_tokens": 124484006.0,
      "step": 103550
    },
    {
      "entropy": 1.8122664958238601,
      "epoch": 0.3210269390146463,
      "grad_norm": 3.879347562789917,
      "learning_rate": 4.4650264658653655e-06,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8467624381184577,
      "num_tokens": 124496792.0,
      "step": 103560
    },
    {
      "entropy": 1.8967302456498145,
      "epoch": 0.321057938139696,
      "grad_norm": 5.540412902832031,
      "learning_rate": 4.464810902605601e-06,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8287664383649826,
      "num_tokens": 124508254.0,
      "step": 103570
    },
    {
      "entropy": 1.8487597212195397,
      "epoch": 0.3210889372647457,
      "grad_norm": 4.171758651733398,
      "learning_rate": 4.464595370563815e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8530851736664772,
      "num_tokens": 124520515.0,
      "step": 103580
    },
    {
      "entropy": 1.8957301035523415,
      "epoch": 0.3211199363897954,
      "grad_norm": 8.205263137817383,
      "learning_rate": 4.464379869732473e-06,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8478555142879486,
      "num_tokens": 124532156.0,
      "step": 103590
    },
    {
      "entropy": 1.8724404126405716,
      "epoch": 0.3211509355148451,
      "grad_norm": 8.49742317199707,
      "learning_rate": 4.464164400104043e-06,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8556983798742295,
      "num_tokens": 124543672.0,
      "step": 103600
    },
    {
      "entropy": 1.8402127534151078,
      "epoch": 0.3211819346398948,
      "grad_norm": 8.611581802368164,
      "learning_rate": 4.4639489616709956e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8543668672442436,
      "num_tokens": 124555679.0,
      "step": 103610
    },
    {
      "entropy": 1.8802747756242753,
      "epoch": 0.3212129337649445,
      "grad_norm": 4.206470012664795,
      "learning_rate": 4.463733554425804e-06,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8467269361019134,
      "num_tokens": 124567696.0,
      "step": 103620
    },
    {
      "entropy": 1.864666721224785,
      "epoch": 0.3212439328899942,
      "grad_norm": 10.374842643737793,
      "learning_rate": 4.4635181783609455e-06,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8502185359597206,
      "num_tokens": 124580191.0,
      "step": 103630
    },
    {
      "entropy": 1.839669795334339,
      "epoch": 0.32127493201504387,
      "grad_norm": 3.81731915473938,
      "learning_rate": 4.463302833468897e-06,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8490176141262055,
      "num_tokens": 124592216.0,
      "step": 103640
    },
    {
      "entropy": 1.8732964277267456,
      "epoch": 0.3213059311400936,
      "grad_norm": 7.773388385772705,
      "learning_rate": 4.463087519742139e-06,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8358982741832733,
      "num_tokens": 124604399.0,
      "step": 103650
    },
    {
      "entropy": 1.8805906429886818,
      "epoch": 0.32133693026514326,
      "grad_norm": 8.990577697753906,
      "learning_rate": 4.462872237173157e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8515235692262649,
      "num_tokens": 124615677.0,
      "step": 103660
    },
    {
      "entropy": 1.8140804409980773,
      "epoch": 0.321367929390193,
      "grad_norm": 8.409754753112793,
      "learning_rate": 4.462656985754436e-06,
      "loss": 0.4065,
      "mean_token_accuracy": 0.8694860532879829,
      "num_tokens": 124628163.0,
      "step": 103670
    },
    {
      "entropy": 1.8692081153392792,
      "epoch": 0.32139892851524265,
      "grad_norm": 8.463071823120117,
      "learning_rate": 4.462441765478465e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8528854250907898,
      "num_tokens": 124640372.0,
      "step": 103680
    },
    {
      "entropy": 1.8829544261097908,
      "epoch": 0.3214299276402924,
      "grad_norm": 9.008349418640137,
      "learning_rate": 4.462226576337735e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8522943645715714,
      "num_tokens": 124652104.0,
      "step": 103690
    },
    {
      "entropy": 1.908381125330925,
      "epoch": 0.32146092676534205,
      "grad_norm": 8.958301544189453,
      "learning_rate": 4.462011418324738e-06,
      "loss": 0.4776,
      "mean_token_accuracy": 0.8482077926397323,
      "num_tokens": 124663567.0,
      "step": 103700
    },
    {
      "entropy": 1.8642259582877159,
      "epoch": 0.3214919258903917,
      "grad_norm": 8.490578651428223,
      "learning_rate": 4.461796291431973e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8547911241650581,
      "num_tokens": 124675305.0,
      "step": 103710
    },
    {
      "entropy": 1.8607799611985683,
      "epoch": 0.32152292501544144,
      "grad_norm": 7.851955413818359,
      "learning_rate": 4.461581195651937e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8484767660498619,
      "num_tokens": 124688192.0,
      "step": 103720
    },
    {
      "entropy": 1.930846455693245,
      "epoch": 0.3215539241404911,
      "grad_norm": 3.855354070663452,
      "learning_rate": 4.461366130977132e-06,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8470954403281212,
      "num_tokens": 124699818.0,
      "step": 103730
    },
    {
      "entropy": 1.7706571273505687,
      "epoch": 0.32158492326554083,
      "grad_norm": 4.847051620483398,
      "learning_rate": 4.461151097400059e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.8597544595599175,
      "num_tokens": 124714446.0,
      "step": 103740
    },
    {
      "entropy": 1.9413534983992577,
      "epoch": 0.3216159223905905,
      "grad_norm": 7.653467655181885,
      "learning_rate": 4.460936094913229e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8460107401013375,
      "num_tokens": 124725683.0,
      "step": 103750
    },
    {
      "entropy": 1.899087278544903,
      "epoch": 0.3216469215156402,
      "grad_norm": 9.189424514770508,
      "learning_rate": 4.460721123509149e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8444273769855499,
      "num_tokens": 124738312.0,
      "step": 103760
    },
    {
      "entropy": 1.951054508984089,
      "epoch": 0.3216779206406899,
      "grad_norm": 8.07810115814209,
      "learning_rate": 4.460506183180329e-06,
      "loss": 0.4815,
      "mean_token_accuracy": 0.841863676905632,
      "num_tokens": 124749524.0,
      "step": 103770
    },
    {
      "entropy": 1.9111177667975425,
      "epoch": 0.3217089197657396,
      "grad_norm": 8.68226146697998,
      "learning_rate": 4.4602912739192835e-06,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8471680104732513,
      "num_tokens": 124760976.0,
      "step": 103780
    },
    {
      "entropy": 1.8448022976517677,
      "epoch": 0.3217399188907893,
      "grad_norm": 7.906839847564697,
      "learning_rate": 4.4600763957185295e-06,
      "loss": 0.4239,
      "mean_token_accuracy": 0.8532932892441749,
      "num_tokens": 124773367.0,
      "step": 103790
    },
    {
      "entropy": 1.9335693135857581,
      "epoch": 0.321770918015839,
      "grad_norm": 8.961040496826172,
      "learning_rate": 4.459861548570586e-06,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8489025041460991,
      "num_tokens": 124784721.0,
      "step": 103800
    },
    {
      "entropy": 1.8875031411647796,
      "epoch": 0.3218019171408887,
      "grad_norm": 3.5718820095062256,
      "learning_rate": 4.459646732467974e-06,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8439980506896972,
      "num_tokens": 124797599.0,
      "step": 103810
    },
    {
      "entropy": 1.8897089153528213,
      "epoch": 0.3218329162659384,
      "grad_norm": 9.0376558303833,
      "learning_rate": 4.459431947403218e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8548488318920135,
      "num_tokens": 124808832.0,
      "step": 103820
    },
    {
      "entropy": 1.8961857289075852,
      "epoch": 0.3218639153909881,
      "grad_norm": 7.7893829345703125,
      "learning_rate": 4.4592171933688435e-06,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8496623903512954,
      "num_tokens": 124821106.0,
      "step": 103830
    },
    {
      "entropy": 1.8757227182388305,
      "epoch": 0.3218949145160378,
      "grad_norm": 4.173471927642822,
      "learning_rate": 4.45900247035738e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8542967364192009,
      "num_tokens": 124833050.0,
      "step": 103840
    },
    {
      "entropy": 1.9881081491708756,
      "epoch": 0.32192591364108747,
      "grad_norm": 7.131890296936035,
      "learning_rate": 4.458787778361361e-06,
      "loss": 0.5528,
      "mean_token_accuracy": 0.8397052332758903,
      "num_tokens": 124844238.0,
      "step": 103850
    },
    {
      "entropy": 1.8623345792293549,
      "epoch": 0.3219569127661372,
      "grad_norm": 3.7538177967071533,
      "learning_rate": 4.458573117373317e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8614621505141258,
      "num_tokens": 124856830.0,
      "step": 103860
    },
    {
      "entropy": 1.8851841911673546,
      "epoch": 0.32198791189118686,
      "grad_norm": 4.068458557128906,
      "learning_rate": 4.458358487385787e-06,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8413153782486915,
      "num_tokens": 124869280.0,
      "step": 103870
    },
    {
      "entropy": 1.8985486201941968,
      "epoch": 0.3220189110162366,
      "grad_norm": 4.279972076416016,
      "learning_rate": 4.458143888391309e-06,
      "loss": 0.4229,
      "mean_token_accuracy": 0.8557350024580955,
      "num_tokens": 124880870.0,
      "step": 103880
    },
    {
      "entropy": 1.834354992210865,
      "epoch": 0.32204991014128626,
      "grad_norm": 3.6236605644226074,
      "learning_rate": 4.457929320382427e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8552977159619332,
      "num_tokens": 124893391.0,
      "step": 103890
    },
    {
      "entropy": 1.9463069021701813,
      "epoch": 0.322080909266336,
      "grad_norm": 8.719969749450684,
      "learning_rate": 4.457714783351681e-06,
      "loss": 0.5187,
      "mean_token_accuracy": 0.8463471621274948,
      "num_tokens": 124904359.0,
      "step": 103900
    },
    {
      "entropy": 1.9136415734887122,
      "epoch": 0.32211190839138565,
      "grad_norm": 8.018022537231445,
      "learning_rate": 4.457500277291621e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8480435863137246,
      "num_tokens": 124916133.0,
      "step": 103910
    },
    {
      "entropy": 1.9327465161681174,
      "epoch": 0.3221429075164354,
      "grad_norm": 9.761283874511719,
      "learning_rate": 4.457285802194794e-06,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8388472273945808,
      "num_tokens": 124927380.0,
      "step": 103920
    },
    {
      "entropy": 1.9528162240982057,
      "epoch": 0.32217390664148504,
      "grad_norm": 8.009254455566406,
      "learning_rate": 4.457071358053754e-06,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8379970923066139,
      "num_tokens": 124937778.0,
      "step": 103930
    },
    {
      "entropy": 1.9495197027921676,
      "epoch": 0.32220490576653477,
      "grad_norm": 8.856219291687012,
      "learning_rate": 4.456856944861052e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8489312395453453,
      "num_tokens": 124948572.0,
      "step": 103940
    },
    {
      "entropy": 1.8287577375769615,
      "epoch": 0.32223590489158443,
      "grad_norm": 4.316620349884033,
      "learning_rate": 4.4566425626092495e-06,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8560908138751984,
      "num_tokens": 124961049.0,
      "step": 103950
    },
    {
      "entropy": 1.8360350668430327,
      "epoch": 0.3222669040166341,
      "grad_norm": 8.45445442199707,
      "learning_rate": 4.456428211290899e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.843099394440651,
      "num_tokens": 124973591.0,
      "step": 103960
    },
    {
      "entropy": 1.829276867955923,
      "epoch": 0.32229790314168383,
      "grad_norm": 2.9788012504577637,
      "learning_rate": 4.456213890898567e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8461241707205772,
      "num_tokens": 124986733.0,
      "step": 103970
    },
    {
      "entropy": 1.951300072669983,
      "epoch": 0.3223289022667335,
      "grad_norm": 7.77920389175415,
      "learning_rate": 4.455999601424818e-06,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8368278592824936,
      "num_tokens": 124998066.0,
      "step": 103980
    },
    {
      "entropy": 1.8031010538339616,
      "epoch": 0.3223599013917832,
      "grad_norm": 4.566207408905029,
      "learning_rate": 4.455785342862216e-06,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8428297311067581,
      "num_tokens": 125011581.0,
      "step": 103990
    },
    {
      "entropy": 1.8515203520655632,
      "epoch": 0.3223909005168329,
      "grad_norm": 7.359809875488281,
      "learning_rate": 4.4555711152033325e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.852726761996746,
      "num_tokens": 125024051.0,
      "step": 104000
    },
    {
      "entropy": 1.8627462826669217,
      "epoch": 0.3224218996418826,
      "grad_norm": 3.8170571327209473,
      "learning_rate": 4.455356918440736e-06,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8479026794433594,
      "num_tokens": 125037097.0,
      "step": 104010
    },
    {
      "entropy": 1.8624150231480598,
      "epoch": 0.3224528987669323,
      "grad_norm": 8.229700088500977,
      "learning_rate": 4.455142752567004e-06,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8403912633657455,
      "num_tokens": 125049190.0,
      "step": 104020
    },
    {
      "entropy": 1.8673087805509567,
      "epoch": 0.322483897891982,
      "grad_norm": 8.417292594909668,
      "learning_rate": 4.454928617574712e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8480266436934472,
      "num_tokens": 125061183.0,
      "step": 104030
    },
    {
      "entropy": 1.83387650847435,
      "epoch": 0.3225148970170317,
      "grad_norm": 7.1093220710754395,
      "learning_rate": 4.4547145134564384e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8557756185531616,
      "num_tokens": 125074081.0,
      "step": 104040
    },
    {
      "entropy": 1.8816956907510758,
      "epoch": 0.3225458961420814,
      "grad_norm": 8.284878730773926,
      "learning_rate": 4.454500440204765e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8433906123042106,
      "num_tokens": 125086116.0,
      "step": 104050
    },
    {
      "entropy": 1.9563316345214843,
      "epoch": 0.32257689526713107,
      "grad_norm": 7.569207668304443,
      "learning_rate": 4.454286397812278e-06,
      "loss": 0.5523,
      "mean_token_accuracy": 0.82982589751482,
      "num_tokens": 125097380.0,
      "step": 104060
    },
    {
      "entropy": 1.941270676255226,
      "epoch": 0.3226078943921808,
      "grad_norm": 9.317317008972168,
      "learning_rate": 4.454072386271562e-06,
      "loss": 0.5532,
      "mean_token_accuracy": 0.8411487266421318,
      "num_tokens": 125108379.0,
      "step": 104070
    },
    {
      "entropy": 1.920466449856758,
      "epoch": 0.32263889351723046,
      "grad_norm": 9.6175537109375,
      "learning_rate": 4.453858405575206e-06,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8352476447820664,
      "num_tokens": 125119711.0,
      "step": 104080
    },
    {
      "entropy": 1.8606573291122914,
      "epoch": 0.3226698926422802,
      "grad_norm": 3.547621488571167,
      "learning_rate": 4.453644455715805e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8499340802431107,
      "num_tokens": 125131618.0,
      "step": 104090
    },
    {
      "entropy": 1.9070662692189218,
      "epoch": 0.32270089176732986,
      "grad_norm": 4.856088638305664,
      "learning_rate": 4.453430536685948e-06,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8461712196469307,
      "num_tokens": 125143282.0,
      "step": 104100
    },
    {
      "entropy": 1.8635522559285165,
      "epoch": 0.3227318908923796,
      "grad_norm": 8.272976875305176,
      "learning_rate": 4.453216648478236e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.85389723777771,
      "num_tokens": 125155283.0,
      "step": 104110
    },
    {
      "entropy": 1.8606613449752332,
      "epoch": 0.32276289001742925,
      "grad_norm": 7.335941314697266,
      "learning_rate": 4.453002791085265e-06,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8420170918107033,
      "num_tokens": 125168312.0,
      "step": 104120
    },
    {
      "entropy": 1.869354782998562,
      "epoch": 0.322793889142479,
      "grad_norm": 8.720457077026367,
      "learning_rate": 4.452788964499638e-06,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8450367793440818,
      "num_tokens": 125180831.0,
      "step": 104130
    },
    {
      "entropy": 1.886098426580429,
      "epoch": 0.32282488826752864,
      "grad_norm": 7.61613130569458,
      "learning_rate": 4.452575168713959e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8547211870551109,
      "num_tokens": 125191572.0,
      "step": 104140
    },
    {
      "entropy": 1.8242208793759347,
      "epoch": 0.32285588739257837,
      "grad_norm": 8.236608505249023,
      "learning_rate": 4.452361403720835e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8449318900704383,
      "num_tokens": 125204559.0,
      "step": 104150
    },
    {
      "entropy": 1.9878936111927032,
      "epoch": 0.32288688651762804,
      "grad_norm": 7.960468769073486,
      "learning_rate": 4.452147669512874e-06,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8334811016917228,
      "num_tokens": 125215145.0,
      "step": 104160
    },
    {
      "entropy": 1.8423168882727623,
      "epoch": 0.32291788564267776,
      "grad_norm": 8.506884574890137,
      "learning_rate": 4.451933966082689e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8522772178053856,
      "num_tokens": 125227069.0,
      "step": 104170
    },
    {
      "entropy": 1.9097376301884652,
      "epoch": 0.32294888476772743,
      "grad_norm": 9.220355033874512,
      "learning_rate": 4.451720293422894e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8438367456197738,
      "num_tokens": 125238513.0,
      "step": 104180
    },
    {
      "entropy": 1.8769617334008217,
      "epoch": 0.32297988389277715,
      "grad_norm": 8.432296752929688,
      "learning_rate": 4.451506651526103e-06,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8506760001182556,
      "num_tokens": 125249982.0,
      "step": 104190
    },
    {
      "entropy": 1.8073933839797973,
      "epoch": 0.3230108830178268,
      "grad_norm": 7.3089985847473145,
      "learning_rate": 4.451293040384938e-06,
      "loss": 0.4174,
      "mean_token_accuracy": 0.8688766479492187,
      "num_tokens": 125263337.0,
      "step": 104200
    },
    {
      "entropy": 1.8363840654492378,
      "epoch": 0.3230418821428765,
      "grad_norm": 4.247748851776123,
      "learning_rate": 4.4510794599920185e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.8401136696338654,
      "num_tokens": 125275329.0,
      "step": 104210
    },
    {
      "entropy": 1.8861171647906303,
      "epoch": 0.3230728812679262,
      "grad_norm": 8.15084457397461,
      "learning_rate": 4.45086591033997e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8480441614985466,
      "num_tokens": 125287001.0,
      "step": 104220
    },
    {
      "entropy": 1.862444320321083,
      "epoch": 0.3231038803929759,
      "grad_norm": 9.114710807800293,
      "learning_rate": 4.450652391421417e-06,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8597635760903358,
      "num_tokens": 125299228.0,
      "step": 104230
    },
    {
      "entropy": 1.8157359957695007,
      "epoch": 0.3231348795180256,
      "grad_norm": 10.471318244934082,
      "learning_rate": 4.45043890322899e-06,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8404217720031738,
      "num_tokens": 125312489.0,
      "step": 104240
    },
    {
      "entropy": 1.8435872822999955,
      "epoch": 0.3231658786430753,
      "grad_norm": 6.866489410400391,
      "learning_rate": 4.45022544575532e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8526515334844589,
      "num_tokens": 125325340.0,
      "step": 104250
    },
    {
      "entropy": 1.7846007272601128,
      "epoch": 0.323196877768125,
      "grad_norm": 8.732168197631836,
      "learning_rate": 4.450012018993041e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8618106812238693,
      "num_tokens": 125338525.0,
      "step": 104260
    },
    {
      "entropy": 1.9067389905452727,
      "epoch": 0.32322787689317467,
      "grad_norm": 9.090784072875977,
      "learning_rate": 4.4497986229347886e-06,
      "loss": 0.4901,
      "mean_token_accuracy": 0.8495014622807503,
      "num_tokens": 125349264.0,
      "step": 104270
    },
    {
      "entropy": 1.8359978944063187,
      "epoch": 0.3232588760182244,
      "grad_norm": 8.766422271728516,
      "learning_rate": 4.449585257573202e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.8507504045963288,
      "num_tokens": 125361176.0,
      "step": 104280
    },
    {
      "entropy": 1.860073594748974,
      "epoch": 0.32328987514327406,
      "grad_norm": 8.850909233093262,
      "learning_rate": 4.4493719229009234e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8514194667339325,
      "num_tokens": 125372673.0,
      "step": 104290
    },
    {
      "entropy": 1.8551115795969964,
      "epoch": 0.3233208742683238,
      "grad_norm": 4.093299388885498,
      "learning_rate": 4.449158618910594e-06,
      "loss": 0.4353,
      "mean_token_accuracy": 0.8481899574398994,
      "num_tokens": 125385018.0,
      "step": 104300
    },
    {
      "entropy": 1.9149028778076171,
      "epoch": 0.32335187339337346,
      "grad_norm": 8.859583854675293,
      "learning_rate": 4.448945345594864e-06,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8369590178132057,
      "num_tokens": 125396234.0,
      "step": 104310
    },
    {
      "entropy": 1.9485036730766296,
      "epoch": 0.3233828725184232,
      "grad_norm": 8.875260353088379,
      "learning_rate": 4.448732102946378e-06,
      "loss": 0.5439,
      "mean_token_accuracy": 0.83717480301857,
      "num_tokens": 125407354.0,
      "step": 104320
    },
    {
      "entropy": 1.8549368754029274,
      "epoch": 0.32341387164347285,
      "grad_norm": 3.1946167945861816,
      "learning_rate": 4.448518890957789e-06,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8576024889945983,
      "num_tokens": 125419710.0,
      "step": 104330
    },
    {
      "entropy": 1.8528822794556619,
      "epoch": 0.3234448707685226,
      "grad_norm": 5.784121036529541,
      "learning_rate": 4.44830570962175e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8541821107268334,
      "num_tokens": 125432442.0,
      "step": 104340
    },
    {
      "entropy": 1.9138613402843476,
      "epoch": 0.32347586989357224,
      "grad_norm": 3.181318998336792,
      "learning_rate": 4.448092558930918e-06,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8456951528787613,
      "num_tokens": 125444223.0,
      "step": 104350
    },
    {
      "entropy": 1.893858689069748,
      "epoch": 0.32350686901862197,
      "grad_norm": 7.411118030548096,
      "learning_rate": 4.447879438877952e-06,
      "loss": 0.5144,
      "mean_token_accuracy": 0.8454016864299774,
      "num_tokens": 125455851.0,
      "step": 104360
    },
    {
      "entropy": 1.9555244013667106,
      "epoch": 0.32353786814367164,
      "grad_norm": 8.233355522155762,
      "learning_rate": 4.447666349455512e-06,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8410166382789612,
      "num_tokens": 125466870.0,
      "step": 104370
    },
    {
      "entropy": 1.9233867451548576,
      "epoch": 0.32356886726872136,
      "grad_norm": 8.369714736938477,
      "learning_rate": 4.44745329065626e-06,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8294639691710473,
      "num_tokens": 125477884.0,
      "step": 104380
    },
    {
      "entropy": 1.842393586039543,
      "epoch": 0.32359986639377103,
      "grad_norm": 8.982789993286133,
      "learning_rate": 4.447240262472865e-06,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8471063405275345,
      "num_tokens": 125490187.0,
      "step": 104390
    },
    {
      "entropy": 1.695570257306099,
      "epoch": 0.32363086551882075,
      "grad_norm": 8.46627426147461,
      "learning_rate": 4.447027264897993e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8552982524037361,
      "num_tokens": 125504771.0,
      "step": 104400
    },
    {
      "entropy": 1.790937101840973,
      "epoch": 0.3236618646438704,
      "grad_norm": 7.850381374359131,
      "learning_rate": 4.446814297924315e-06,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8472680017352104,
      "num_tokens": 125517462.0,
      "step": 104410
    },
    {
      "entropy": 1.9535409420728684,
      "epoch": 0.32369286376892015,
      "grad_norm": 8.090740203857422,
      "learning_rate": 4.446601361544507e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8454528123140335,
      "num_tokens": 125528256.0,
      "step": 104420
    },
    {
      "entropy": 1.8886640325188637,
      "epoch": 0.3237238628939698,
      "grad_norm": 4.096344470977783,
      "learning_rate": 4.44638845575124e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8485276773571968,
      "num_tokens": 125540038.0,
      "step": 104430
    },
    {
      "entropy": 1.8927355989813806,
      "epoch": 0.32375486201901954,
      "grad_norm": 8.601818084716797,
      "learning_rate": 4.446175580537197e-06,
      "loss": 0.5351,
      "mean_token_accuracy": 0.8406930550932884,
      "num_tokens": 125551783.0,
      "step": 104440
    },
    {
      "entropy": 1.933559662103653,
      "epoch": 0.3237858611440692,
      "grad_norm": 7.8219475746154785,
      "learning_rate": 4.445962735895055e-06,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8485094651579856,
      "num_tokens": 125562679.0,
      "step": 104450
    },
    {
      "entropy": 1.8612630635499954,
      "epoch": 0.3238168602691189,
      "grad_norm": 8.244562149047852,
      "learning_rate": 4.445749921817498e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8419337302446366,
      "num_tokens": 125574193.0,
      "step": 104460
    },
    {
      "entropy": 1.8549982547760009,
      "epoch": 0.3238478593941686,
      "grad_norm": 9.33348274230957,
      "learning_rate": 4.445537138297214e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8472777247428894,
      "num_tokens": 125585671.0,
      "step": 104470
    },
    {
      "entropy": 1.8255490154027938,
      "epoch": 0.32387885851921827,
      "grad_norm": 7.582674503326416,
      "learning_rate": 4.445324385326889e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8529926791787148,
      "num_tokens": 125599259.0,
      "step": 104480
    },
    {
      "entropy": 1.8867202073335647,
      "epoch": 0.323909857644268,
      "grad_norm": 7.966434478759766,
      "learning_rate": 4.445111662899213e-06,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8491407498717308,
      "num_tokens": 125611435.0,
      "step": 104490
    },
    {
      "entropy": 1.8497378468513488,
      "epoch": 0.32394085676931766,
      "grad_norm": 5.026673793792725,
      "learning_rate": 4.444898971006879e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8529240384697914,
      "num_tokens": 125623829.0,
      "step": 104500
    },
    {
      "entropy": 1.8570104882121086,
      "epoch": 0.3239718558943674,
      "grad_norm": 3.9983022212982178,
      "learning_rate": 4.444686309642584e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8561410129070282,
      "num_tokens": 125636076.0,
      "step": 104510
    },
    {
      "entropy": 1.9286787793040276,
      "epoch": 0.32400285501941706,
      "grad_norm": 9.436726570129395,
      "learning_rate": 4.444473678799025e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8409220159053803,
      "num_tokens": 125647410.0,
      "step": 104520
    },
    {
      "entropy": 1.790224689245224,
      "epoch": 0.3240338541444668,
      "grad_norm": 7.947468280792236,
      "learning_rate": 4.444261078468901e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8576122537255287,
      "num_tokens": 125660817.0,
      "step": 104530
    },
    {
      "entropy": 1.979808408021927,
      "epoch": 0.32406485326951645,
      "grad_norm": 8.681998252868652,
      "learning_rate": 4.444048508644915e-06,
      "loss": 0.5421,
      "mean_token_accuracy": 0.8427085399627685,
      "num_tokens": 125671400.0,
      "step": 104540
    },
    {
      "entropy": 1.8585129737854005,
      "epoch": 0.3240958523945662,
      "grad_norm": 8.156949043273926,
      "learning_rate": 4.443835969319773e-06,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8462090104818344,
      "num_tokens": 125684225.0,
      "step": 104550
    },
    {
      "entropy": 1.8992905005812646,
      "epoch": 0.32412685151961584,
      "grad_norm": 10.150556564331055,
      "learning_rate": 4.443623460486183e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.838335144519806,
      "num_tokens": 125696461.0,
      "step": 104560
    },
    {
      "entropy": 1.8747254326939582,
      "epoch": 0.32415785064466557,
      "grad_norm": 8.261116981506348,
      "learning_rate": 4.443410982136853e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8448391363024712,
      "num_tokens": 125708703.0,
      "step": 104570
    },
    {
      "entropy": 1.812967798113823,
      "epoch": 0.32418884976971524,
      "grad_norm": 9.305374145507812,
      "learning_rate": 4.443198534264497e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.853294064104557,
      "num_tokens": 125721322.0,
      "step": 104580
    },
    {
      "entropy": 1.8749259188771248,
      "epoch": 0.32421984889476496,
      "grad_norm": 8.172529220581055,
      "learning_rate": 4.442986116861831e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.86143379509449,
      "num_tokens": 125733217.0,
      "step": 104590
    },
    {
      "entropy": 1.9058199599385262,
      "epoch": 0.32425084801981463,
      "grad_norm": 4.497680187225342,
      "learning_rate": 4.442773729921569e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.832387951016426,
      "num_tokens": 125744499.0,
      "step": 104600
    },
    {
      "entropy": 1.9057392075657844,
      "epoch": 0.32428184714486435,
      "grad_norm": 9.339095115661621,
      "learning_rate": 4.4425613734364346e-06,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8399216368794441,
      "num_tokens": 125755625.0,
      "step": 104610
    },
    {
      "entropy": 1.887702539563179,
      "epoch": 0.324312846269914,
      "grad_norm": 9.061288833618164,
      "learning_rate": 4.442349047399148e-06,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8430167078971863,
      "num_tokens": 125767881.0,
      "step": 104620
    },
    {
      "entropy": 1.8627898827195168,
      "epoch": 0.32434384539496375,
      "grad_norm": 7.091198921203613,
      "learning_rate": 4.442136751802433e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8567918986082077,
      "num_tokens": 125779283.0,
      "step": 104630
    },
    {
      "entropy": 1.8987063318490982,
      "epoch": 0.3243748445200134,
      "grad_norm": 7.691847324371338,
      "learning_rate": 4.441924486639018e-06,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8585168570280075,
      "num_tokens": 125790750.0,
      "step": 104640
    },
    {
      "entropy": 1.8565407916903496,
      "epoch": 0.32440584364506314,
      "grad_norm": 8.805460929870605,
      "learning_rate": 4.441712251901632e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8448611825704575,
      "num_tokens": 125803070.0,
      "step": 104650
    },
    {
      "entropy": 1.8467104405164718,
      "epoch": 0.3244368427701128,
      "grad_norm": 7.713294982910156,
      "learning_rate": 4.4415000475830064e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.85001719892025,
      "num_tokens": 125814740.0,
      "step": 104660
    },
    {
      "entropy": 1.8211202561855315,
      "epoch": 0.32446784189516253,
      "grad_norm": 4.372882843017578,
      "learning_rate": 4.441287873675877e-06,
      "loss": 0.4047,
      "mean_token_accuracy": 0.8514401495456696,
      "num_tokens": 125826554.0,
      "step": 104670
    },
    {
      "entropy": 1.7906322300434112,
      "epoch": 0.3244988410202122,
      "grad_norm": 7.335409641265869,
      "learning_rate": 4.44107573017298e-06,
      "loss": 0.4032,
      "mean_token_accuracy": 0.8584406465291977,
      "num_tokens": 125839095.0,
      "step": 104680
    },
    {
      "entropy": 1.887892808020115,
      "epoch": 0.3245298401452619,
      "grad_norm": 9.185553550720215,
      "learning_rate": 4.4408636170670526e-06,
      "loss": 0.4883,
      "mean_token_accuracy": 0.839701421558857,
      "num_tokens": 125850893.0,
      "step": 104690
    },
    {
      "entropy": 1.9207829117774964,
      "epoch": 0.3245608392703116,
      "grad_norm": 10.154474258422852,
      "learning_rate": 4.4406515343508405e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8411696791648865,
      "num_tokens": 125862536.0,
      "step": 104700
    },
    {
      "entropy": 1.9450317040085792,
      "epoch": 0.32459183839536127,
      "grad_norm": 10.254415512084961,
      "learning_rate": 4.440439482017084e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8435680419206619,
      "num_tokens": 125873937.0,
      "step": 104710
    },
    {
      "entropy": 1.9040479019284249,
      "epoch": 0.324622837520411,
      "grad_norm": 7.563967704772949,
      "learning_rate": 4.440227460058531e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8446063458919525,
      "num_tokens": 125885785.0,
      "step": 104720
    },
    {
      "entropy": 1.9021693363785743,
      "epoch": 0.32465383664546066,
      "grad_norm": 8.840468406677246,
      "learning_rate": 4.440015468467932e-06,
      "loss": 0.4776,
      "mean_token_accuracy": 0.8490529716014862,
      "num_tokens": 125897435.0,
      "step": 104730
    },
    {
      "entropy": 1.9175179213285447,
      "epoch": 0.3246848357705104,
      "grad_norm": 3.9579575061798096,
      "learning_rate": 4.439803507238037e-06,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8449933081865311,
      "num_tokens": 125908914.0,
      "step": 104740
    },
    {
      "entropy": 1.8500194996595383,
      "epoch": 0.32471583489556005,
      "grad_norm": 8.99378490447998,
      "learning_rate": 4.439591576361599e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8460746437311173,
      "num_tokens": 125921013.0,
      "step": 104750
    },
    {
      "entropy": 1.8048820734024047,
      "epoch": 0.3247468340206098,
      "grad_norm": 2.689271926879883,
      "learning_rate": 4.439379675831374e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8416306912899018,
      "num_tokens": 125934812.0,
      "step": 104760
    },
    {
      "entropy": 1.8943803861737252,
      "epoch": 0.32477783314565944,
      "grad_norm": 9.720478057861328,
      "learning_rate": 4.439167805640121e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.8504800125956535,
      "num_tokens": 125946337.0,
      "step": 104770
    },
    {
      "entropy": 1.8520411089062692,
      "epoch": 0.32480883227070917,
      "grad_norm": 7.526730060577393,
      "learning_rate": 4.438955965780603e-06,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8448957860469818,
      "num_tokens": 125958824.0,
      "step": 104780
    },
    {
      "entropy": 1.9352994859218597,
      "epoch": 0.32483983139575884,
      "grad_norm": 9.028653144836426,
      "learning_rate": 4.438744156245582e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8472319498658181,
      "num_tokens": 125970046.0,
      "step": 104790
    },
    {
      "entropy": 1.8998494163155555,
      "epoch": 0.32487083052080856,
      "grad_norm": 7.81031608581543,
      "learning_rate": 4.438532377027824e-06,
      "loss": 0.5041,
      "mean_token_accuracy": 0.847889693081379,
      "num_tokens": 125981669.0,
      "step": 104800
    },
    {
      "entropy": 1.8678305372595787,
      "epoch": 0.32490182964585823,
      "grad_norm": 6.961506366729736,
      "learning_rate": 4.438320628120095e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8497982889413833,
      "num_tokens": 125994098.0,
      "step": 104810
    },
    {
      "entropy": 1.8759206786751748,
      "epoch": 0.32493282877090796,
      "grad_norm": 7.766717433929443,
      "learning_rate": 4.43810890951517e-06,
      "loss": 0.5073,
      "mean_token_accuracy": 0.8452546551823616,
      "num_tokens": 126006395.0,
      "step": 104820
    },
    {
      "entropy": 1.9110927432775497,
      "epoch": 0.3249638278959576,
      "grad_norm": 9.222915649414062,
      "learning_rate": 4.437897221205818e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8430734008550644,
      "num_tokens": 126017620.0,
      "step": 104830
    },
    {
      "entropy": 1.81925430893898,
      "epoch": 0.32499482702100735,
      "grad_norm": 8.071293830871582,
      "learning_rate": 4.4376855631848185e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8518892988562584,
      "num_tokens": 126030236.0,
      "step": 104840
    },
    {
      "entropy": 1.8442829206585885,
      "epoch": 0.325025826146057,
      "grad_norm": 4.356393337249756,
      "learning_rate": 4.437473935444945e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8437263607978821,
      "num_tokens": 126043043.0,
      "step": 104850
    },
    {
      "entropy": 1.8087534308433533,
      "epoch": 0.32505682527110674,
      "grad_norm": 8.250895500183105,
      "learning_rate": 4.437262337978981e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8567241251468658,
      "num_tokens": 126055544.0,
      "step": 104860
    },
    {
      "entropy": 1.866674281656742,
      "epoch": 0.3250878243961564,
      "grad_norm": 7.741235733032227,
      "learning_rate": 4.437050770779709e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8507555142045021,
      "num_tokens": 126067524.0,
      "step": 104870
    },
    {
      "entropy": 1.9428170025348663,
      "epoch": 0.32511882352120614,
      "grad_norm": 7.691112995147705,
      "learning_rate": 4.436839233839913e-06,
      "loss": 0.5591,
      "mean_token_accuracy": 0.833220262825489,
      "num_tokens": 126078929.0,
      "step": 104880
    },
    {
      "entropy": 1.8663004338741302,
      "epoch": 0.3251498226462558,
      "grad_norm": 6.292846202850342,
      "learning_rate": 4.436627727152381e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8499148935079575,
      "num_tokens": 126091005.0,
      "step": 104890
    },
    {
      "entropy": 1.9085410133004188,
      "epoch": 0.32518082177130553,
      "grad_norm": 8.354232788085938,
      "learning_rate": 4.436416250709903e-06,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8494470819830895,
      "num_tokens": 126102258.0,
      "step": 104900
    },
    {
      "entropy": 1.9257890343666078,
      "epoch": 0.3252118208963552,
      "grad_norm": 8.245911598205566,
      "learning_rate": 4.436204804505272e-06,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8361289039254188,
      "num_tokens": 126112823.0,
      "step": 104910
    },
    {
      "entropy": 1.8960036650300025,
      "epoch": 0.3252428200214049,
      "grad_norm": 7.979340553283691,
      "learning_rate": 4.435993388531282e-06,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8566076219081878,
      "num_tokens": 126123821.0,
      "step": 104920
    },
    {
      "entropy": 1.7642639800906181,
      "epoch": 0.3252738191464546,
      "grad_norm": 4.4087324142456055,
      "learning_rate": 4.435782002780731e-06,
      "loss": 0.3795,
      "mean_token_accuracy": 0.8499569103121758,
      "num_tokens": 126136484.0,
      "step": 104930
    },
    {
      "entropy": 1.8943352848291397,
      "epoch": 0.32530481827150426,
      "grad_norm": 7.900808334350586,
      "learning_rate": 4.435570647246417e-06,
      "loss": 0.5372,
      "mean_token_accuracy": 0.8378649175167083,
      "num_tokens": 126147568.0,
      "step": 104940
    },
    {
      "entropy": 1.7571329042315482,
      "epoch": 0.325335817396554,
      "grad_norm": 5.164160251617432,
      "learning_rate": 4.435359321921144e-06,
      "loss": 0.4033,
      "mean_token_accuracy": 0.8640245348215103,
      "num_tokens": 126161243.0,
      "step": 104950
    },
    {
      "entropy": 1.8740806519985198,
      "epoch": 0.32536681652160365,
      "grad_norm": 7.368401050567627,
      "learning_rate": 4.435148026797714e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8456128895282745,
      "num_tokens": 126173067.0,
      "step": 104960
    },
    {
      "entropy": 1.852415107935667,
      "epoch": 0.3253978156466534,
      "grad_norm": 8.494685173034668,
      "learning_rate": 4.434936761868937e-06,
      "loss": 0.4704,
      "mean_token_accuracy": 0.8431513279676437,
      "num_tokens": 126185590.0,
      "step": 104970
    },
    {
      "entropy": 1.872846459597349,
      "epoch": 0.32542881477170305,
      "grad_norm": 2.7987353801727295,
      "learning_rate": 4.434725527127619e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.8551968723535538,
      "num_tokens": 126197812.0,
      "step": 104980
    },
    {
      "entropy": 1.8627401351928712,
      "epoch": 0.32545981389675277,
      "grad_norm": 4.279714584350586,
      "learning_rate": 4.434514322566573e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8442763239145279,
      "num_tokens": 126210066.0,
      "step": 104990
    },
    {
      "entropy": 1.88872180134058,
      "epoch": 0.32549081302180244,
      "grad_norm": 7.514123439788818,
      "learning_rate": 4.434303148178613e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8437312439084053,
      "num_tokens": 126222842.0,
      "step": 105000
    },
    {
      "entropy": 1.9054294735193253,
      "epoch": 0.32552181214685216,
      "grad_norm": 7.55472993850708,
      "learning_rate": 4.434092003956556e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8568596675992012,
      "num_tokens": 126233568.0,
      "step": 105010
    },
    {
      "entropy": 1.9124674081802369,
      "epoch": 0.32555281127190183,
      "grad_norm": 10.53081226348877,
      "learning_rate": 4.4338808898932204e-06,
      "loss": 0.5074,
      "mean_token_accuracy": 0.84600650370121,
      "num_tokens": 126245643.0,
      "step": 105020
    },
    {
      "entropy": 1.8103389233350753,
      "epoch": 0.32558381039695156,
      "grad_norm": 3.0053796768188477,
      "learning_rate": 4.433669805981426e-06,
      "loss": 0.434,
      "mean_token_accuracy": 0.8554301127791405,
      "num_tokens": 126257601.0,
      "step": 105030
    },
    {
      "entropy": 1.9138575494289398,
      "epoch": 0.3256148095220012,
      "grad_norm": 8.38183879852295,
      "learning_rate": 4.433458752213998e-06,
      "loss": 0.5186,
      "mean_token_accuracy": 0.838289988040924,
      "num_tokens": 126269101.0,
      "step": 105040
    },
    {
      "entropy": 1.8735656663775444,
      "epoch": 0.32564580864705095,
      "grad_norm": 8.337263107299805,
      "learning_rate": 4.433247728583761e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8512678787112236,
      "num_tokens": 126279995.0,
      "step": 105050
    },
    {
      "entropy": 1.8865850910544395,
      "epoch": 0.3256768077721006,
      "grad_norm": 10.122349739074707,
      "learning_rate": 4.433036735083546e-06,
      "loss": 0.456,
      "mean_token_accuracy": 0.8520815536379814,
      "num_tokens": 126291441.0,
      "step": 105060
    },
    {
      "entropy": 1.8525515541434288,
      "epoch": 0.32570780689715034,
      "grad_norm": 9.650886535644531,
      "learning_rate": 4.43282577170618e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8546164348721504,
      "num_tokens": 126303505.0,
      "step": 105070
    },
    {
      "entropy": 1.813267020881176,
      "epoch": 0.3257388060222,
      "grad_norm": 8.041412353515625,
      "learning_rate": 4.432614838444499e-06,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8510597214102745,
      "num_tokens": 126316240.0,
      "step": 105080
    },
    {
      "entropy": 1.837722858786583,
      "epoch": 0.32576980514724974,
      "grad_norm": 7.805492401123047,
      "learning_rate": 4.432403935291336e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8534246236085892,
      "num_tokens": 126328521.0,
      "step": 105090
    },
    {
      "entropy": 1.916102121770382,
      "epoch": 0.3258008042722994,
      "grad_norm": 8.766565322875977,
      "learning_rate": 4.432193062239532e-06,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8374799504876137,
      "num_tokens": 126340386.0,
      "step": 105100
    },
    {
      "entropy": 1.894291676580906,
      "epoch": 0.32583180339734913,
      "grad_norm": 7.835315227508545,
      "learning_rate": 4.431982219281925e-06,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8560818731784821,
      "num_tokens": 126351585.0,
      "step": 105110
    },
    {
      "entropy": 1.9061674028635025,
      "epoch": 0.3258628025223988,
      "grad_norm": 8.130203247070312,
      "learning_rate": 4.431771406411358e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8496122330427169,
      "num_tokens": 126362872.0,
      "step": 105120
    },
    {
      "entropy": 1.980933591723442,
      "epoch": 0.3258938016474485,
      "grad_norm": 8.286996841430664,
      "learning_rate": 4.431560623620675e-06,
      "loss": 0.5517,
      "mean_token_accuracy": 0.8369853526353837,
      "num_tokens": 126374021.0,
      "step": 105130
    },
    {
      "entropy": 1.8085208341479302,
      "epoch": 0.3259248007724982,
      "grad_norm": 6.969970226287842,
      "learning_rate": 4.431349870902727e-06,
      "loss": 0.4279,
      "mean_token_accuracy": 0.8551606863737107,
      "num_tokens": 126387654.0,
      "step": 105140
    },
    {
      "entropy": 1.9031474754214286,
      "epoch": 0.3259557998975479,
      "grad_norm": 7.189389705657959,
      "learning_rate": 4.431139148250362e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8574759498238563,
      "num_tokens": 126398864.0,
      "step": 105150
    },
    {
      "entropy": 1.7899470388889314,
      "epoch": 0.3259867990225976,
      "grad_norm": 7.750818252563477,
      "learning_rate": 4.430928455656429e-06,
      "loss": 0.3847,
      "mean_token_accuracy": 0.8657332420349121,
      "num_tokens": 126412367.0,
      "step": 105160
    },
    {
      "entropy": 1.812622857093811,
      "epoch": 0.3260177981476473,
      "grad_norm": 6.185579776763916,
      "learning_rate": 4.4307177931137864e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8634996026754379,
      "num_tokens": 126424902.0,
      "step": 105170
    },
    {
      "entropy": 1.9225602239370345,
      "epoch": 0.326048797272697,
      "grad_norm": 9.488457679748535,
      "learning_rate": 4.4305071606152906e-06,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8386317417025566,
      "num_tokens": 126436598.0,
      "step": 105180
    },
    {
      "entropy": 1.8649600803852082,
      "epoch": 0.32607979639774665,
      "grad_norm": 9.695698738098145,
      "learning_rate": 4.4302965581538e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8493688553571701,
      "num_tokens": 126447796.0,
      "step": 105190
    },
    {
      "entropy": 1.8659547924995423,
      "epoch": 0.32611079552279637,
      "grad_norm": 8.478187561035156,
      "learning_rate": 4.4300859857221765e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8515862852334977,
      "num_tokens": 126460081.0,
      "step": 105200
    },
    {
      "entropy": 1.8974830508232117,
      "epoch": 0.32614179464784604,
      "grad_norm": 4.3100457191467285,
      "learning_rate": 4.429875443313283e-06,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8492521673440934,
      "num_tokens": 126471858.0,
      "step": 105210
    },
    {
      "entropy": 1.8549850180745124,
      "epoch": 0.32617279377289576,
      "grad_norm": 8.894652366638184,
      "learning_rate": 4.429664930919989e-06,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8504776194691658,
      "num_tokens": 126483795.0,
      "step": 105220
    },
    {
      "entropy": 1.8823614329099656,
      "epoch": 0.32620379289794543,
      "grad_norm": 8.619571685791016,
      "learning_rate": 4.429454448535162e-06,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8416207283735275,
      "num_tokens": 126495379.0,
      "step": 105230
    },
    {
      "entropy": 1.8804425790905952,
      "epoch": 0.32623479202299516,
      "grad_norm": 8.832228660583496,
      "learning_rate": 4.429243996151671e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8424814388155937,
      "num_tokens": 126507418.0,
      "step": 105240
    },
    {
      "entropy": 1.9131366968154908,
      "epoch": 0.3262657911480448,
      "grad_norm": 9.161233901977539,
      "learning_rate": 4.4290335737623915e-06,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8429681926965713,
      "num_tokens": 126519465.0,
      "step": 105250
    },
    {
      "entropy": 1.8811295062303544,
      "epoch": 0.32629679027309455,
      "grad_norm": 7.556185245513916,
      "learning_rate": 4.4288231813602e-06,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8429519325494766,
      "num_tokens": 126531139.0,
      "step": 105260
    },
    {
      "entropy": 1.9287168130278587,
      "epoch": 0.3263277893981442,
      "grad_norm": 8.719147682189941,
      "learning_rate": 4.428612818937974e-06,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8285974323749542,
      "num_tokens": 126542914.0,
      "step": 105270
    },
    {
      "entropy": 1.8551797360181808,
      "epoch": 0.32635878852319394,
      "grad_norm": 3.8900067806243896,
      "learning_rate": 4.428402486488593e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8506741672754288,
      "num_tokens": 126554599.0,
      "step": 105280
    },
    {
      "entropy": 1.9169548124074935,
      "epoch": 0.3263897876482436,
      "grad_norm": 7.38561487197876,
      "learning_rate": 4.428192184004942e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8483023956418038,
      "num_tokens": 126565301.0,
      "step": 105290
    },
    {
      "entropy": 1.9362420484423637,
      "epoch": 0.32642078677329334,
      "grad_norm": 8.075181007385254,
      "learning_rate": 4.427981911479907e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.837260690331459,
      "num_tokens": 126576660.0,
      "step": 105300
    },
    {
      "entropy": 1.9895741552114488,
      "epoch": 0.326451785898343,
      "grad_norm": 8.598695755004883,
      "learning_rate": 4.427771668906373e-06,
      "loss": 0.5641,
      "mean_token_accuracy": 0.8313561111688614,
      "num_tokens": 126587261.0,
      "step": 105310
    },
    {
      "entropy": 1.9371146634221077,
      "epoch": 0.32648278502339273,
      "grad_norm": 4.246029376983643,
      "learning_rate": 4.427561456277231e-06,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8425526946783066,
      "num_tokens": 126598325.0,
      "step": 105320
    },
    {
      "entropy": 1.897105310857296,
      "epoch": 0.3265137841484424,
      "grad_norm": 4.175163269042969,
      "learning_rate": 4.427351273585373e-06,
      "loss": 0.4996,
      "mean_token_accuracy": 0.8386947169899941,
      "num_tokens": 126610456.0,
      "step": 105330
    },
    {
      "entropy": 1.9232771515846252,
      "epoch": 0.3265447832734921,
      "grad_norm": 7.9354071617126465,
      "learning_rate": 4.427141120823697e-06,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8433753445744514,
      "num_tokens": 126621448.0,
      "step": 105340
    },
    {
      "entropy": 1.9362381488084792,
      "epoch": 0.3265757823985418,
      "grad_norm": 8.758517265319824,
      "learning_rate": 4.426930997985096e-06,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8483623921871185,
      "num_tokens": 126632162.0,
      "step": 105350
    },
    {
      "entropy": 1.8089591830968856,
      "epoch": 0.3266067815235915,
      "grad_norm": 8.597465515136719,
      "learning_rate": 4.426720905062472e-06,
      "loss": 0.4347,
      "mean_token_accuracy": 0.851777009665966,
      "num_tokens": 126645035.0,
      "step": 105360
    },
    {
      "entropy": 1.9275547727942466,
      "epoch": 0.3266377806486412,
      "grad_norm": 7.728219509124756,
      "learning_rate": 4.426510842048728e-06,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8419729739427566,
      "num_tokens": 126656366.0,
      "step": 105370
    },
    {
      "entropy": 1.947744831442833,
      "epoch": 0.3266687797736909,
      "grad_norm": 8.231205940246582,
      "learning_rate": 4.426300808936765e-06,
      "loss": 0.5361,
      "mean_token_accuracy": 0.8379727810621261,
      "num_tokens": 126667203.0,
      "step": 105380
    },
    {
      "entropy": 1.89700628221035,
      "epoch": 0.3266997788987406,
      "grad_norm": 8.626848220825195,
      "learning_rate": 4.426090805719492e-06,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8507444426417351,
      "num_tokens": 126679228.0,
      "step": 105390
    },
    {
      "entropy": 1.7944568783044814,
      "epoch": 0.3267307780237903,
      "grad_norm": 7.826314926147461,
      "learning_rate": 4.4258808323898175e-06,
      "loss": 0.3949,
      "mean_token_accuracy": 0.8589142486453056,
      "num_tokens": 126692017.0,
      "step": 105400
    },
    {
      "entropy": 1.8504990950226783,
      "epoch": 0.32676177714883997,
      "grad_norm": 8.685942649841309,
      "learning_rate": 4.425670888940653e-06,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8543372765183449,
      "num_tokens": 126704537.0,
      "step": 105410
    },
    {
      "entropy": 1.774632167816162,
      "epoch": 0.3267927762738897,
      "grad_norm": 8.145844459533691,
      "learning_rate": 4.425460975364912e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.853350467979908,
      "num_tokens": 126717875.0,
      "step": 105420
    },
    {
      "entropy": 1.878597044944763,
      "epoch": 0.32682377539893936,
      "grad_norm": 8.8296480178833,
      "learning_rate": 4.425251091655509e-06,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8486263751983643,
      "num_tokens": 126729109.0,
      "step": 105430
    },
    {
      "entropy": 1.8818610459566116,
      "epoch": 0.32685477452398903,
      "grad_norm": 7.772421836853027,
      "learning_rate": 4.425041237805365e-06,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8495032295584679,
      "num_tokens": 126741332.0,
      "step": 105440
    },
    {
      "entropy": 1.8752919748425483,
      "epoch": 0.32688577364903876,
      "grad_norm": 7.812198162078857,
      "learning_rate": 4.4248314138074e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8592524334788323,
      "num_tokens": 126753407.0,
      "step": 105450
    },
    {
      "entropy": 1.8762979090213776,
      "epoch": 0.3269167727740884,
      "grad_norm": 3.573888063430786,
      "learning_rate": 4.424621619654536e-06,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8430203974246979,
      "num_tokens": 126764669.0,
      "step": 105460
    },
    {
      "entropy": 1.9314564675092698,
      "epoch": 0.32694777189913815,
      "grad_norm": 8.204230308532715,
      "learning_rate": 4.4244118553397e-06,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8395652711391449,
      "num_tokens": 126775741.0,
      "step": 105470
    },
    {
      "entropy": 1.8367981255054473,
      "epoch": 0.3269787710241878,
      "grad_norm": 8.883301734924316,
      "learning_rate": 4.424202120855818e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8451298654079438,
      "num_tokens": 126788018.0,
      "step": 105480
    },
    {
      "entropy": 1.9303217083215714,
      "epoch": 0.32700977014923754,
      "grad_norm": 7.516885757446289,
      "learning_rate": 4.423992416195822e-06,
      "loss": 0.524,
      "mean_token_accuracy": 0.8447012811899185,
      "num_tokens": 126798946.0,
      "step": 105490
    },
    {
      "entropy": 1.8869639337062836,
      "epoch": 0.3270407692742872,
      "grad_norm": 7.857840538024902,
      "learning_rate": 4.4237827413526425e-06,
      "loss": 0.5241,
      "mean_token_accuracy": 0.8342716425657273,
      "num_tokens": 126810950.0,
      "step": 105500
    },
    {
      "entropy": 1.8075002878904343,
      "epoch": 0.32707176839933694,
      "grad_norm": 3.914418935775757,
      "learning_rate": 4.423573096319217e-06,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8586590677499771,
      "num_tokens": 126823453.0,
      "step": 105510
    },
    {
      "entropy": 1.898749789595604,
      "epoch": 0.3271027675243866,
      "grad_norm": 8.268956184387207,
      "learning_rate": 4.423363481088481e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8393109634518623,
      "num_tokens": 126835289.0,
      "step": 105520
    },
    {
      "entropy": 1.7726179018616677,
      "epoch": 0.32713376664943633,
      "grad_norm": 3.696336030960083,
      "learning_rate": 4.423153895653373e-06,
      "loss": 0.4025,
      "mean_token_accuracy": 0.8711417764425278,
      "num_tokens": 126847849.0,
      "step": 105530
    },
    {
      "entropy": 1.7812189802527427,
      "epoch": 0.327164765774486,
      "grad_norm": 7.559261798858643,
      "learning_rate": 4.422944340006837e-06,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8589012801647187,
      "num_tokens": 126860795.0,
      "step": 105540
    },
    {
      "entropy": 1.869552193582058,
      "epoch": 0.3271957648995357,
      "grad_norm": 7.7391252517700195,
      "learning_rate": 4.4227348141418165e-06,
      "loss": 0.5163,
      "mean_token_accuracy": 0.8421175703406334,
      "num_tokens": 126872641.0,
      "step": 105550
    },
    {
      "entropy": 1.847855243086815,
      "epoch": 0.3272267640245854,
      "grad_norm": 8.301652908325195,
      "learning_rate": 4.422525318051257e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8523299112915993,
      "num_tokens": 126884824.0,
      "step": 105560
    },
    {
      "entropy": 1.8750713467597961,
      "epoch": 0.3272577631496351,
      "grad_norm": 9.099956512451172,
      "learning_rate": 4.422315851728109e-06,
      "loss": 0.4838,
      "mean_token_accuracy": 0.848430560529232,
      "num_tokens": 126896206.0,
      "step": 105570
    },
    {
      "entropy": 1.8710100874304771,
      "epoch": 0.3272887622746848,
      "grad_norm": 7.930703163146973,
      "learning_rate": 4.422106415165322e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8564017862081528,
      "num_tokens": 126907868.0,
      "step": 105580
    },
    {
      "entropy": 1.9006529331207276,
      "epoch": 0.3273197613997345,
      "grad_norm": 3.9899120330810547,
      "learning_rate": 4.4218970083558505e-06,
      "loss": 0.5064,
      "mean_token_accuracy": 0.8440683215856553,
      "num_tokens": 126919807.0,
      "step": 105590
    },
    {
      "entropy": 1.8714349627494813,
      "epoch": 0.3273507605247842,
      "grad_norm": 8.205395698547363,
      "learning_rate": 4.421687631292651e-06,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8427043676376342,
      "num_tokens": 126931174.0,
      "step": 105600
    },
    {
      "entropy": 1.8943498641252519,
      "epoch": 0.3273817596498339,
      "grad_norm": 5.008574485778809,
      "learning_rate": 4.4214782839686805e-06,
      "loss": 0.4976,
      "mean_token_accuracy": 0.8429151371121406,
      "num_tokens": 126942843.0,
      "step": 105610
    },
    {
      "entropy": 1.8473748803138732,
      "epoch": 0.32741275877488357,
      "grad_norm": 8.37645435333252,
      "learning_rate": 4.421268966376901e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8443385511636734,
      "num_tokens": 126954725.0,
      "step": 105620
    },
    {
      "entropy": 1.7291858568787575,
      "epoch": 0.3274437578999333,
      "grad_norm": 3.739603042602539,
      "learning_rate": 4.421059678510274e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8665387943387032,
      "num_tokens": 126968620.0,
      "step": 105630
    },
    {
      "entropy": 1.7893247798085212,
      "epoch": 0.32747475702498297,
      "grad_norm": 4.73966121673584,
      "learning_rate": 4.420850420361765e-06,
      "loss": 0.3692,
      "mean_token_accuracy": 0.8631637006998062,
      "num_tokens": 126981603.0,
      "step": 105640
    },
    {
      "entropy": 1.8417418763041495,
      "epoch": 0.3275057561500327,
      "grad_norm": 3.766573190689087,
      "learning_rate": 4.420641191924342e-06,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8520455986261368,
      "num_tokens": 126994028.0,
      "step": 105650
    },
    {
      "entropy": 1.8928628534078598,
      "epoch": 0.32753675527508236,
      "grad_norm": 10.196480751037598,
      "learning_rate": 4.420431993190975e-06,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8365366339683533,
      "num_tokens": 127005092.0,
      "step": 105660
    },
    {
      "entropy": 1.8437321752309799,
      "epoch": 0.3275677544001321,
      "grad_norm": 3.234116792678833,
      "learning_rate": 4.4202228241546354e-06,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8564884319901467,
      "num_tokens": 127017233.0,
      "step": 105670
    },
    {
      "entropy": 1.8227815330028534,
      "epoch": 0.32759875352518175,
      "grad_norm": 6.8498921394348145,
      "learning_rate": 4.420013684808299e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8651960000395775,
      "num_tokens": 127028817.0,
      "step": 105680
    },
    {
      "entropy": 1.8443077996373176,
      "epoch": 0.3276297526502314,
      "grad_norm": 5.660989761352539,
      "learning_rate": 4.419804575144942e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8490277513861656,
      "num_tokens": 127041425.0,
      "step": 105690
    },
    {
      "entropy": 1.8642980486154557,
      "epoch": 0.32766075177528114,
      "grad_norm": 7.390862464904785,
      "learning_rate": 4.419595495157543e-06,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8533131554722786,
      "num_tokens": 127053051.0,
      "step": 105700
    },
    {
      "entropy": 1.906336459517479,
      "epoch": 0.3276917509003308,
      "grad_norm": 7.918664455413818,
      "learning_rate": 4.419386444839084e-06,
      "loss": 0.536,
      "mean_token_accuracy": 0.8386664062738418,
      "num_tokens": 127064707.0,
      "step": 105710
    },
    {
      "entropy": 1.8976897805929185,
      "epoch": 0.32772275002538054,
      "grad_norm": 8.025975227355957,
      "learning_rate": 4.419177424182549e-06,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8574541077017784,
      "num_tokens": 127075139.0,
      "step": 105720
    },
    {
      "entropy": 1.7558037593960762,
      "epoch": 0.3277537491504302,
      "grad_norm": 8.160879135131836,
      "learning_rate": 4.418968433180924e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8632358491420746,
      "num_tokens": 127088887.0,
      "step": 105730
    },
    {
      "entropy": 1.7168943211436272,
      "epoch": 0.32778474827547993,
      "grad_norm": 4.3874359130859375,
      "learning_rate": 4.418759471827199e-06,
      "loss": 0.3729,
      "mean_token_accuracy": 0.861461877822876,
      "num_tokens": 127103106.0,
      "step": 105740
    },
    {
      "entropy": 1.84348274320364,
      "epoch": 0.3278157474005296,
      "grad_norm": 4.105869770050049,
      "learning_rate": 4.418550540114362e-06,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8481574520468712,
      "num_tokens": 127115076.0,
      "step": 105750
    },
    {
      "entropy": 1.803061343729496,
      "epoch": 0.3278467465255793,
      "grad_norm": 4.1507368087768555,
      "learning_rate": 4.418341638035409e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8516246452927589,
      "num_tokens": 127127443.0,
      "step": 105760
    },
    {
      "entropy": 1.821767383813858,
      "epoch": 0.327877745650629,
      "grad_norm": 3.866901159286499,
      "learning_rate": 4.4181327655833315e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8582574099302291,
      "num_tokens": 127139967.0,
      "step": 105770
    },
    {
      "entropy": 1.8919250801205636,
      "epoch": 0.3279087447756787,
      "grad_norm": 7.636270999908447,
      "learning_rate": 4.417923922751132e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8451293349266052,
      "num_tokens": 127151560.0,
      "step": 105780
    },
    {
      "entropy": 1.8248420521616935,
      "epoch": 0.3279397439007284,
      "grad_norm": 3.9394943714141846,
      "learning_rate": 4.417715109531807e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8457361742854118,
      "num_tokens": 127163965.0,
      "step": 105790
    },
    {
      "entropy": 1.8081183552742004,
      "epoch": 0.3279707430257781,
      "grad_norm": 9.723323822021484,
      "learning_rate": 4.41750632591836e-06,
      "loss": 0.4581,
      "mean_token_accuracy": 0.856559830904007,
      "num_tokens": 127176304.0,
      "step": 105800
    },
    {
      "entropy": 1.9597806930541992,
      "epoch": 0.3280017421508278,
      "grad_norm": 9.514747619628906,
      "learning_rate": 4.417297571903797e-06,
      "loss": 0.5702,
      "mean_token_accuracy": 0.823980063199997,
      "num_tokens": 127187023.0,
      "step": 105810
    },
    {
      "entropy": 1.8679911822080613,
      "epoch": 0.3280327412758775,
      "grad_norm": 8.901237487792969,
      "learning_rate": 4.4170888474811235e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8576633468270302,
      "num_tokens": 127199443.0,
      "step": 105820
    },
    {
      "entropy": 1.7843319460749627,
      "epoch": 0.3280637404009272,
      "grad_norm": 7.45582914352417,
      "learning_rate": 4.4168801526433495e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.8751977890729904,
      "num_tokens": 127212000.0,
      "step": 105830
    },
    {
      "entropy": 1.8773673444986343,
      "epoch": 0.3280947395259769,
      "grad_norm": 7.825649261474609,
      "learning_rate": 4.416671487383486e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8506253302097321,
      "num_tokens": 127222912.0,
      "step": 105840
    },
    {
      "entropy": 1.9522982880473136,
      "epoch": 0.32812573865102657,
      "grad_norm": 11.103384017944336,
      "learning_rate": 4.416462851694547e-06,
      "loss": 0.5427,
      "mean_token_accuracy": 0.8257913753390312,
      "num_tokens": 127234149.0,
      "step": 105850
    },
    {
      "entropy": 1.9224087953567506,
      "epoch": 0.3281567377760763,
      "grad_norm": 8.676215171813965,
      "learning_rate": 4.4162542455695495e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8379293918609619,
      "num_tokens": 127245446.0,
      "step": 105860
    },
    {
      "entropy": 1.8061835870146752,
      "epoch": 0.32818773690112596,
      "grad_norm": 3.818633794784546,
      "learning_rate": 4.416045669001512e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8506641060113906,
      "num_tokens": 127258271.0,
      "step": 105870
    },
    {
      "entropy": 1.891014301776886,
      "epoch": 0.3282187360261757,
      "grad_norm": 8.472275733947754,
      "learning_rate": 4.415837121983454e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8417830243706703,
      "num_tokens": 127269244.0,
      "step": 105880
    },
    {
      "entropy": 1.8804811149835587,
      "epoch": 0.32824973515122535,
      "grad_norm": 7.921799659729004,
      "learning_rate": 4.415628604508402e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.843516594171524,
      "num_tokens": 127281119.0,
      "step": 105890
    },
    {
      "entropy": 1.8675379946827888,
      "epoch": 0.3282807342762751,
      "grad_norm": 7.780767917633057,
      "learning_rate": 4.415420116569378e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8497234523296356,
      "num_tokens": 127292982.0,
      "step": 105900
    },
    {
      "entropy": 1.8299174696207046,
      "epoch": 0.32831173340132475,
      "grad_norm": 8.369937896728516,
      "learning_rate": 4.41521165815941e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8549463152885437,
      "num_tokens": 127305007.0,
      "step": 105910
    },
    {
      "entropy": 1.902072112262249,
      "epoch": 0.32834273252637447,
      "grad_norm": 6.59075403213501,
      "learning_rate": 4.4150032292715315e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.84717957675457,
      "num_tokens": 127316897.0,
      "step": 105920
    },
    {
      "entropy": 1.9103587403893472,
      "epoch": 0.32837373165142414,
      "grad_norm": 9.852415084838867,
      "learning_rate": 4.414794829898772e-06,
      "loss": 0.518,
      "mean_token_accuracy": 0.8442836970090866,
      "num_tokens": 127328580.0,
      "step": 105930
    },
    {
      "entropy": 1.8694602742791175,
      "epoch": 0.3284047307764738,
      "grad_norm": 9.172961235046387,
      "learning_rate": 4.4145864600341656e-06,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8534001842141151,
      "num_tokens": 127339695.0,
      "step": 105940
    },
    {
      "entropy": 1.8509365767240524,
      "epoch": 0.32843572990152353,
      "grad_norm": 7.372804641723633,
      "learning_rate": 4.414378119670751e-06,
      "loss": 0.4497,
      "mean_token_accuracy": 0.85338544100523,
      "num_tokens": 127350996.0,
      "step": 105950
    },
    {
      "entropy": 1.9594619423151016,
      "epoch": 0.3284667290265732,
      "grad_norm": 9.022714614868164,
      "learning_rate": 4.414169808801567e-06,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8271567165851593,
      "num_tokens": 127361650.0,
      "step": 105960
    },
    {
      "entropy": 1.788424487411976,
      "epoch": 0.3284977281516229,
      "grad_norm": 8.611357688903809,
      "learning_rate": 4.413961527419656e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8564797580242157,
      "num_tokens": 127375584.0,
      "step": 105970
    },
    {
      "entropy": 1.8492983341217042,
      "epoch": 0.3285287272766726,
      "grad_norm": 6.395881175994873,
      "learning_rate": 4.4137532755180615e-06,
      "loss": 0.4253,
      "mean_token_accuracy": 0.8596012353897095,
      "num_tokens": 127388011.0,
      "step": 105980
    },
    {
      "entropy": 1.9068158611655235,
      "epoch": 0.3285597264017223,
      "grad_norm": 9.72159481048584,
      "learning_rate": 4.4135450530898296e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8352061986923218,
      "num_tokens": 127399562.0,
      "step": 105990
    },
    {
      "entropy": 1.8263181149959564,
      "epoch": 0.328590725526772,
      "grad_norm": 9.199592590332031,
      "learning_rate": 4.413336860128008e-06,
      "loss": 0.4156,
      "mean_token_accuracy": 0.856217360496521,
      "num_tokens": 127412251.0,
      "step": 106000
    },
    {
      "entropy": 1.7998489513993263,
      "epoch": 0.3286217246518217,
      "grad_norm": 7.553385257720947,
      "learning_rate": 4.413128696625648e-06,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8605973869562149,
      "num_tokens": 127424851.0,
      "step": 106010
    },
    {
      "entropy": 1.8556075662374496,
      "epoch": 0.3286527237768714,
      "grad_norm": 9.884352684020996,
      "learning_rate": 4.412920562575802e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8411578819155693,
      "num_tokens": 127436563.0,
      "step": 106020
    },
    {
      "entropy": 1.8433560684323311,
      "epoch": 0.3286837229019211,
      "grad_norm": 9.31692123413086,
      "learning_rate": 4.412712457971527e-06,
      "loss": 0.4248,
      "mean_token_accuracy": 0.85454620718956,
      "num_tokens": 127448302.0,
      "step": 106030
    },
    {
      "entropy": 1.8037659108638764,
      "epoch": 0.3287147220269708,
      "grad_norm": 4.995815753936768,
      "learning_rate": 4.412504382805881e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8571658223867417,
      "num_tokens": 127459774.0,
      "step": 106040
    },
    {
      "entropy": 1.8213044986128808,
      "epoch": 0.3287457211520205,
      "grad_norm": 8.79011344909668,
      "learning_rate": 4.412296337071922e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8526079818606377,
      "num_tokens": 127472645.0,
      "step": 106050
    },
    {
      "entropy": 1.9069361835718155,
      "epoch": 0.32877672027707017,
      "grad_norm": 9.181472778320312,
      "learning_rate": 4.412088320762712e-06,
      "loss": 0.481,
      "mean_token_accuracy": 0.8438616082072258,
      "num_tokens": 127483985.0,
      "step": 106060
    },
    {
      "entropy": 1.7741859510540963,
      "epoch": 0.3288077194021199,
      "grad_norm": 4.265276908874512,
      "learning_rate": 4.411880333871319e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.85208670347929,
      "num_tokens": 127497519.0,
      "step": 106070
    },
    {
      "entropy": 1.774897436797619,
      "epoch": 0.32883871852716956,
      "grad_norm": 3.4330549240112305,
      "learning_rate": 4.411672376390806e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.862097978591919,
      "num_tokens": 127510803.0,
      "step": 106080
    },
    {
      "entropy": 1.7656168833374977,
      "epoch": 0.3288697176522193,
      "grad_norm": 7.550422191619873,
      "learning_rate": 4.411464448314243e-06,
      "loss": 0.3932,
      "mean_token_accuracy": 0.859065192937851,
      "num_tokens": 127523417.0,
      "step": 106090
    },
    {
      "entropy": 1.8322931870818138,
      "epoch": 0.32890071677726895,
      "grad_norm": 8.406190872192383,
      "learning_rate": 4.411256549634704e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8471446886658669,
      "num_tokens": 127536405.0,
      "step": 106100
    },
    {
      "entropy": 1.914437648653984,
      "epoch": 0.3289317159023187,
      "grad_norm": 10.335504531860352,
      "learning_rate": 4.411048680345259e-06,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8394327759742737,
      "num_tokens": 127546905.0,
      "step": 106110
    },
    {
      "entropy": 1.863400039076805,
      "epoch": 0.32896271502736835,
      "grad_norm": 7.929013252258301,
      "learning_rate": 4.410840840438987e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8584112614393234,
      "num_tokens": 127558189.0,
      "step": 106120
    },
    {
      "entropy": 1.8088774591684342,
      "epoch": 0.32899371415241807,
      "grad_norm": 12.394187927246094,
      "learning_rate": 4.410633029908964e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8598813712596893,
      "num_tokens": 127570793.0,
      "step": 106130
    },
    {
      "entropy": 1.8091413155198097,
      "epoch": 0.32902471327746774,
      "grad_norm": 7.606610298156738,
      "learning_rate": 4.4104252487482726e-06,
      "loss": 0.4373,
      "mean_token_accuracy": 0.8585825636982918,
      "num_tokens": 127582513.0,
      "step": 106140
    },
    {
      "entropy": 1.8211368069052696,
      "epoch": 0.32905571240251746,
      "grad_norm": 3.8526268005371094,
      "learning_rate": 4.4102174969499945e-06,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8475786969065666,
      "num_tokens": 127595268.0,
      "step": 106150
    },
    {
      "entropy": 1.8827776312828064,
      "epoch": 0.32908671152756713,
      "grad_norm": 8.69912052154541,
      "learning_rate": 4.410009774507214e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8499000027775765,
      "num_tokens": 127606870.0,
      "step": 106160
    },
    {
      "entropy": 1.9105200260877608,
      "epoch": 0.32911771065261686,
      "grad_norm": 8.402679443359375,
      "learning_rate": 4.40980208141302e-06,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8454275488853454,
      "num_tokens": 127617735.0,
      "step": 106170
    },
    {
      "entropy": 1.8559418886899948,
      "epoch": 0.3291487097776665,
      "grad_norm": 10.870802879333496,
      "learning_rate": 4.4095944176605015e-06,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8455240190029144,
      "num_tokens": 127629429.0,
      "step": 106180
    },
    {
      "entropy": 1.8259831920266152,
      "epoch": 0.3291797089027162,
      "grad_norm": 8.05710506439209,
      "learning_rate": 4.40938678324275e-06,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8483387231826782,
      "num_tokens": 127642106.0,
      "step": 106190
    },
    {
      "entropy": 1.8355090886354446,
      "epoch": 0.3292107080277659,
      "grad_norm": 6.109237194061279,
      "learning_rate": 4.40917917815286e-06,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8482315883040428,
      "num_tokens": 127653739.0,
      "step": 106200
    },
    {
      "entropy": 1.8437444880604743,
      "epoch": 0.3292417071528156,
      "grad_norm": 3.581789493560791,
      "learning_rate": 4.408971602383929e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8457508951425552,
      "num_tokens": 127665387.0,
      "step": 106210
    },
    {
      "entropy": 1.7818902999162674,
      "epoch": 0.3292727062778653,
      "grad_norm": 8.96249771118164,
      "learning_rate": 4.408764055929055e-06,
      "loss": 0.4116,
      "mean_token_accuracy": 0.8549827590584755,
      "num_tokens": 127678366.0,
      "step": 106220
    },
    {
      "entropy": 1.8756959035992622,
      "epoch": 0.329303705402915,
      "grad_norm": 8.410722732543945,
      "learning_rate": 4.40855653878134e-06,
      "loss": 0.4863,
      "mean_token_accuracy": 0.8546160280704498,
      "num_tokens": 127689938.0,
      "step": 106230
    },
    {
      "entropy": 1.782509133219719,
      "epoch": 0.3293347045279647,
      "grad_norm": 4.164084434509277,
      "learning_rate": 4.408349050933885e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8557716026902199,
      "num_tokens": 127703501.0,
      "step": 106240
    },
    {
      "entropy": 1.9288730174303055,
      "epoch": 0.3293657036530144,
      "grad_norm": 9.329178810119629,
      "learning_rate": 4.408141592379797e-06,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8449020832777023,
      "num_tokens": 127714411.0,
      "step": 106250
    },
    {
      "entropy": 1.9135432869195939,
      "epoch": 0.3293967027780641,
      "grad_norm": 7.811339378356934,
      "learning_rate": 4.4079341631121846e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.843287567794323,
      "num_tokens": 127725960.0,
      "step": 106260
    },
    {
      "entropy": 1.8159632161259651,
      "epoch": 0.32942770190311377,
      "grad_norm": 10.536284446716309,
      "learning_rate": 4.407726763124156e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8523865416646004,
      "num_tokens": 127738519.0,
      "step": 106270
    },
    {
      "entropy": 1.6889009296894073,
      "epoch": 0.3294587010281635,
      "grad_norm": 3.854187488555908,
      "learning_rate": 4.407519392408826e-06,
      "loss": 0.3543,
      "mean_token_accuracy": 0.8675200149416924,
      "num_tokens": 127753002.0,
      "step": 106280
    },
    {
      "entropy": 1.9123955562710762,
      "epoch": 0.32948970015321316,
      "grad_norm": 8.895185470581055,
      "learning_rate": 4.4073120509593084e-06,
      "loss": 0.5115,
      "mean_token_accuracy": 0.843984666466713,
      "num_tokens": 127764315.0,
      "step": 106290
    },
    {
      "entropy": 1.8070273652672768,
      "epoch": 0.3295206992782629,
      "grad_norm": 6.719578266143799,
      "learning_rate": 4.4071047387687186e-06,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8462962463498116,
      "num_tokens": 127777448.0,
      "step": 106300
    },
    {
      "entropy": 1.9051207095384597,
      "epoch": 0.32955169840331255,
      "grad_norm": 6.615823268890381,
      "learning_rate": 4.406897455830177e-06,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8489801079034806,
      "num_tokens": 127788724.0,
      "step": 106310
    },
    {
      "entropy": 1.790643498301506,
      "epoch": 0.3295826975283623,
      "grad_norm": 4.387211322784424,
      "learning_rate": 4.4066902021368055e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8558807000517845,
      "num_tokens": 127801391.0,
      "step": 106320
    },
    {
      "entropy": 1.9066824808716774,
      "epoch": 0.32961369665341195,
      "grad_norm": 8.245688438415527,
      "learning_rate": 4.406482977681727e-06,
      "loss": 0.5221,
      "mean_token_accuracy": 0.8399421364068985,
      "num_tokens": 127812990.0,
      "step": 106330
    },
    {
      "entropy": 1.7780104890465736,
      "epoch": 0.32964469577846167,
      "grad_norm": 7.971060276031494,
      "learning_rate": 4.406275782458069e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8576152637600899,
      "num_tokens": 127825366.0,
      "step": 106340
    },
    {
      "entropy": 1.7980201825499535,
      "epoch": 0.32967569490351134,
      "grad_norm": 10.708061218261719,
      "learning_rate": 4.406068616458957e-06,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8627904281020164,
      "num_tokens": 127837420.0,
      "step": 106350
    },
    {
      "entropy": 1.788325347006321,
      "epoch": 0.32970669402856106,
      "grad_norm": 7.981190204620361,
      "learning_rate": 4.405861479677525e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8535309046506881,
      "num_tokens": 127850016.0,
      "step": 106360
    },
    {
      "entropy": 1.8833219155669212,
      "epoch": 0.32973769315361073,
      "grad_norm": 7.901392936706543,
      "learning_rate": 4.4056543721069024e-06,
      "loss": 0.5564,
      "mean_token_accuracy": 0.83796806037426,
      "num_tokens": 127861358.0,
      "step": 106370
    },
    {
      "entropy": 1.9057514756917953,
      "epoch": 0.32976869227866046,
      "grad_norm": 7.878623962402344,
      "learning_rate": 4.405447293740227e-06,
      "loss": 0.5158,
      "mean_token_accuracy": 0.850669352710247,
      "num_tokens": 127872047.0,
      "step": 106380
    },
    {
      "entropy": 1.8584971472620964,
      "epoch": 0.3297996914037101,
      "grad_norm": 3.818636894226074,
      "learning_rate": 4.405240244570635e-06,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8438824072480202,
      "num_tokens": 127884719.0,
      "step": 106390
    },
    {
      "entropy": 1.9173836246132852,
      "epoch": 0.32983069052875985,
      "grad_norm": 9.852693557739258,
      "learning_rate": 4.405033224591264e-06,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8338365435600281,
      "num_tokens": 127896128.0,
      "step": 106400
    },
    {
      "entropy": 1.8377716928720473,
      "epoch": 0.3298616896538095,
      "grad_norm": 5.034619331359863,
      "learning_rate": 4.404826233795259e-06,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8457611680030823,
      "num_tokens": 127908855.0,
      "step": 106410
    },
    {
      "entropy": 1.7964683637022971,
      "epoch": 0.32989268877885924,
      "grad_norm": 4.223313808441162,
      "learning_rate": 4.4046192721757625e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.847607783973217,
      "num_tokens": 127921526.0,
      "step": 106420
    },
    {
      "entropy": 1.806885550916195,
      "epoch": 0.3299236879039089,
      "grad_norm": 3.7442190647125244,
      "learning_rate": 4.404412339725922e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8525063678622246,
      "num_tokens": 127934569.0,
      "step": 106430
    },
    {
      "entropy": 1.763914243876934,
      "epoch": 0.3299546870289586,
      "grad_norm": 9.370475769042969,
      "learning_rate": 4.404205436438884e-06,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8634126842021942,
      "num_tokens": 127947927.0,
      "step": 106440
    },
    {
      "entropy": 1.8372886329889297,
      "epoch": 0.3299856861540083,
      "grad_norm": 7.662943363189697,
      "learning_rate": 4.403998562307801e-06,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8420701563358307,
      "num_tokens": 127960835.0,
      "step": 106450
    },
    {
      "entropy": 1.8602828592061997,
      "epoch": 0.330016685279058,
      "grad_norm": 8.085721969604492,
      "learning_rate": 4.403791717325825e-06,
      "loss": 0.4263,
      "mean_token_accuracy": 0.8534385710954666,
      "num_tokens": 127973028.0,
      "step": 106460
    },
    {
      "entropy": 1.8481948718428611,
      "epoch": 0.3300476844041077,
      "grad_norm": 9.864887237548828,
      "learning_rate": 4.403584901486113e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8566409692168235,
      "num_tokens": 127985321.0,
      "step": 106470
    },
    {
      "entropy": 1.8503880456089974,
      "epoch": 0.33007868352915737,
      "grad_norm": 8.877874374389648,
      "learning_rate": 4.403378114781821e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8484792679548263,
      "num_tokens": 127996333.0,
      "step": 106480
    },
    {
      "entropy": 1.850000935792923,
      "epoch": 0.3301096826542071,
      "grad_norm": 8.956764221191406,
      "learning_rate": 4.403171357206109e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8442446753382683,
      "num_tokens": 128007956.0,
      "step": 106490
    },
    {
      "entropy": 1.84800376445055,
      "epoch": 0.33014068177925676,
      "grad_norm": 7.336550712585449,
      "learning_rate": 4.402964628752139e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8560839980840683,
      "num_tokens": 128019797.0,
      "step": 106500
    },
    {
      "entropy": 1.887277567386627,
      "epoch": 0.3301716809043065,
      "grad_norm": 8.57533073425293,
      "learning_rate": 4.4027579294130766e-06,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8528825983405113,
      "num_tokens": 128030487.0,
      "step": 106510
    },
    {
      "entropy": 1.8442515268921853,
      "epoch": 0.33020268002935615,
      "grad_norm": 8.904195785522461,
      "learning_rate": 4.402551259182087e-06,
      "loss": 0.4448,
      "mean_token_accuracy": 0.8501736581325531,
      "num_tokens": 128042554.0,
      "step": 106520
    },
    {
      "entropy": 1.8353326499462128,
      "epoch": 0.3302336791544059,
      "grad_norm": 7.989103317260742,
      "learning_rate": 4.402344618052339e-06,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8517720341682434,
      "num_tokens": 128054608.0,
      "step": 106530
    },
    {
      "entropy": 1.8498850226402284,
      "epoch": 0.33026467827945555,
      "grad_norm": 4.291723251342773,
      "learning_rate": 4.402138006017006e-06,
      "loss": 0.4413,
      "mean_token_accuracy": 0.848425455391407,
      "num_tokens": 128066823.0,
      "step": 106540
    },
    {
      "entropy": 1.8223819851875305,
      "epoch": 0.3302956774045053,
      "grad_norm": 4.217907428741455,
      "learning_rate": 4.401931423069258e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8563816383481025,
      "num_tokens": 128078929.0,
      "step": 106550
    },
    {
      "entropy": 1.943525066971779,
      "epoch": 0.33032667652955494,
      "grad_norm": 6.72135066986084,
      "learning_rate": 4.401724869202272e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8440445035696029,
      "num_tokens": 128089603.0,
      "step": 106560
    },
    {
      "entropy": 1.8807444974780083,
      "epoch": 0.33035767565460467,
      "grad_norm": 9.296396255493164,
      "learning_rate": 4.401518344409226e-06,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8367669865489006,
      "num_tokens": 128101462.0,
      "step": 106570
    },
    {
      "entropy": 1.8421072617173195,
      "epoch": 0.33038867477965433,
      "grad_norm": 4.264364719390869,
      "learning_rate": 4.4013118486833e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8508803278207779,
      "num_tokens": 128113035.0,
      "step": 106580
    },
    {
      "entropy": 1.9045246794819832,
      "epoch": 0.33041967390470406,
      "grad_norm": 7.94566011428833,
      "learning_rate": 4.401105382017675e-06,
      "loss": 0.5491,
      "mean_token_accuracy": 0.83597691655159,
      "num_tokens": 128124609.0,
      "step": 106590
    },
    {
      "entropy": 1.8478467270731926,
      "epoch": 0.3304506730297537,
      "grad_norm": 7.042159557342529,
      "learning_rate": 4.400898944405538e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8560745418071747,
      "num_tokens": 128136516.0,
      "step": 106600
    },
    {
      "entropy": 1.8195524469017983,
      "epoch": 0.33048167215480345,
      "grad_norm": 8.00984001159668,
      "learning_rate": 4.400692535840075e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8634827017784119,
      "num_tokens": 128148763.0,
      "step": 106610
    },
    {
      "entropy": 1.864709873497486,
      "epoch": 0.3305126712798531,
      "grad_norm": 7.909924030303955,
      "learning_rate": 4.4004861563144735e-06,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8368150681257248,
      "num_tokens": 128160851.0,
      "step": 106620
    },
    {
      "entropy": 1.8297677680850029,
      "epoch": 0.33054367040490285,
      "grad_norm": 4.41657018661499,
      "learning_rate": 4.4002798058219256e-06,
      "loss": 0.4139,
      "mean_token_accuracy": 0.8527476787567139,
      "num_tokens": 128173705.0,
      "step": 106630
    },
    {
      "entropy": 1.78669353723526,
      "epoch": 0.3305746695299525,
      "grad_norm": 8.336793899536133,
      "learning_rate": 4.400073484355625e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8534336164593697,
      "num_tokens": 128187455.0,
      "step": 106640
    },
    {
      "entropy": 1.872094802558422,
      "epoch": 0.33060566865500224,
      "grad_norm": 6.102949142456055,
      "learning_rate": 4.399867191908766e-06,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8369955107569694,
      "num_tokens": 128199573.0,
      "step": 106650
    },
    {
      "entropy": 1.8278701439499856,
      "epoch": 0.3306366677800519,
      "grad_norm": 6.1908793449401855,
      "learning_rate": 4.399660928474549e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8451722010970115,
      "num_tokens": 128212545.0,
      "step": 106660
    },
    {
      "entropy": 1.80742659419775,
      "epoch": 0.3306676669051016,
      "grad_norm": 9.044242858886719,
      "learning_rate": 4.399454694046172e-06,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8628106907010078,
      "num_tokens": 128225069.0,
      "step": 106670
    },
    {
      "entropy": 1.8953262731432914,
      "epoch": 0.3306986660301513,
      "grad_norm": 7.473639965057373,
      "learning_rate": 4.399248488616839e-06,
      "loss": 0.4912,
      "mean_token_accuracy": 0.846845431625843,
      "num_tokens": 128236139.0,
      "step": 106680
    },
    {
      "entropy": 1.8824147969484328,
      "epoch": 0.33072966515520097,
      "grad_norm": 3.392375946044922,
      "learning_rate": 4.399042312179753e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.847435437142849,
      "num_tokens": 128247581.0,
      "step": 106690
    },
    {
      "entropy": 1.8657734468579292,
      "epoch": 0.3307606642802507,
      "grad_norm": 8.0246000289917,
      "learning_rate": 4.398836164728121e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8519965663552285,
      "num_tokens": 128259094.0,
      "step": 106700
    },
    {
      "entropy": 1.9068785265088082,
      "epoch": 0.33079166340530036,
      "grad_norm": 10.08130168914795,
      "learning_rate": 4.398630046255153e-06,
      "loss": 0.4959,
      "mean_token_accuracy": 0.8427546977996826,
      "num_tokens": 128270336.0,
      "step": 106710
    },
    {
      "entropy": 1.6765178635716438,
      "epoch": 0.3308226625303501,
      "grad_norm": 3.5709166526794434,
      "learning_rate": 4.39842395675406e-06,
      "loss": 0.3849,
      "mean_token_accuracy": 0.8611236557364463,
      "num_tokens": 128284344.0,
      "step": 106720
    },
    {
      "entropy": 1.8389539405703546,
      "epoch": 0.33085366165539976,
      "grad_norm": 11.363747596740723,
      "learning_rate": 4.398217896218056e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8520830184221267,
      "num_tokens": 128296693.0,
      "step": 106730
    },
    {
      "entropy": 1.807772381603718,
      "epoch": 0.3308846607804495,
      "grad_norm": 4.15220308303833,
      "learning_rate": 4.398011864640357e-06,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8365904971957207,
      "num_tokens": 128309280.0,
      "step": 106740
    },
    {
      "entropy": 1.8760442197322846,
      "epoch": 0.33091565990549915,
      "grad_norm": 8.140005111694336,
      "learning_rate": 4.397805862014179e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8473433390259743,
      "num_tokens": 128320766.0,
      "step": 106750
    },
    {
      "entropy": 1.884317898750305,
      "epoch": 0.3309466590305489,
      "grad_norm": 8.764328956604004,
      "learning_rate": 4.397599888332744e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8561802804470062,
      "num_tokens": 128332462.0,
      "step": 106760
    },
    {
      "entropy": 1.790523299574852,
      "epoch": 0.33097765815559854,
      "grad_norm": 7.192575454711914,
      "learning_rate": 4.397393943589274e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.841584998369217,
      "num_tokens": 128344386.0,
      "step": 106770
    },
    {
      "entropy": 1.765225899219513,
      "epoch": 0.33100865728064827,
      "grad_norm": 4.198668956756592,
      "learning_rate": 4.397188027776993e-06,
      "loss": 0.3832,
      "mean_token_accuracy": 0.8681812718510628,
      "num_tokens": 128357282.0,
      "step": 106780
    },
    {
      "entropy": 1.8882387548685073,
      "epoch": 0.33103965640569794,
      "grad_norm": 8.834793090820312,
      "learning_rate": 4.396982140889129e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8361070469021797,
      "num_tokens": 128368511.0,
      "step": 106790
    },
    {
      "entropy": 1.9408668845891952,
      "epoch": 0.33107065553074766,
      "grad_norm": 7.283912181854248,
      "learning_rate": 4.396776282918909e-06,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8402132421731949,
      "num_tokens": 128379340.0,
      "step": 106800
    },
    {
      "entropy": 1.9052711009979248,
      "epoch": 0.33110165465579733,
      "grad_norm": 9.804369926452637,
      "learning_rate": 4.396570453859568e-06,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8372917637228966,
      "num_tokens": 128390571.0,
      "step": 106810
    },
    {
      "entropy": 1.8480284005403518,
      "epoch": 0.33113265378084705,
      "grad_norm": 8.21649169921875,
      "learning_rate": 4.396364653704335e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8489649355411529,
      "num_tokens": 128402248.0,
      "step": 106820
    },
    {
      "entropy": 1.8601251021027565,
      "epoch": 0.3311636529058967,
      "grad_norm": 3.575690746307373,
      "learning_rate": 4.396158882446449e-06,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8458746924996376,
      "num_tokens": 128414209.0,
      "step": 106830
    },
    {
      "entropy": 1.8285237357020379,
      "epoch": 0.33119465203094645,
      "grad_norm": 9.334534645080566,
      "learning_rate": 4.3959531400791465e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8475871294736862,
      "num_tokens": 128426736.0,
      "step": 106840
    },
    {
      "entropy": 1.8750257775187493,
      "epoch": 0.3312256511559961,
      "grad_norm": 5.553659439086914,
      "learning_rate": 4.395747426595669e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8419804513454437,
      "num_tokens": 128438360.0,
      "step": 106850
    },
    {
      "entropy": 1.870549239218235,
      "epoch": 0.33125665028104584,
      "grad_norm": 6.454956531524658,
      "learning_rate": 4.395541741989258e-06,
      "loss": 0.529,
      "mean_token_accuracy": 0.8351958334445954,
      "num_tokens": 128449971.0,
      "step": 106860
    },
    {
      "entropy": 1.8797499716281891,
      "epoch": 0.3312876494060955,
      "grad_norm": 6.935102939605713,
      "learning_rate": 4.395336086253158e-06,
      "loss": 0.435,
      "mean_token_accuracy": 0.8618351891636848,
      "num_tokens": 128461856.0,
      "step": 106870
    },
    {
      "entropy": 1.8809518918395043,
      "epoch": 0.33131864853114523,
      "grad_norm": 7.828016757965088,
      "learning_rate": 4.395130459380615e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.844907422363758,
      "num_tokens": 128473604.0,
      "step": 106880
    },
    {
      "entropy": 1.8513751283288002,
      "epoch": 0.3313496476561949,
      "grad_norm": 8.020812034606934,
      "learning_rate": 4.39492486136488e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8413770824670792,
      "num_tokens": 128485181.0,
      "step": 106890
    },
    {
      "entropy": 1.8325881719589234,
      "epoch": 0.3313806467812446,
      "grad_norm": 8.994607925415039,
      "learning_rate": 4.3947192921992015e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.854704113304615,
      "num_tokens": 128497524.0,
      "step": 106900
    },
    {
      "entropy": 1.7628432139754295,
      "epoch": 0.3314116459062943,
      "grad_norm": 4.272587299346924,
      "learning_rate": 4.394513751876836e-06,
      "loss": 0.4269,
      "mean_token_accuracy": 0.853622929751873,
      "num_tokens": 128510023.0,
      "step": 106910
    },
    {
      "entropy": 1.8294406726956367,
      "epoch": 0.33144264503134396,
      "grad_norm": 7.545256614685059,
      "learning_rate": 4.394308240391038e-06,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8351798385381699,
      "num_tokens": 128522418.0,
      "step": 106920
    },
    {
      "entropy": 1.765715080499649,
      "epoch": 0.3314736441563937,
      "grad_norm": 6.0750508308410645,
      "learning_rate": 4.394102757735063e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8552381888031959,
      "num_tokens": 128535920.0,
      "step": 106930
    },
    {
      "entropy": 1.8687207847833633,
      "epoch": 0.33150464328144336,
      "grad_norm": 4.047634124755859,
      "learning_rate": 4.393897303902175e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8374577537178993,
      "num_tokens": 128548220.0,
      "step": 106940
    },
    {
      "entropy": 1.7995543777942657,
      "epoch": 0.3315356424064931,
      "grad_norm": 2.202331781387329,
      "learning_rate": 4.3936918788856334e-06,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8424097955226898,
      "num_tokens": 128561328.0,
      "step": 106950
    },
    {
      "entropy": 1.7521587029099464,
      "epoch": 0.33156664153154275,
      "grad_norm": 2.6816599369049072,
      "learning_rate": 4.393486482678704e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8565811112523078,
      "num_tokens": 128575383.0,
      "step": 106960
    },
    {
      "entropy": 1.841699157655239,
      "epoch": 0.3315976406565925,
      "grad_norm": 7.942765235900879,
      "learning_rate": 4.393281115274653e-06,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8441447481513024,
      "num_tokens": 128587295.0,
      "step": 106970
    },
    {
      "entropy": 1.8485121354460716,
      "epoch": 0.33162863978164214,
      "grad_norm": 3.5550448894500732,
      "learning_rate": 4.39307577666675e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8580849900841713,
      "num_tokens": 128598970.0,
      "step": 106980
    },
    {
      "entropy": 1.8623722419142723,
      "epoch": 0.33165963890669187,
      "grad_norm": 9.27586555480957,
      "learning_rate": 4.3928704668482655e-06,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8380335718393326,
      "num_tokens": 128611672.0,
      "step": 106990
    },
    {
      "entropy": 1.8408109113574027,
      "epoch": 0.33169063803174154,
      "grad_norm": 8.369120597839355,
      "learning_rate": 4.392665185812473e-06,
      "loss": 0.519,
      "mean_token_accuracy": 0.8391941249370575,
      "num_tokens": 128623923.0,
      "step": 107000
    },
    {
      "entropy": 1.9453389286994933,
      "epoch": 0.33172163715679126,
      "grad_norm": 8.78305435180664,
      "learning_rate": 4.392459933552647e-06,
      "loss": 0.5335,
      "mean_token_accuracy": 0.8356341332197189,
      "num_tokens": 128634685.0,
      "step": 107010
    },
    {
      "entropy": 1.8125431835651398,
      "epoch": 0.33175263628184093,
      "grad_norm": 3.5980794429779053,
      "learning_rate": 4.392254710062065e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.854391148686409,
      "num_tokens": 128648003.0,
      "step": 107020
    },
    {
      "entropy": 1.8818465709686278,
      "epoch": 0.33178363540689065,
      "grad_norm": 9.284706115722656,
      "learning_rate": 4.3920495153340094e-06,
      "loss": 0.5199,
      "mean_token_accuracy": 0.838673610985279,
      "num_tokens": 128659608.0,
      "step": 107030
    },
    {
      "entropy": 1.8223556622862815,
      "epoch": 0.3318146345319403,
      "grad_norm": 5.312669277191162,
      "learning_rate": 4.39184434936176e-06,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8526388436555863,
      "num_tokens": 128672024.0,
      "step": 107040
    },
    {
      "entropy": 1.8191813245415687,
      "epoch": 0.33184563365699005,
      "grad_norm": 9.039751052856445,
      "learning_rate": 4.3916392121386025e-06,
      "loss": 0.4316,
      "mean_token_accuracy": 0.8523767486214637,
      "num_tokens": 128684326.0,
      "step": 107050
    },
    {
      "entropy": 1.8492727160453797,
      "epoch": 0.3318766327820397,
      "grad_norm": 4.566840648651123,
      "learning_rate": 4.3914341036578225e-06,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8528471812605858,
      "num_tokens": 128696669.0,
      "step": 107060
    },
    {
      "entropy": 1.8731192171573638,
      "epoch": 0.33190763190708944,
      "grad_norm": 10.486961364746094,
      "learning_rate": 4.391229023912708e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8397856906056405,
      "num_tokens": 128708878.0,
      "step": 107070
    },
    {
      "entropy": 1.757045941054821,
      "epoch": 0.3319386310321391,
      "grad_norm": 4.080169677734375,
      "learning_rate": 4.3910239728965514e-06,
      "loss": 0.3642,
      "mean_token_accuracy": 0.8683019027113914,
      "num_tokens": 128722019.0,
      "step": 107080
    },
    {
      "entropy": 1.8372290149331092,
      "epoch": 0.33196963015718883,
      "grad_norm": 7.9230265617370605,
      "learning_rate": 4.390818950602644e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8605737552046776,
      "num_tokens": 128733732.0,
      "step": 107090
    },
    {
      "entropy": 1.7193499386310578,
      "epoch": 0.3320006292822385,
      "grad_norm": 3.894108533859253,
      "learning_rate": 4.390613957024283e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.8623716205358505,
      "num_tokens": 128747568.0,
      "step": 107100
    },
    {
      "entropy": 1.8526916906237603,
      "epoch": 0.3320316284072882,
      "grad_norm": 7.582345008850098,
      "learning_rate": 4.390408992154765e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8529738247394562,
      "num_tokens": 128758676.0,
      "step": 107110
    },
    {
      "entropy": 1.7845040023326875,
      "epoch": 0.3320626275323379,
      "grad_norm": 4.077572822570801,
      "learning_rate": 4.390204055987388e-06,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8509064018726349,
      "num_tokens": 128770630.0,
      "step": 107120
    },
    {
      "entropy": 1.9034978345036506,
      "epoch": 0.3320936266573876,
      "grad_norm": 11.147791862487793,
      "learning_rate": 4.389999148515457e-06,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8462053269147873,
      "num_tokens": 128781845.0,
      "step": 107130
    },
    {
      "entropy": 1.8408338278532028,
      "epoch": 0.3321246257824373,
      "grad_norm": 5.386666297912598,
      "learning_rate": 4.389794269732274e-06,
      "loss": 0.5242,
      "mean_token_accuracy": 0.8391245663166046,
      "num_tokens": 128794017.0,
      "step": 107140
    },
    {
      "entropy": 1.8361081883311272,
      "epoch": 0.332155624907487,
      "grad_norm": 7.174749374389648,
      "learning_rate": 4.389589419631145e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.8466487154364586,
      "num_tokens": 128805772.0,
      "step": 107150
    },
    {
      "entropy": 1.8699164032936095,
      "epoch": 0.3321866240325367,
      "grad_norm": 9.90065860748291,
      "learning_rate": 4.389384598205379e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8560563385486603,
      "num_tokens": 128817741.0,
      "step": 107160
    },
    {
      "entropy": 1.8497094124555589,
      "epoch": 0.33221762315758635,
      "grad_norm": 8.163619995117188,
      "learning_rate": 4.389179805448286e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8564316019415855,
      "num_tokens": 128829031.0,
      "step": 107170
    },
    {
      "entropy": 1.8336839243769645,
      "epoch": 0.3322486222826361,
      "grad_norm": 8.241865158081055,
      "learning_rate": 4.38897504135318e-06,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8471016883850098,
      "num_tokens": 128841424.0,
      "step": 107180
    },
    {
      "entropy": 1.7827384427189827,
      "epoch": 0.33227962140768574,
      "grad_norm": 4.903016090393066,
      "learning_rate": 4.388770305913374e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8514184907078743,
      "num_tokens": 128855066.0,
      "step": 107190
    },
    {
      "entropy": 1.89354208111763,
      "epoch": 0.33231062053273547,
      "grad_norm": 8.681282043457031,
      "learning_rate": 4.388565599122187e-06,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8418336838483811,
      "num_tokens": 128866035.0,
      "step": 107200
    },
    {
      "entropy": 1.8403149604797364,
      "epoch": 0.33234161965778514,
      "grad_norm": 4.251253604888916,
      "learning_rate": 4.3883609209729375e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8545697897672653,
      "num_tokens": 128878048.0,
      "step": 107210
    },
    {
      "entropy": 1.9954572170972824,
      "epoch": 0.33237261878283486,
      "grad_norm": 8.248795509338379,
      "learning_rate": 4.388156271458946e-06,
      "loss": 0.5474,
      "mean_token_accuracy": 0.8341551095247268,
      "num_tokens": 128888607.0,
      "step": 107220
    },
    {
      "entropy": 1.7177160628139974,
      "epoch": 0.33240361790788453,
      "grad_norm": 7.692461967468262,
      "learning_rate": 4.387951650573538e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8625841051340103,
      "num_tokens": 128903267.0,
      "step": 107230
    },
    {
      "entropy": 1.9298606514930725,
      "epoch": 0.33243461703293425,
      "grad_norm": 7.782931327819824,
      "learning_rate": 4.387747058310038e-06,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8428981378674507,
      "num_tokens": 128914211.0,
      "step": 107240
    },
    {
      "entropy": 1.9102041110396386,
      "epoch": 0.3324656161579839,
      "grad_norm": 9.960110664367676,
      "learning_rate": 4.387542494661775e-06,
      "loss": 0.5518,
      "mean_token_accuracy": 0.8341942340135574,
      "num_tokens": 128925160.0,
      "step": 107250
    },
    {
      "entropy": 1.7972071468830109,
      "epoch": 0.33249661528303365,
      "grad_norm": 8.017036437988281,
      "learning_rate": 4.3873379596220784e-06,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8529744014143944,
      "num_tokens": 128937771.0,
      "step": 107260
    },
    {
      "entropy": 1.8894135251641273,
      "epoch": 0.3325276144080833,
      "grad_norm": 8.887062072753906,
      "learning_rate": 4.3871334531842805e-06,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8535187065601348,
      "num_tokens": 128949419.0,
      "step": 107270
    },
    {
      "entropy": 1.8428677216172218,
      "epoch": 0.33255861353313304,
      "grad_norm": 6.1778645515441895,
      "learning_rate": 4.386928975341716e-06,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8626132607460022,
      "num_tokens": 128960944.0,
      "step": 107280
    },
    {
      "entropy": 1.8553937733173371,
      "epoch": 0.3325896126581827,
      "grad_norm": 8.516434669494629,
      "learning_rate": 4.386724526087722e-06,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8496822208166123,
      "num_tokens": 128972466.0,
      "step": 107290
    },
    {
      "entropy": 1.8534843027591705,
      "epoch": 0.33262061178323243,
      "grad_norm": 8.477677345275879,
      "learning_rate": 4.386520105415637e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8520261511206627,
      "num_tokens": 128984702.0,
      "step": 107300
    },
    {
      "entropy": 1.8577327355742455,
      "epoch": 0.3326516109082821,
      "grad_norm": 8.226881980895996,
      "learning_rate": 4.386315713318802e-06,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8503885567188263,
      "num_tokens": 128996533.0,
      "step": 107310
    },
    {
      "entropy": 1.766034336388111,
      "epoch": 0.3326826100333318,
      "grad_norm": 7.096409320831299,
      "learning_rate": 4.386111349790562e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8526156499981881,
      "num_tokens": 129009556.0,
      "step": 107320
    },
    {
      "entropy": 1.823246991634369,
      "epoch": 0.3327136091583815,
      "grad_norm": 8.441973686218262,
      "learning_rate": 4.385907014824259e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8562351033091545,
      "num_tokens": 129021912.0,
      "step": 107330
    },
    {
      "entropy": 1.8534847319126129,
      "epoch": 0.3327446082834312,
      "grad_norm": 4.784743785858154,
      "learning_rate": 4.385702708413244e-06,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8512447312474251,
      "num_tokens": 129033807.0,
      "step": 107340
    },
    {
      "entropy": 1.8405146718025207,
      "epoch": 0.3327756074084809,
      "grad_norm": 3.7886173725128174,
      "learning_rate": 4.385498430550864e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8435701325535774,
      "num_tokens": 129046120.0,
      "step": 107350
    },
    {
      "entropy": 1.7437313303351403,
      "epoch": 0.3328066065335306,
      "grad_norm": 4.013438701629639,
      "learning_rate": 4.385294181230472e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.8705343827605247,
      "num_tokens": 129060159.0,
      "step": 107360
    },
    {
      "entropy": 1.8268103674054146,
      "epoch": 0.3328376056585803,
      "grad_norm": 10.785335540771484,
      "learning_rate": 4.385089960445422e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.853385554254055,
      "num_tokens": 129072056.0,
      "step": 107370
    },
    {
      "entropy": 1.7955839529633522,
      "epoch": 0.33286860478363,
      "grad_norm": 3.143990993499756,
      "learning_rate": 4.384885768189071e-06,
      "loss": 0.4295,
      "mean_token_accuracy": 0.854556767642498,
      "num_tokens": 129084936.0,
      "step": 107380
    },
    {
      "entropy": 1.9022680431604386,
      "epoch": 0.3328996039086797,
      "grad_norm": 7.960306167602539,
      "learning_rate": 4.384681604454776e-06,
      "loss": 0.5184,
      "mean_token_accuracy": 0.8433623731136322,
      "num_tokens": 129096132.0,
      "step": 107390
    },
    {
      "entropy": 1.8644019782543182,
      "epoch": 0.3329306030337294,
      "grad_norm": 9.219409942626953,
      "learning_rate": 4.384477469235899e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8375217631459236,
      "num_tokens": 129108399.0,
      "step": 107400
    },
    {
      "entropy": 1.8267999082803725,
      "epoch": 0.33296160215877907,
      "grad_norm": 9.918363571166992,
      "learning_rate": 4.384273362525801e-06,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8537021040916443,
      "num_tokens": 129120240.0,
      "step": 107410
    },
    {
      "entropy": 1.8040613248944282,
      "epoch": 0.33299260128382874,
      "grad_norm": 9.216553688049316,
      "learning_rate": 4.384069284317849e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.8598408699035645,
      "num_tokens": 129133042.0,
      "step": 107420
    },
    {
      "entropy": 1.8831330671906472,
      "epoch": 0.33302360040887846,
      "grad_norm": 7.262103080749512,
      "learning_rate": 4.383865234605409e-06,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8518488049507141,
      "num_tokens": 129144475.0,
      "step": 107430
    },
    {
      "entropy": 1.8981991022825242,
      "epoch": 0.33305459953392813,
      "grad_norm": 8.177460670471191,
      "learning_rate": 4.38366121338185e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8512781083583831,
      "num_tokens": 129156011.0,
      "step": 107440
    },
    {
      "entropy": 1.829132416844368,
      "epoch": 0.33308559865897786,
      "grad_norm": 9.38082504272461,
      "learning_rate": 4.383457220640543e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8530718177556992,
      "num_tokens": 129168301.0,
      "step": 107450
    },
    {
      "entropy": 1.76314737200737,
      "epoch": 0.3331165977840275,
      "grad_norm": 8.310257911682129,
      "learning_rate": 4.383253256374863e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8589179947972297,
      "num_tokens": 129181867.0,
      "step": 107460
    },
    {
      "entropy": 1.7426991537213326,
      "epoch": 0.33314759690907725,
      "grad_norm": 4.443640232086182,
      "learning_rate": 4.383049320578185e-06,
      "loss": 0.4003,
      "mean_token_accuracy": 0.8596662864089012,
      "num_tokens": 129194857.0,
      "step": 107470
    },
    {
      "entropy": 1.8887342095375061,
      "epoch": 0.3331785960341269,
      "grad_norm": 9.242331504821777,
      "learning_rate": 4.382845413243886e-06,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8357272610068321,
      "num_tokens": 129206228.0,
      "step": 107480
    },
    {
      "entropy": 1.7917708232998848,
      "epoch": 0.33320959515917664,
      "grad_norm": 7.552341938018799,
      "learning_rate": 4.382641534365348e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8635924890637398,
      "num_tokens": 129219549.0,
      "step": 107490
    },
    {
      "entropy": 1.9377831488847732,
      "epoch": 0.3332405942842263,
      "grad_norm": 8.330388069152832,
      "learning_rate": 4.38243768393595e-06,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8412655621767045,
      "num_tokens": 129230964.0,
      "step": 107500
    },
    {
      "entropy": 1.8014700457453727,
      "epoch": 0.33327159340927603,
      "grad_norm": 9.167746543884277,
      "learning_rate": 4.382233861949079e-06,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8664462238550186,
      "num_tokens": 129242857.0,
      "step": 107510
    },
    {
      "entropy": 1.8836747616529466,
      "epoch": 0.3333025925343257,
      "grad_norm": 8.401406288146973,
      "learning_rate": 4.382030068398122e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.846863467991352,
      "num_tokens": 129254438.0,
      "step": 107520
    },
    {
      "entropy": 1.8405706137418747,
      "epoch": 0.33333359165937543,
      "grad_norm": 3.3866302967071533,
      "learning_rate": 4.381826303276464e-06,
      "loss": 0.4663,
      "mean_token_accuracy": 0.849305622279644,
      "num_tokens": 129266522.0,
      "step": 107530
    },
    {
      "entropy": 1.8931007355451583,
      "epoch": 0.3333645907844251,
      "grad_norm": 9.184843063354492,
      "learning_rate": 4.3816225665775e-06,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8357789561152458,
      "num_tokens": 129278660.0,
      "step": 107540
    },
    {
      "entropy": 1.8619947463274003,
      "epoch": 0.3333955899094748,
      "grad_norm": 8.761635780334473,
      "learning_rate": 4.38141885829462e-06,
      "loss": 0.4635,
      "mean_token_accuracy": 0.8441411420702934,
      "num_tokens": 129290127.0,
      "step": 107550
    },
    {
      "entropy": 1.8187642887234687,
      "epoch": 0.3334265890345245,
      "grad_norm": 7.852782249450684,
      "learning_rate": 4.3812151784212205e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8587213516235351,
      "num_tokens": 129302352.0,
      "step": 107560
    },
    {
      "entropy": 1.8852619290351869,
      "epoch": 0.3334575881595742,
      "grad_norm": 8.888199806213379,
      "learning_rate": 4.3810115269506985e-06,
      "loss": 0.5021,
      "mean_token_accuracy": 0.850848950445652,
      "num_tokens": 129313337.0,
      "step": 107570
    },
    {
      "entropy": 1.8726006522774696,
      "epoch": 0.3334885872846239,
      "grad_norm": 7.686779022216797,
      "learning_rate": 4.380807903876452e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8449873507022858,
      "num_tokens": 129324483.0,
      "step": 107580
    },
    {
      "entropy": 1.857853028178215,
      "epoch": 0.3335195864096736,
      "grad_norm": 9.525311470031738,
      "learning_rate": 4.3806043091918855e-06,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8560262963175773,
      "num_tokens": 129336845.0,
      "step": 107590
    },
    {
      "entropy": 1.8294879227876664,
      "epoch": 0.3335505855347233,
      "grad_norm": 3.713810920715332,
      "learning_rate": 4.3804007428904e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8501683503389359,
      "num_tokens": 129349320.0,
      "step": 107600
    },
    {
      "entropy": 1.7990604281425475,
      "epoch": 0.333581584659773,
      "grad_norm": 8.975895881652832,
      "learning_rate": 4.380197204965402e-06,
      "loss": 0.4444,
      "mean_token_accuracy": 0.850322599709034,
      "num_tokens": 129361696.0,
      "step": 107610
    },
    {
      "entropy": 1.8823053747415543,
      "epoch": 0.33361258378482267,
      "grad_norm": 9.130423545837402,
      "learning_rate": 4.3799936954103e-06,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8529215559363366,
      "num_tokens": 129373047.0,
      "step": 107620
    },
    {
      "entropy": 1.7972995683550834,
      "epoch": 0.3336435829098724,
      "grad_norm": 6.733953952789307,
      "learning_rate": 4.3797902142185034e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8613699153065681,
      "num_tokens": 129386076.0,
      "step": 107630
    },
    {
      "entropy": 1.9032483518123626,
      "epoch": 0.33367458203492206,
      "grad_norm": 8.466991424560547,
      "learning_rate": 4.379586761383426e-06,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8323118224740028,
      "num_tokens": 129397394.0,
      "step": 107640
    },
    {
      "entropy": 1.8645583361387252,
      "epoch": 0.3337055811599718,
      "grad_norm": 9.792939186096191,
      "learning_rate": 4.379383336898479e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8481997266411782,
      "num_tokens": 129408714.0,
      "step": 107650
    },
    {
      "entropy": 1.9275704205036164,
      "epoch": 0.33373658028502146,
      "grad_norm": 8.203200340270996,
      "learning_rate": 4.3791799407570814e-06,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8284709110856057,
      "num_tokens": 129420176.0,
      "step": 107660
    },
    {
      "entropy": 1.9315796703100205,
      "epoch": 0.3337675794100711,
      "grad_norm": 7.3529953956604,
      "learning_rate": 4.378976572952653e-06,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8478155717253685,
      "num_tokens": 129430827.0,
      "step": 107670
    },
    {
      "entropy": 1.7806938499212266,
      "epoch": 0.33379857853512085,
      "grad_norm": 3.7703471183776855,
      "learning_rate": 4.378773233478612e-06,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8491846337914467,
      "num_tokens": 129443966.0,
      "step": 107680
    },
    {
      "entropy": 1.7839350268244742,
      "epoch": 0.3338295776601705,
      "grad_norm": 7.560736179351807,
      "learning_rate": 4.378569922328383e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8620262175798417,
      "num_tokens": 129457050.0,
      "step": 107690
    },
    {
      "entropy": 1.8768269151449204,
      "epoch": 0.33386057678522024,
      "grad_norm": 8.099947929382324,
      "learning_rate": 4.37836663949539e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.844311997294426,
      "num_tokens": 129468839.0,
      "step": 107700
    },
    {
      "entropy": 1.8897486120462417,
      "epoch": 0.3338915759102699,
      "grad_norm": 8.095099449157715,
      "learning_rate": 4.3781633849730605e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8418814584612846,
      "num_tokens": 129480361.0,
      "step": 107710
    },
    {
      "entropy": 1.8843684524297715,
      "epoch": 0.33392257503531964,
      "grad_norm": 5.762594223022461,
      "learning_rate": 4.377960158754823e-06,
      "loss": 0.482,
      "mean_token_accuracy": 0.8446230858564376,
      "num_tokens": 129492583.0,
      "step": 107720
    },
    {
      "entropy": 1.969536703824997,
      "epoch": 0.3339535741603693,
      "grad_norm": 8.858036041259766,
      "learning_rate": 4.377756960834111e-06,
      "loss": 0.509,
      "mean_token_accuracy": 0.844764456152916,
      "num_tokens": 129503430.0,
      "step": 107730
    },
    {
      "entropy": 1.8859269142150878,
      "epoch": 0.33398457328541903,
      "grad_norm": 4.305178165435791,
      "learning_rate": 4.377553791204358e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8533265367150307,
      "num_tokens": 129514786.0,
      "step": 107740
    },
    {
      "entropy": 1.859623844921589,
      "epoch": 0.3340155724104687,
      "grad_norm": 7.1102190017700195,
      "learning_rate": 4.377350649858999e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8454333543777466,
      "num_tokens": 129526854.0,
      "step": 107750
    },
    {
      "entropy": 1.897015392780304,
      "epoch": 0.3340465715355184,
      "grad_norm": 8.836033821105957,
      "learning_rate": 4.377147536791471e-06,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8378755912184715,
      "num_tokens": 129537781.0,
      "step": 107760
    },
    {
      "entropy": 1.8058869987726212,
      "epoch": 0.3340775706605681,
      "grad_norm": 2.6924660205841064,
      "learning_rate": 4.376944451995214e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8562029376626015,
      "num_tokens": 129550764.0,
      "step": 107770
    },
    {
      "entropy": 1.8765144675970078,
      "epoch": 0.3341085697856178,
      "grad_norm": 7.857125759124756,
      "learning_rate": 4.376741395463672e-06,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8540433526039124,
      "num_tokens": 129562412.0,
      "step": 107780
    },
    {
      "entropy": 1.8461247250437736,
      "epoch": 0.3341395689106675,
      "grad_norm": 8.548455238342285,
      "learning_rate": 4.376538367190288e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8496983379125596,
      "num_tokens": 129573651.0,
      "step": 107790
    },
    {
      "entropy": 1.8069075807929038,
      "epoch": 0.3341705680357172,
      "grad_norm": 9.008338928222656,
      "learning_rate": 4.37633536716851e-06,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8606833323836327,
      "num_tokens": 129586574.0,
      "step": 107800
    },
    {
      "entropy": 1.8529318809509276,
      "epoch": 0.3342015671607669,
      "grad_norm": 3.940270185470581,
      "learning_rate": 4.376132395391783e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8578683048486709,
      "num_tokens": 129598430.0,
      "step": 107810
    },
    {
      "entropy": 1.8817769691348076,
      "epoch": 0.3342325662858166,
      "grad_norm": 8.67603588104248,
      "learning_rate": 4.375929451853561e-06,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8386638849973679,
      "num_tokens": 129610580.0,
      "step": 107820
    },
    {
      "entropy": 1.8078914038836955,
      "epoch": 0.33426356541086627,
      "grad_norm": 10.10114574432373,
      "learning_rate": 4.375726536547296e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8479559436440468,
      "num_tokens": 129623248.0,
      "step": 107830
    },
    {
      "entropy": 1.8651557683944702,
      "epoch": 0.334294564535916,
      "grad_norm": 7.359059810638428,
      "learning_rate": 4.375523649466441e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8485989525914193,
      "num_tokens": 129635182.0,
      "step": 107840
    },
    {
      "entropy": 1.744455586373806,
      "epoch": 0.33432556366096566,
      "grad_norm": 5.441449165344238,
      "learning_rate": 4.375320790604457e-06,
      "loss": 0.3794,
      "mean_token_accuracy": 0.8629415735602379,
      "num_tokens": 129649244.0,
      "step": 107850
    },
    {
      "entropy": 1.7221811696887017,
      "epoch": 0.3343565627860154,
      "grad_norm": 3.5538432598114014,
      "learning_rate": 4.375117959954799e-06,
      "loss": 0.353,
      "mean_token_accuracy": 0.8725274935364723,
      "num_tokens": 129662708.0,
      "step": 107860
    },
    {
      "entropy": 1.7094054415822029,
      "epoch": 0.33438756191106506,
      "grad_norm": 11.59929370880127,
      "learning_rate": 4.37491515751093e-06,
      "loss": 0.39,
      "mean_token_accuracy": 0.8701628282666206,
      "num_tokens": 129676262.0,
      "step": 107870
    },
    {
      "entropy": 1.7959278047084808,
      "epoch": 0.3344185610361148,
      "grad_norm": 3.8675200939178467,
      "learning_rate": 4.374712383266314e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8563177272677421,
      "num_tokens": 129689029.0,
      "step": 107880
    },
    {
      "entropy": 1.8378399103879928,
      "epoch": 0.33444956016116445,
      "grad_norm": 7.299082279205322,
      "learning_rate": 4.3745096372144166e-06,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8417981162667274,
      "num_tokens": 129701201.0,
      "step": 107890
    },
    {
      "entropy": 1.8643528178334237,
      "epoch": 0.3344805592862142,
      "grad_norm": 7.960921287536621,
      "learning_rate": 4.374306919348705e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8524877533316613,
      "num_tokens": 129712743.0,
      "step": 107900
    },
    {
      "entropy": 1.900659802556038,
      "epoch": 0.33451155841126384,
      "grad_norm": 7.667184829711914,
      "learning_rate": 4.374104229662648e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8470824211835861,
      "num_tokens": 129723529.0,
      "step": 107910
    },
    {
      "entropy": 1.8972845152020454,
      "epoch": 0.3345425575363135,
      "grad_norm": 7.8113484382629395,
      "learning_rate": 4.3739015681497185e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8441723167896271,
      "num_tokens": 129734887.0,
      "step": 107920
    },
    {
      "entropy": 1.8530030235648156,
      "epoch": 0.33457355666136324,
      "grad_norm": 7.784739971160889,
      "learning_rate": 4.37369893480339e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8602523878216743,
      "num_tokens": 129746166.0,
      "step": 107930
    },
    {
      "entropy": 1.8418569251894952,
      "epoch": 0.3346045557864129,
      "grad_norm": 3.6897940635681152,
      "learning_rate": 4.37349632961714e-06,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8551413804292679,
      "num_tokens": 129758785.0,
      "step": 107940
    },
    {
      "entropy": 1.8382650315761566,
      "epoch": 0.33463555491146263,
      "grad_norm": 6.814600467681885,
      "learning_rate": 4.373293752584445e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8480258256196975,
      "num_tokens": 129771240.0,
      "step": 107950
    },
    {
      "entropy": 1.7906376250088214,
      "epoch": 0.3346665540365123,
      "grad_norm": 8.821784973144531,
      "learning_rate": 4.373091203698785e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.8606226801872253,
      "num_tokens": 129783410.0,
      "step": 107960
    },
    {
      "entropy": 1.850863453745842,
      "epoch": 0.334697553161562,
      "grad_norm": 3.5716798305511475,
      "learning_rate": 4.372888682953645e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8479766875505448,
      "num_tokens": 129794773.0,
      "step": 107970
    },
    {
      "entropy": 1.8356069535017014,
      "epoch": 0.3347285522866117,
      "grad_norm": 8.035232543945312,
      "learning_rate": 4.372686190342508e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8509008720517158,
      "num_tokens": 129807004.0,
      "step": 107980
    },
    {
      "entropy": 1.8803606986999513,
      "epoch": 0.3347595514116614,
      "grad_norm": 7.623292446136475,
      "learning_rate": 4.3724837258588594e-06,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8468181252479553,
      "num_tokens": 129818561.0,
      "step": 107990
    },
    {
      "entropy": 1.8860761135816575,
      "epoch": 0.3347905505367111,
      "grad_norm": 9.617867469787598,
      "learning_rate": 4.37228128949619e-06,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8485965311527253,
      "num_tokens": 129830143.0,
      "step": 108000
    },
    {
      "entropy": 1.912020392715931,
      "epoch": 0.3348215496617608,
      "grad_norm": 8.843141555786133,
      "learning_rate": 4.37207888124799e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8369658708572387,
      "num_tokens": 129841114.0,
      "step": 108010
    },
    {
      "entropy": 1.9048634201288224,
      "epoch": 0.3348525487868105,
      "grad_norm": 6.282487392425537,
      "learning_rate": 4.3718765011077526e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8487017884850502,
      "num_tokens": 129851752.0,
      "step": 108020
    },
    {
      "entropy": 1.9025760471820832,
      "epoch": 0.3348835479118602,
      "grad_norm": 9.245036125183105,
      "learning_rate": 4.371674149068973e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8425028279423714,
      "num_tokens": 129863830.0,
      "step": 108030
    },
    {
      "entropy": 1.863828182220459,
      "epoch": 0.33491454703690987,
      "grad_norm": 3.375025510787964,
      "learning_rate": 4.3714718251251484e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8575912222266198,
      "num_tokens": 129875417.0,
      "step": 108040
    },
    {
      "entropy": 1.749958410859108,
      "epoch": 0.3349455461619596,
      "grad_norm": 3.8491711616516113,
      "learning_rate": 4.371269529269777e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.8608524903655053,
      "num_tokens": 129888923.0,
      "step": 108050
    },
    {
      "entropy": 1.8047884538769723,
      "epoch": 0.33497654528700926,
      "grad_norm": 4.140578746795654,
      "learning_rate": 4.371067261496363e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8508468925952911,
      "num_tokens": 129901648.0,
      "step": 108060
    },
    {
      "entropy": 1.9018609017133712,
      "epoch": 0.335007544412059,
      "grad_norm": 7.736186504364014,
      "learning_rate": 4.3708650217984065e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8536763086915016,
      "num_tokens": 129912303.0,
      "step": 108070
    },
    {
      "entropy": 1.8833556294441223,
      "epoch": 0.33503854353710866,
      "grad_norm": 8.170893669128418,
      "learning_rate": 4.370662810169415e-06,
      "loss": 0.454,
      "mean_token_accuracy": 0.8535693794488907,
      "num_tokens": 129923394.0,
      "step": 108080
    },
    {
      "entropy": 1.7835390165448188,
      "epoch": 0.3350695426621584,
      "grad_norm": 8.293116569519043,
      "learning_rate": 4.370460626602896e-06,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8520323082804679,
      "num_tokens": 129936227.0,
      "step": 108090
    },
    {
      "entropy": 1.8735915690660476,
      "epoch": 0.33510054178720805,
      "grad_norm": 8.832682609558105,
      "learning_rate": 4.37025847109236e-06,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8542598769068718,
      "num_tokens": 129947875.0,
      "step": 108100
    },
    {
      "entropy": 1.8861595943570137,
      "epoch": 0.3351315409122578,
      "grad_norm": 4.6345601081848145,
      "learning_rate": 4.370056343631318e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8438780650496482,
      "num_tokens": 129959918.0,
      "step": 108110
    },
    {
      "entropy": 1.8636915504932403,
      "epoch": 0.33516254003730744,
      "grad_norm": 9.650102615356445,
      "learning_rate": 4.3698542442132845e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8606543466448784,
      "num_tokens": 129970846.0,
      "step": 108120
    },
    {
      "entropy": 1.911002305150032,
      "epoch": 0.33519353916235717,
      "grad_norm": 9.570003509521484,
      "learning_rate": 4.369652172831775e-06,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8438562154769897,
      "num_tokens": 129981852.0,
      "step": 108130
    },
    {
      "entropy": 1.8741035476326942,
      "epoch": 0.33522453828740684,
      "grad_norm": 6.404723644256592,
      "learning_rate": 4.3694501294803084e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8497413903474808,
      "num_tokens": 129993561.0,
      "step": 108140
    },
    {
      "entropy": 1.881354147195816,
      "epoch": 0.3352555374124565,
      "grad_norm": 8.92978286743164,
      "learning_rate": 4.369248114152405e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8553690627217293,
      "num_tokens": 130005553.0,
      "step": 108150
    },
    {
      "entropy": 1.9167394667863846,
      "epoch": 0.33528653653750623,
      "grad_norm": 6.4430341720581055,
      "learning_rate": 4.369046126841588e-06,
      "loss": 0.4929,
      "mean_token_accuracy": 0.848999661207199,
      "num_tokens": 130016296.0,
      "step": 108160
    },
    {
      "entropy": 1.7608971558511257,
      "epoch": 0.3353175356625559,
      "grad_norm": 11.467843055725098,
      "learning_rate": 4.3688441675413805e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8577596142888069,
      "num_tokens": 130029257.0,
      "step": 108170
    },
    {
      "entropy": 1.7902075618505477,
      "epoch": 0.3353485347876056,
      "grad_norm": 7.754476070404053,
      "learning_rate": 4.3686422362453095e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8457227498292923,
      "num_tokens": 130042376.0,
      "step": 108180
    },
    {
      "entropy": 1.80088053047657,
      "epoch": 0.3353795339126553,
      "grad_norm": 9.529817581176758,
      "learning_rate": 4.368440332946905e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8530313909053803,
      "num_tokens": 130054996.0,
      "step": 108190
    },
    {
      "entropy": 1.88881815969944,
      "epoch": 0.335410533037705,
      "grad_norm": 7.529262542724609,
      "learning_rate": 4.368238457639695e-06,
      "loss": 0.4901,
      "mean_token_accuracy": 0.8425343319773674,
      "num_tokens": 130066402.0,
      "step": 108200
    },
    {
      "entropy": 1.9495450556278229,
      "epoch": 0.3354415321627547,
      "grad_norm": 7.71326208114624,
      "learning_rate": 4.368036610317216e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8409694746136666,
      "num_tokens": 130077178.0,
      "step": 108210
    },
    {
      "entropy": 1.898306019604206,
      "epoch": 0.3354725312878044,
      "grad_norm": 7.798426151275635,
      "learning_rate": 4.367834790973002e-06,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8483818009495735,
      "num_tokens": 130088667.0,
      "step": 108220
    },
    {
      "entropy": 1.8723418027162553,
      "epoch": 0.3355035304128541,
      "grad_norm": 4.362555503845215,
      "learning_rate": 4.367632999600588e-06,
      "loss": 0.4794,
      "mean_token_accuracy": 0.8386352211236954,
      "num_tokens": 130101166.0,
      "step": 108230
    },
    {
      "entropy": 1.8666965618729592,
      "epoch": 0.3355345295379038,
      "grad_norm": 6.942992687225342,
      "learning_rate": 4.367431236193516e-06,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8470498412847519,
      "num_tokens": 130113744.0,
      "step": 108240
    },
    {
      "entropy": 1.811617687344551,
      "epoch": 0.33556552866295347,
      "grad_norm": 10.553791999816895,
      "learning_rate": 4.367229500745324e-06,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8510163232684136,
      "num_tokens": 130126707.0,
      "step": 108250
    },
    {
      "entropy": 1.9607309699058533,
      "epoch": 0.3355965277880032,
      "grad_norm": 8.012662887573242,
      "learning_rate": 4.367027793249559e-06,
      "loss": 0.5177,
      "mean_token_accuracy": 0.8463420748710633,
      "num_tokens": 130137777.0,
      "step": 108260
    },
    {
      "entropy": 1.7869203016161919,
      "epoch": 0.33562752691305286,
      "grad_norm": 7.589692115783691,
      "learning_rate": 4.366826113699764e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8630554795265197,
      "num_tokens": 130150461.0,
      "step": 108270
    },
    {
      "entropy": 1.880982668697834,
      "epoch": 0.3356585260381026,
      "grad_norm": 7.124267101287842,
      "learning_rate": 4.366624462089486e-06,
      "loss": 0.5833,
      "mean_token_accuracy": 0.827752648293972,
      "num_tokens": 130162409.0,
      "step": 108280
    },
    {
      "entropy": 1.8591894909739495,
      "epoch": 0.33568952516315226,
      "grad_norm": 7.850539684295654,
      "learning_rate": 4.3664228384122775e-06,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8456223160028458,
      "num_tokens": 130175026.0,
      "step": 108290
    },
    {
      "entropy": 1.8133115381002427,
      "epoch": 0.335720524288202,
      "grad_norm": 3.5843214988708496,
      "learning_rate": 4.366221242661688e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8574982568621635,
      "num_tokens": 130187161.0,
      "step": 108300
    },
    {
      "entropy": 1.8298716515302658,
      "epoch": 0.33575152341325165,
      "grad_norm": 7.22581148147583,
      "learning_rate": 4.366019674831272e-06,
      "loss": 0.4093,
      "mean_token_accuracy": 0.8628581598401069,
      "num_tokens": 130199254.0,
      "step": 108310
    },
    {
      "entropy": 1.8459315612912177,
      "epoch": 0.3357825225383014,
      "grad_norm": 7.195556640625,
      "learning_rate": 4.365818134914586e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.855617669224739,
      "num_tokens": 130211367.0,
      "step": 108320
    },
    {
      "entropy": 1.7838531404733657,
      "epoch": 0.33581352166335104,
      "grad_norm": 3.379918098449707,
      "learning_rate": 4.3656166229051865e-06,
      "loss": 0.3927,
      "mean_token_accuracy": 0.868264564871788,
      "num_tokens": 130223512.0,
      "step": 108330
    },
    {
      "entropy": 1.805635717511177,
      "epoch": 0.33584452078840077,
      "grad_norm": 8.333756446838379,
      "learning_rate": 4.365415138796635e-06,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8466704249382019,
      "num_tokens": 130235508.0,
      "step": 108340
    },
    {
      "entropy": 1.8783694118261338,
      "epoch": 0.33587551991345044,
      "grad_norm": 8.413524627685547,
      "learning_rate": 4.3652136825824914e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8432982712984085,
      "num_tokens": 130247392.0,
      "step": 108350
    },
    {
      "entropy": 1.8464765295386314,
      "epoch": 0.33590651903850016,
      "grad_norm": 10.091898918151855,
      "learning_rate": 4.365012254256323e-06,
      "loss": 0.5256,
      "mean_token_accuracy": 0.836361040174961,
      "num_tokens": 130259254.0,
      "step": 108360
    },
    {
      "entropy": 1.8891173303127289,
      "epoch": 0.33593751816354983,
      "grad_norm": 7.264171123504639,
      "learning_rate": 4.3648108538116935e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.854046767950058,
      "num_tokens": 130270895.0,
      "step": 108370
    },
    {
      "entropy": 1.7990408152341844,
      "epoch": 0.33596851728859956,
      "grad_norm": 7.887672424316406,
      "learning_rate": 4.364609481242173e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8649446800351143,
      "num_tokens": 130283074.0,
      "step": 108380
    },
    {
      "entropy": 1.8251417383551598,
      "epoch": 0.3359995164136492,
      "grad_norm": 8.665481567382812,
      "learning_rate": 4.36440813654133e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8497830405831337,
      "num_tokens": 130295411.0,
      "step": 108390
    },
    {
      "entropy": 1.8189852401614188,
      "epoch": 0.3360305155386989,
      "grad_norm": 7.711446762084961,
      "learning_rate": 4.36420681970274e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8486074671149254,
      "num_tokens": 130307723.0,
      "step": 108400
    },
    {
      "entropy": 1.8464395835995675,
      "epoch": 0.3360615146637486,
      "grad_norm": 8.451082229614258,
      "learning_rate": 4.364005530719975e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.845195434987545,
      "num_tokens": 130320326.0,
      "step": 108410
    },
    {
      "entropy": 1.9397107988595963,
      "epoch": 0.3360925137887983,
      "grad_norm": 8.73201847076416,
      "learning_rate": 4.363804269586612e-06,
      "loss": 0.5236,
      "mean_token_accuracy": 0.8377207443118095,
      "num_tokens": 130330931.0,
      "step": 108420
    },
    {
      "entropy": 1.8559130504727364,
      "epoch": 0.336123512913848,
      "grad_norm": 7.468655109405518,
      "learning_rate": 4.363603036296231e-06,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8583543822169304,
      "num_tokens": 130342645.0,
      "step": 108430
    },
    {
      "entropy": 1.8359305024147035,
      "epoch": 0.3361545120388977,
      "grad_norm": 6.1533522605896,
      "learning_rate": 4.363401830842411e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8503975555300712,
      "num_tokens": 130354736.0,
      "step": 108440
    },
    {
      "entropy": 1.8281599283218384,
      "epoch": 0.3361855111639474,
      "grad_norm": 10.031996726989746,
      "learning_rate": 4.3632006532187375e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8479070514440536,
      "num_tokens": 130366811.0,
      "step": 108450
    },
    {
      "entropy": 1.8181740254163743,
      "epoch": 0.3362165102889971,
      "grad_norm": 9.026259422302246,
      "learning_rate": 4.3629995034187926e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8515908911824226,
      "num_tokens": 130378443.0,
      "step": 108460
    },
    {
      "entropy": 1.653176798671484,
      "epoch": 0.3362475094140468,
      "grad_norm": 3.760293960571289,
      "learning_rate": 4.362798381436164e-06,
      "loss": 0.3436,
      "mean_token_accuracy": 0.8677978113293647,
      "num_tokens": 130393766.0,
      "step": 108470
    },
    {
      "entropy": 1.8992000639438629,
      "epoch": 0.33627850853909647,
      "grad_norm": 8.01069164276123,
      "learning_rate": 4.362597287264443e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8475998684763908,
      "num_tokens": 130405181.0,
      "step": 108480
    },
    {
      "entropy": 1.8159596145153045,
      "epoch": 0.3363095076641462,
      "grad_norm": 4.691624641418457,
      "learning_rate": 4.362396220897218e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8557477623224259,
      "num_tokens": 130417472.0,
      "step": 108490
    },
    {
      "entropy": 1.8598027050495147,
      "epoch": 0.33634050678919586,
      "grad_norm": 4.708345413208008,
      "learning_rate": 4.362195182328085e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8497734665870667,
      "num_tokens": 130429769.0,
      "step": 108500
    },
    {
      "entropy": 1.8282673090696335,
      "epoch": 0.3363715059142456,
      "grad_norm": 7.435179233551025,
      "learning_rate": 4.361994171550637e-06,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8541338846087456,
      "num_tokens": 130441794.0,
      "step": 108510
    },
    {
      "entropy": 1.875466302037239,
      "epoch": 0.33640250503929525,
      "grad_norm": 9.038471221923828,
      "learning_rate": 4.361793188558471e-06,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8360525846481324,
      "num_tokens": 130453505.0,
      "step": 108520
    },
    {
      "entropy": 1.9005172535777093,
      "epoch": 0.336433504164345,
      "grad_norm": 8.704236030578613,
      "learning_rate": 4.361592233345188e-06,
      "loss": 0.5057,
      "mean_token_accuracy": 0.8375848770141602,
      "num_tokens": 130465066.0,
      "step": 108530
    },
    {
      "entropy": 1.811057348549366,
      "epoch": 0.33646450328939465,
      "grad_norm": 10.785872459411621,
      "learning_rate": 4.361391305904388e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8490529522299767,
      "num_tokens": 130477719.0,
      "step": 108540
    },
    {
      "entropy": 1.8520541563630104,
      "epoch": 0.33649550241444437,
      "grad_norm": 7.816697597503662,
      "learning_rate": 4.361190406229676e-06,
      "loss": 0.4668,
      "mean_token_accuracy": 0.849137333035469,
      "num_tokens": 130489472.0,
      "step": 108550
    },
    {
      "entropy": 1.867296999692917,
      "epoch": 0.33652650153949404,
      "grad_norm": 4.023216724395752,
      "learning_rate": 4.360989534314658e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8456032857298851,
      "num_tokens": 130500834.0,
      "step": 108560
    },
    {
      "entropy": 1.9245821714401246,
      "epoch": 0.33655750066454376,
      "grad_norm": 7.590324401855469,
      "learning_rate": 4.36078869015294e-06,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8328892081975937,
      "num_tokens": 130512705.0,
      "step": 108570
    },
    {
      "entropy": 1.8899716123938561,
      "epoch": 0.33658849978959343,
      "grad_norm": 8.06901741027832,
      "learning_rate": 4.360587873738132e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.847601892054081,
      "num_tokens": 130524196.0,
      "step": 108580
    },
    {
      "entropy": 1.8788098588585853,
      "epoch": 0.33661949891464316,
      "grad_norm": 8.891491889953613,
      "learning_rate": 4.360387085063847e-06,
      "loss": 0.5064,
      "mean_token_accuracy": 0.844712820649147,
      "num_tokens": 130536389.0,
      "step": 108590
    },
    {
      "entropy": 1.8847306236624717,
      "epoch": 0.3366504980396928,
      "grad_norm": 4.331157684326172,
      "learning_rate": 4.3601863241237e-06,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8504201784729958,
      "num_tokens": 130548320.0,
      "step": 108600
    },
    {
      "entropy": 1.8718352302908898,
      "epoch": 0.33668149716474255,
      "grad_norm": 7.971691131591797,
      "learning_rate": 4.359985590911303e-06,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8353804409503937,
      "num_tokens": 130560292.0,
      "step": 108610
    },
    {
      "entropy": 1.8931324303150177,
      "epoch": 0.3367124962897922,
      "grad_norm": 7.773447513580322,
      "learning_rate": 4.359784885420276e-06,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8583063185214996,
      "num_tokens": 130570906.0,
      "step": 108620
    },
    {
      "entropy": 1.844712384045124,
      "epoch": 0.33674349541484194,
      "grad_norm": 4.043328285217285,
      "learning_rate": 4.3595842076442394e-06,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8548864290118218,
      "num_tokens": 130583434.0,
      "step": 108630
    },
    {
      "entropy": 1.851549918949604,
      "epoch": 0.3367744945398916,
      "grad_norm": 8.612878799438477,
      "learning_rate": 4.359383557576814e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8524287804961205,
      "num_tokens": 130595397.0,
      "step": 108640
    },
    {
      "entropy": 1.90091462880373,
      "epoch": 0.3368054936649413,
      "grad_norm": 9.851509094238281,
      "learning_rate": 4.359182935211626e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8377526298165321,
      "num_tokens": 130606655.0,
      "step": 108650
    },
    {
      "entropy": 1.8972817674279212,
      "epoch": 0.336836492789991,
      "grad_norm": 8.680237770080566,
      "learning_rate": 4.3589823405423e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8444389298558235,
      "num_tokens": 130617788.0,
      "step": 108660
    },
    {
      "entropy": 1.8298688307404518,
      "epoch": 0.3368674919150407,
      "grad_norm": 9.806267738342285,
      "learning_rate": 4.358781773562466e-06,
      "loss": 0.4908,
      "mean_token_accuracy": 0.8443355247378349,
      "num_tokens": 130629777.0,
      "step": 108670
    },
    {
      "entropy": 1.8319950178265572,
      "epoch": 0.3368984910400904,
      "grad_norm": 7.356091499328613,
      "learning_rate": 4.358581234265752e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8494518354535103,
      "num_tokens": 130642431.0,
      "step": 108680
    },
    {
      "entropy": 1.9162609457969666,
      "epoch": 0.33692949016514007,
      "grad_norm": 4.026700973510742,
      "learning_rate": 4.3583807226457904e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8415546149015427,
      "num_tokens": 130653469.0,
      "step": 108690
    },
    {
      "entropy": 1.846106144785881,
      "epoch": 0.3369604892901898,
      "grad_norm": 8.164050102233887,
      "learning_rate": 4.358180238696217e-06,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8469562456011772,
      "num_tokens": 130665500.0,
      "step": 108700
    },
    {
      "entropy": 1.8313665881752967,
      "epoch": 0.33699148841523946,
      "grad_norm": 9.194726943969727,
      "learning_rate": 4.357979782410669e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8402792170643807,
      "num_tokens": 130677813.0,
      "step": 108710
    },
    {
      "entropy": 1.9255888879299163,
      "epoch": 0.3370224875402892,
      "grad_norm": 10.219868659973145,
      "learning_rate": 4.357779353782783e-06,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8486444607377053,
      "num_tokens": 130689073.0,
      "step": 108720
    },
    {
      "entropy": 1.87244506329298,
      "epoch": 0.33705348666533885,
      "grad_norm": 8.656083106994629,
      "learning_rate": 4.3575789528062e-06,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8389718398451805,
      "num_tokens": 130701943.0,
      "step": 108730
    },
    {
      "entropy": 1.8355872690677644,
      "epoch": 0.3370844857903886,
      "grad_norm": 7.93245792388916,
      "learning_rate": 4.357378579474562e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8615069106221199,
      "num_tokens": 130714006.0,
      "step": 108740
    },
    {
      "entropy": 1.8478736653923988,
      "epoch": 0.33711548491543825,
      "grad_norm": 9.331077575683594,
      "learning_rate": 4.357178233781516e-06,
      "loss": 0.474,
      "mean_token_accuracy": 0.8405976980924607,
      "num_tokens": 130726086.0,
      "step": 108750
    },
    {
      "entropy": 1.8251071408391,
      "epoch": 0.33714648404048797,
      "grad_norm": 9.389089584350586,
      "learning_rate": 4.356977915720707e-06,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8374730452895165,
      "num_tokens": 130738640.0,
      "step": 108760
    },
    {
      "entropy": 1.8872618451714516,
      "epoch": 0.33717748316553764,
      "grad_norm": 6.961200714111328,
      "learning_rate": 4.356777625285783e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8394622817635536,
      "num_tokens": 130750629.0,
      "step": 108770
    },
    {
      "entropy": 1.8409572780132293,
      "epoch": 0.33720848229058736,
      "grad_norm": 4.427051067352295,
      "learning_rate": 4.356577362470397e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8474222555756569,
      "num_tokens": 130763704.0,
      "step": 108780
    },
    {
      "entropy": 1.8041258588433267,
      "epoch": 0.33723948141563703,
      "grad_norm": 2.4223392009735107,
      "learning_rate": 4.3563771272681995e-06,
      "loss": 0.494,
      "mean_token_accuracy": 0.8473551660776139,
      "num_tokens": 130776752.0,
      "step": 108790
    },
    {
      "entropy": 1.8579876363277434,
      "epoch": 0.33727048054068676,
      "grad_norm": 4.547750949859619,
      "learning_rate": 4.356176919672846e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8426438301801682,
      "num_tokens": 130789486.0,
      "step": 108800
    },
    {
      "entropy": 1.756131686270237,
      "epoch": 0.3373014796657364,
      "grad_norm": 4.558795928955078,
      "learning_rate": 4.355976739677995e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8526749208569526,
      "num_tokens": 130803159.0,
      "step": 108810
    },
    {
      "entropy": 1.736269749701023,
      "epoch": 0.33733247879078615,
      "grad_norm": 3.4384872913360596,
      "learning_rate": 4.355776587277302e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8524326965212822,
      "num_tokens": 130817219.0,
      "step": 108820
    },
    {
      "entropy": 1.7983601108193397,
      "epoch": 0.3373634779158358,
      "grad_norm": 7.676399230957031,
      "learning_rate": 4.355576462464431e-06,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8553432121872901,
      "num_tokens": 130829997.0,
      "step": 108830
    },
    {
      "entropy": 1.849858644604683,
      "epoch": 0.33739447704088554,
      "grad_norm": 6.735903739929199,
      "learning_rate": 4.3553763652330446e-06,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8419752418994904,
      "num_tokens": 130842661.0,
      "step": 108840
    },
    {
      "entropy": 1.9195626258850098,
      "epoch": 0.3374254761659352,
      "grad_norm": 8.325475692749023,
      "learning_rate": 4.355176295576807e-06,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8380102962255478,
      "num_tokens": 130854094.0,
      "step": 108850
    },
    {
      "entropy": 1.6827729061245917,
      "epoch": 0.33745647529098494,
      "grad_norm": 3.7391862869262695,
      "learning_rate": 4.354976253489386e-06,
      "loss": 0.345,
      "mean_token_accuracy": 0.8674192950129509,
      "num_tokens": 130868296.0,
      "step": 108860
    },
    {
      "entropy": 1.8783330723643303,
      "epoch": 0.3374874744160346,
      "grad_norm": 7.508564472198486,
      "learning_rate": 4.35477623896445e-06,
      "loss": 0.4763,
      "mean_token_accuracy": 0.8525745928287506,
      "num_tokens": 130879523.0,
      "step": 108870
    },
    {
      "entropy": 1.831697914004326,
      "epoch": 0.33751847354108433,
      "grad_norm": 7.616162300109863,
      "learning_rate": 4.35457625199567e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8564835578203202,
      "num_tokens": 130891707.0,
      "step": 108880
    },
    {
      "entropy": 1.780026839673519,
      "epoch": 0.337549472666134,
      "grad_norm": 6.331145763397217,
      "learning_rate": 4.354376292576721e-06,
      "loss": 0.3985,
      "mean_token_accuracy": 0.8688737958669662,
      "num_tokens": 130904551.0,
      "step": 108890
    },
    {
      "entropy": 1.8686219871044158,
      "epoch": 0.33758047179118367,
      "grad_norm": 8.733838081359863,
      "learning_rate": 4.354176360701276e-06,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8480039656162262,
      "num_tokens": 130916170.0,
      "step": 108900
    },
    {
      "entropy": 1.8567358165979386,
      "epoch": 0.3376114709162334,
      "grad_norm": 9.616422653198242,
      "learning_rate": 4.353976456363014e-06,
      "loss": 0.4977,
      "mean_token_accuracy": 0.83982073366642,
      "num_tokens": 130927254.0,
      "step": 108910
    },
    {
      "entropy": 1.7685384809970857,
      "epoch": 0.33764247004128306,
      "grad_norm": 9.59286117553711,
      "learning_rate": 4.353776579555613e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8706537902355194,
      "num_tokens": 130939683.0,
      "step": 108920
    },
    {
      "entropy": 1.864783415198326,
      "epoch": 0.3376734691663328,
      "grad_norm": 8.993294715881348,
      "learning_rate": 4.353576730272754e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.847262118756771,
      "num_tokens": 130951156.0,
      "step": 108930
    },
    {
      "entropy": 1.850709395110607,
      "epoch": 0.33770446829138245,
      "grad_norm": 8.352401733398438,
      "learning_rate": 4.3533769085081226e-06,
      "loss": 0.5349,
      "mean_token_accuracy": 0.8358264163136482,
      "num_tokens": 130962743.0,
      "step": 108940
    },
    {
      "entropy": 1.8151063159108163,
      "epoch": 0.3377354674164322,
      "grad_norm": 7.70052433013916,
      "learning_rate": 4.353177114255402e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8555310294032097,
      "num_tokens": 130974981.0,
      "step": 108950
    },
    {
      "entropy": 1.813981081545353,
      "epoch": 0.33776646654148185,
      "grad_norm": 7.85250997543335,
      "learning_rate": 4.352977347508281e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8673159837722778,
      "num_tokens": 130985978.0,
      "step": 108960
    },
    {
      "entropy": 1.8001227349042892,
      "epoch": 0.33779746566653157,
      "grad_norm": 7.5097270011901855,
      "learning_rate": 4.352777608260448e-06,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8529728427529335,
      "num_tokens": 130998505.0,
      "step": 108970
    },
    {
      "entropy": 1.812097106873989,
      "epoch": 0.33782846479158124,
      "grad_norm": 8.388304710388184,
      "learning_rate": 4.352577896505595e-06,
      "loss": 0.45,
      "mean_token_accuracy": 0.8446347638964653,
      "num_tokens": 131010534.0,
      "step": 108980
    },
    {
      "entropy": 1.8360211491584777,
      "epoch": 0.33785946391663096,
      "grad_norm": 9.120504379272461,
      "learning_rate": 4.352378212237415e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8418739259243011,
      "num_tokens": 131022899.0,
      "step": 108990
    },
    {
      "entropy": 1.9238962471485137,
      "epoch": 0.33789046304168063,
      "grad_norm": 8.026063919067383,
      "learning_rate": 4.352178555449604e-06,
      "loss": 0.5547,
      "mean_token_accuracy": 0.8331388279795646,
      "num_tokens": 131034537.0,
      "step": 109000
    },
    {
      "entropy": 1.8263150677084923,
      "epoch": 0.33792146216673036,
      "grad_norm": 8.960397720336914,
      "learning_rate": 4.351978926135859e-06,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8622379198670387,
      "num_tokens": 131046023.0,
      "step": 109010
    },
    {
      "entropy": 1.8406116291880608,
      "epoch": 0.33795246129178,
      "grad_norm": 8.759995460510254,
      "learning_rate": 4.351779324289881e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8474781602621079,
      "num_tokens": 131057713.0,
      "step": 109020
    },
    {
      "entropy": 1.8527844935655593,
      "epoch": 0.33798346041682975,
      "grad_norm": 8.07425594329834,
      "learning_rate": 4.35157974990537e-06,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8427453622221946,
      "num_tokens": 131069288.0,
      "step": 109030
    },
    {
      "entropy": 1.8225320369005202,
      "epoch": 0.3380144595418794,
      "grad_norm": 8.338834762573242,
      "learning_rate": 4.35138020297603e-06,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8498498395085334,
      "num_tokens": 131082196.0,
      "step": 109040
    },
    {
      "entropy": 1.7835667803883553,
      "epoch": 0.33804545866692914,
      "grad_norm": 4.169846534729004,
      "learning_rate": 4.351180683495567e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8465461641550064,
      "num_tokens": 131094762.0,
      "step": 109050
    },
    {
      "entropy": 1.8385676950216294,
      "epoch": 0.3380764577919788,
      "grad_norm": 7.802093505859375,
      "learning_rate": 4.350981191457688e-06,
      "loss": 0.5106,
      "mean_token_accuracy": 0.8392044931650162,
      "num_tokens": 131107519.0,
      "step": 109060
    },
    {
      "entropy": 1.9263346076011658,
      "epoch": 0.33810745691702854,
      "grad_norm": 8.671483039855957,
      "learning_rate": 4.350781726856105e-06,
      "loss": 0.5573,
      "mean_token_accuracy": 0.8391161412000656,
      "num_tokens": 131118004.0,
      "step": 109070
    },
    {
      "entropy": 1.7481406077742576,
      "epoch": 0.3381384560420782,
      "grad_norm": 4.902276039123535,
      "learning_rate": 4.3505822896845245e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8569299519062042,
      "num_tokens": 131132148.0,
      "step": 109080
    },
    {
      "entropy": 1.8396126255393028,
      "epoch": 0.33816945516712793,
      "grad_norm": 9.361617088317871,
      "learning_rate": 4.350382879936665e-06,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8506528735160828,
      "num_tokens": 131143760.0,
      "step": 109090
    },
    {
      "entropy": 1.7284615464508533,
      "epoch": 0.3382004542921776,
      "grad_norm": 3.8647422790527344,
      "learning_rate": 4.350183497606242e-06,
      "loss": 0.3681,
      "mean_token_accuracy": 0.8593399003148079,
      "num_tokens": 131158374.0,
      "step": 109100
    },
    {
      "entropy": 1.84774319678545,
      "epoch": 0.3382314534172273,
      "grad_norm": 7.73523473739624,
      "learning_rate": 4.34998414268697e-06,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8502051830291748,
      "num_tokens": 131170450.0,
      "step": 109110
    },
    {
      "entropy": 1.8346653819084167,
      "epoch": 0.338262452542277,
      "grad_norm": 9.322132110595703,
      "learning_rate": 4.349784815172573e-06,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8492701262235641,
      "num_tokens": 131182861.0,
      "step": 109120
    },
    {
      "entropy": 1.899721224606037,
      "epoch": 0.3382934516673267,
      "grad_norm": 10.050145149230957,
      "learning_rate": 4.349585515056768e-06,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8410122975707054,
      "num_tokens": 131194694.0,
      "step": 109130
    },
    {
      "entropy": 1.8425235763192176,
      "epoch": 0.3383244507923764,
      "grad_norm": 12.264300346374512,
      "learning_rate": 4.349386242333283e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8546083837747573,
      "num_tokens": 131205712.0,
      "step": 109140
    },
    {
      "entropy": 1.896941477060318,
      "epoch": 0.33835544991742605,
      "grad_norm": 8.371997833251953,
      "learning_rate": 4.349186996995841e-06,
      "loss": 0.5291,
      "mean_token_accuracy": 0.8351884350180626,
      "num_tokens": 131217095.0,
      "step": 109150
    },
    {
      "entropy": 1.863653865456581,
      "epoch": 0.3383864490424758,
      "grad_norm": 8.408138275146484,
      "learning_rate": 4.3489877790381716e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8465486764907837,
      "num_tokens": 131228887.0,
      "step": 109160
    },
    {
      "entropy": 1.7314985610544682,
      "epoch": 0.33841744816752545,
      "grad_norm": 3.9695277214050293,
      "learning_rate": 4.348788588454003e-06,
      "loss": 0.3653,
      "mean_token_accuracy": 0.8568756997585296,
      "num_tokens": 131242417.0,
      "step": 109170
    },
    {
      "entropy": 1.7864164792001247,
      "epoch": 0.33844844729257517,
      "grad_norm": 3.7821221351623535,
      "learning_rate": 4.348589425237069e-06,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8489362448453903,
      "num_tokens": 131255401.0,
      "step": 109180
    },
    {
      "entropy": 1.8366502463817596,
      "epoch": 0.33847944641762484,
      "grad_norm": 7.946313381195068,
      "learning_rate": 4.348390289381101e-06,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8512516096234322,
      "num_tokens": 131267388.0,
      "step": 109190
    },
    {
      "entropy": 1.800801232457161,
      "epoch": 0.33851044554267457,
      "grad_norm": 8.96656322479248,
      "learning_rate": 4.348191180879837e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8497388571500778,
      "num_tokens": 131280203.0,
      "step": 109200
    },
    {
      "entropy": 1.8128297343850135,
      "epoch": 0.33854144466772423,
      "grad_norm": 7.974830150604248,
      "learning_rate": 4.347992099727013e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8562871024012566,
      "num_tokens": 131292537.0,
      "step": 109210
    },
    {
      "entropy": 1.8276149809360505,
      "epoch": 0.33857244379277396,
      "grad_norm": 7.294045448303223,
      "learning_rate": 4.347793045916371e-06,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8447601184248924,
      "num_tokens": 131304251.0,
      "step": 109220
    },
    {
      "entropy": 1.9087721765041352,
      "epoch": 0.3386034429178236,
      "grad_norm": 9.217912673950195,
      "learning_rate": 4.3475940194416516e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8431843280792236,
      "num_tokens": 131315640.0,
      "step": 109230
    },
    {
      "entropy": 1.8142705112695694,
      "epoch": 0.33863444204287335,
      "grad_norm": 8.342325210571289,
      "learning_rate": 4.347395020296598e-06,
      "loss": 0.4458,
      "mean_token_accuracy": 0.84480060338974,
      "num_tokens": 131328126.0,
      "step": 109240
    },
    {
      "entropy": 1.8766521513462067,
      "epoch": 0.338665441167923,
      "grad_norm": 7.5658745765686035,
      "learning_rate": 4.347196048474958e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8464653968811036,
      "num_tokens": 131339951.0,
      "step": 109250
    },
    {
      "entropy": 1.7888486787676812,
      "epoch": 0.33869644029297274,
      "grad_norm": 8.956585884094238,
      "learning_rate": 4.346997103970477e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8637697413563729,
      "num_tokens": 131352826.0,
      "step": 109260
    },
    {
      "entropy": 1.9030503317713738,
      "epoch": 0.3387274394180224,
      "grad_norm": 9.701215744018555,
      "learning_rate": 4.3467981867769075e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8389164701104164,
      "num_tokens": 131364743.0,
      "step": 109270
    },
    {
      "entropy": 1.904677079617977,
      "epoch": 0.33875843854307214,
      "grad_norm": 7.389717102050781,
      "learning_rate": 4.346599296888e-06,
      "loss": 0.472,
      "mean_token_accuracy": 0.8455563247203827,
      "num_tokens": 131376346.0,
      "step": 109280
    },
    {
      "entropy": 1.853348232060671,
      "epoch": 0.3387894376681218,
      "grad_norm": 8.647706031799316,
      "learning_rate": 4.346400434297507e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8570475921034812,
      "num_tokens": 131388648.0,
      "step": 109290
    },
    {
      "entropy": 1.854999852180481,
      "epoch": 0.33882043679317153,
      "grad_norm": 5.456407070159912,
      "learning_rate": 4.346201598999186e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.845246997475624,
      "num_tokens": 131401406.0,
      "step": 109300
    },
    {
      "entropy": 1.9005474954843522,
      "epoch": 0.3388514359182212,
      "grad_norm": 7.554257869720459,
      "learning_rate": 4.346002790986796e-06,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8443148121237755,
      "num_tokens": 131412576.0,
      "step": 109310
    },
    {
      "entropy": 1.957038275897503,
      "epoch": 0.3388824350432709,
      "grad_norm": 8.818881034851074,
      "learning_rate": 4.3458040102540945e-06,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8357266262173653,
      "num_tokens": 131424034.0,
      "step": 109320
    },
    {
      "entropy": 1.7917650774121285,
      "epoch": 0.3389134341683206,
      "grad_norm": 7.657764911651611,
      "learning_rate": 4.345605256794846e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8518206834793091,
      "num_tokens": 131437119.0,
      "step": 109330
    },
    {
      "entropy": 1.9100136697292327,
      "epoch": 0.3389444332933703,
      "grad_norm": 7.320883274078369,
      "learning_rate": 4.345406530602812e-06,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8504310637712479,
      "num_tokens": 131448308.0,
      "step": 109340
    },
    {
      "entropy": 1.8857989877462387,
      "epoch": 0.33897543241842,
      "grad_norm": 8.363656044006348,
      "learning_rate": 4.3452078316717594e-06,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8354710638523102,
      "num_tokens": 131460352.0,
      "step": 109350
    },
    {
      "entropy": 1.829300546646118,
      "epoch": 0.3390064315434697,
      "grad_norm": 3.853241443634033,
      "learning_rate": 4.345009159995455e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8451276645064354,
      "num_tokens": 131472410.0,
      "step": 109360
    },
    {
      "entropy": 1.9137260258197784,
      "epoch": 0.3390374306685194,
      "grad_norm": 8.751523971557617,
      "learning_rate": 4.344810515567671e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8443731889128685,
      "num_tokens": 131484493.0,
      "step": 109370
    },
    {
      "entropy": 1.839902514219284,
      "epoch": 0.3390684297935691,
      "grad_norm": 8.783478736877441,
      "learning_rate": 4.344611898382176e-06,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8553899496793747,
      "num_tokens": 131496822.0,
      "step": 109380
    },
    {
      "entropy": 1.896230974793434,
      "epoch": 0.3390994289186188,
      "grad_norm": 8.84762954711914,
      "learning_rate": 4.3444133084327464e-06,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8462594136595726,
      "num_tokens": 131508185.0,
      "step": 109390
    },
    {
      "entropy": 1.889091356098652,
      "epoch": 0.33913042804366844,
      "grad_norm": 9.512887954711914,
      "learning_rate": 4.344214745713158e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.8467561930418015,
      "num_tokens": 131519594.0,
      "step": 109400
    },
    {
      "entropy": 1.9061472669243813,
      "epoch": 0.33916142716871817,
      "grad_norm": 9.376303672790527,
      "learning_rate": 4.344016210217188e-06,
      "loss": 0.5523,
      "mean_token_accuracy": 0.8321489855647087,
      "num_tokens": 131531669.0,
      "step": 109410
    },
    {
      "entropy": 1.8296129301190376,
      "epoch": 0.33919242629376783,
      "grad_norm": 3.8118650913238525,
      "learning_rate": 4.343817701938615e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8557925269007682,
      "num_tokens": 131544918.0,
      "step": 109420
    },
    {
      "entropy": 1.8758042737841607,
      "epoch": 0.33922342541881756,
      "grad_norm": 8.741567611694336,
      "learning_rate": 4.3436192208712234e-06,
      "loss": 0.4798,
      "mean_token_accuracy": 0.849548727273941,
      "num_tokens": 131557550.0,
      "step": 109430
    },
    {
      "entropy": 1.8672025337815286,
      "epoch": 0.33925442454386723,
      "grad_norm": 7.4333624839782715,
      "learning_rate": 4.343420767008793e-06,
      "loss": 0.489,
      "mean_token_accuracy": 0.8443247124552726,
      "num_tokens": 131569386.0,
      "step": 109440
    },
    {
      "entropy": 1.9236103370785713,
      "epoch": 0.33928542366891695,
      "grad_norm": 9.117382049560547,
      "learning_rate": 4.343222340345114e-06,
      "loss": 0.5212,
      "mean_token_accuracy": 0.8418242424726486,
      "num_tokens": 131580901.0,
      "step": 109450
    },
    {
      "entropy": 1.7963170796632766,
      "epoch": 0.3393164227939666,
      "grad_norm": 7.715478897094727,
      "learning_rate": 4.343023940873973e-06,
      "loss": 0.3927,
      "mean_token_accuracy": 0.8599506393074989,
      "num_tokens": 131594068.0,
      "step": 109460
    },
    {
      "entropy": 1.7216453760862351,
      "epoch": 0.33934742191901635,
      "grad_norm": 7.901419162750244,
      "learning_rate": 4.342825568589158e-06,
      "loss": 0.3747,
      "mean_token_accuracy": 0.8583090871572494,
      "num_tokens": 131608444.0,
      "step": 109470
    },
    {
      "entropy": 1.9911436915397644,
      "epoch": 0.339378421044066,
      "grad_norm": 8.45919132232666,
      "learning_rate": 4.342627223484461e-06,
      "loss": 0.5489,
      "mean_token_accuracy": 0.8311255067586899,
      "num_tokens": 131619482.0,
      "step": 109480
    },
    {
      "entropy": 1.8852492436766624,
      "epoch": 0.33940942016911574,
      "grad_norm": 7.630876064300537,
      "learning_rate": 4.342428905553678e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.8610611036419868,
      "num_tokens": 131632114.0,
      "step": 109490
    },
    {
      "entropy": 1.7850932255387306,
      "epoch": 0.3394404192941654,
      "grad_norm": 7.689153671264648,
      "learning_rate": 4.342230614790603e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8630185857415199,
      "num_tokens": 131645342.0,
      "step": 109500
    },
    {
      "entropy": 1.8250619187951087,
      "epoch": 0.33947141841921513,
      "grad_norm": 9.863249778747559,
      "learning_rate": 4.342032351189033e-06,
      "loss": 0.4457,
      "mean_token_accuracy": 0.8580538392066955,
      "num_tokens": 131657561.0,
      "step": 109510
    },
    {
      "entropy": 1.8692012056708336,
      "epoch": 0.3395024175442648,
      "grad_norm": 7.83109712600708,
      "learning_rate": 4.341834114742769e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8402553409337997,
      "num_tokens": 131669689.0,
      "step": 109520
    },
    {
      "entropy": 1.8381701841950417,
      "epoch": 0.3395334166693145,
      "grad_norm": 9.981649398803711,
      "learning_rate": 4.341635905445612e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8586729139089584,
      "num_tokens": 131681608.0,
      "step": 109530
    },
    {
      "entropy": 1.879381312429905,
      "epoch": 0.3395644157943642,
      "grad_norm": 3.283308506011963,
      "learning_rate": 4.341437723291367e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.846294941008091,
      "num_tokens": 131693388.0,
      "step": 109540
    },
    {
      "entropy": 1.812347574532032,
      "epoch": 0.3395954149194139,
      "grad_norm": 8.128570556640625,
      "learning_rate": 4.341239568273838e-06,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8501547113060951,
      "num_tokens": 131706396.0,
      "step": 109550
    },
    {
      "entropy": 1.8860585197806359,
      "epoch": 0.3396264140444636,
      "grad_norm": 8.69900894165039,
      "learning_rate": 4.341041440386833e-06,
      "loss": 0.4872,
      "mean_token_accuracy": 0.844724814593792,
      "num_tokens": 131718236.0,
      "step": 109560
    },
    {
      "entropy": 1.7079974353313445,
      "epoch": 0.3396574131695133,
      "grad_norm": 2.750688314437866,
      "learning_rate": 4.34084333962416e-06,
      "loss": 0.3525,
      "mean_token_accuracy": 0.8660884618759155,
      "num_tokens": 131732740.0,
      "step": 109570
    },
    {
      "entropy": 1.9225364074110984,
      "epoch": 0.339688412294563,
      "grad_norm": 8.820713996887207,
      "learning_rate": 4.340645265979634e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8516004055738449,
      "num_tokens": 131743968.0,
      "step": 109580
    },
    {
      "entropy": 1.9468763768672943,
      "epoch": 0.3397194114196127,
      "grad_norm": 9.133736610412598,
      "learning_rate": 4.340447219447068e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8412905469536781,
      "num_tokens": 131754741.0,
      "step": 109590
    },
    {
      "entropy": 1.99237479865551,
      "epoch": 0.3397504105446624,
      "grad_norm": 8.745469093322754,
      "learning_rate": 4.340249200020274e-06,
      "loss": 0.5682,
      "mean_token_accuracy": 0.8262647077441215,
      "num_tokens": 131766026.0,
      "step": 109600
    },
    {
      "entropy": 1.9354695051908493,
      "epoch": 0.3397814096697121,
      "grad_norm": 7.548391819000244,
      "learning_rate": 4.340051207693073e-06,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8456323340535163,
      "num_tokens": 131776690.0,
      "step": 109610
    },
    {
      "entropy": 1.8958056300878525,
      "epoch": 0.33981240879476177,
      "grad_norm": 4.0965256690979,
      "learning_rate": 4.339853242459283e-06,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8384672284126282,
      "num_tokens": 131788190.0,
      "step": 109620
    },
    {
      "entropy": 1.8747544452548026,
      "epoch": 0.3398434079198115,
      "grad_norm": 9.14088249206543,
      "learning_rate": 4.339655304312725e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.8471368581056595,
      "num_tokens": 131799932.0,
      "step": 109630
    },
    {
      "entropy": 1.9145475149154663,
      "epoch": 0.33987440704486116,
      "grad_norm": 8.718486785888672,
      "learning_rate": 4.339457393247224e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8452249899506569,
      "num_tokens": 131810626.0,
      "step": 109640
    },
    {
      "entropy": 1.9000482648611068,
      "epoch": 0.33990540616991083,
      "grad_norm": 5.583610534667969,
      "learning_rate": 4.339259509256604e-06,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8375597059726715,
      "num_tokens": 131821937.0,
      "step": 109650
    },
    {
      "entropy": 1.7997367039322854,
      "epoch": 0.33993640529496055,
      "grad_norm": 7.190251350402832,
      "learning_rate": 4.339061652334693e-06,
      "loss": 0.412,
      "mean_token_accuracy": 0.8610153466463089,
      "num_tokens": 131834731.0,
      "step": 109660
    },
    {
      "entropy": 1.8661419078707695,
      "epoch": 0.3399674044200102,
      "grad_norm": 7.567733287811279,
      "learning_rate": 4.33886382247532e-06,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8479538485407829,
      "num_tokens": 131846172.0,
      "step": 109670
    },
    {
      "entropy": 1.8096156984567642,
      "epoch": 0.33999840354505995,
      "grad_norm": 8.170653343200684,
      "learning_rate": 4.338666019672315e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8594262331724167,
      "num_tokens": 131859132.0,
      "step": 109680
    },
    {
      "entropy": 1.8735254988074304,
      "epoch": 0.3400294026701096,
      "grad_norm": 4.316888809204102,
      "learning_rate": 4.3384682439195146e-06,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8408343940973282,
      "num_tokens": 131870647.0,
      "step": 109690
    },
    {
      "entropy": 1.8006722897291183,
      "epoch": 0.34006040179515934,
      "grad_norm": 3.814378499984741,
      "learning_rate": 4.338270495210751e-06,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8559133544564247,
      "num_tokens": 131882827.0,
      "step": 109700
    },
    {
      "entropy": 1.8134817466139794,
      "epoch": 0.340091400920209,
      "grad_norm": 4.081186771392822,
      "learning_rate": 4.338072773539862e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8579586237668991,
      "num_tokens": 131894503.0,
      "step": 109710
    },
    {
      "entropy": 1.8244572654366493,
      "epoch": 0.34012240004525873,
      "grad_norm": 3.6130762100219727,
      "learning_rate": 4.337875078900688e-06,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8462868794798851,
      "num_tokens": 131907275.0,
      "step": 109720
    },
    {
      "entropy": 1.911041909456253,
      "epoch": 0.3401533991703084,
      "grad_norm": 10.027078628540039,
      "learning_rate": 4.337677411287069e-06,
      "loss": 0.4936,
      "mean_token_accuracy": 0.855143415927887,
      "num_tokens": 131918604.0,
      "step": 109730
    },
    {
      "entropy": 1.7550349622964858,
      "epoch": 0.3401843982953581,
      "grad_norm": 4.361454010009766,
      "learning_rate": 4.337479770692849e-06,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8564995616674423,
      "num_tokens": 131931986.0,
      "step": 109740
    },
    {
      "entropy": 1.8325242474675179,
      "epoch": 0.3402153974204078,
      "grad_norm": 7.0168890953063965,
      "learning_rate": 4.337282157111871e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.856087788939476,
      "num_tokens": 131944417.0,
      "step": 109750
    },
    {
      "entropy": 1.8343531802296638,
      "epoch": 0.3402463965454575,
      "grad_norm": 8.980559349060059,
      "learning_rate": 4.337084570537985e-06,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8543761387467385,
      "num_tokens": 131956829.0,
      "step": 109760
    },
    {
      "entropy": 1.9214718401432038,
      "epoch": 0.3402773956705072,
      "grad_norm": 9.840754508972168,
      "learning_rate": 4.336887010965037e-06,
      "loss": 0.5471,
      "mean_token_accuracy": 0.8250068485736847,
      "num_tokens": 131967623.0,
      "step": 109770
    },
    {
      "entropy": 1.7958107739686966,
      "epoch": 0.3403083947955569,
      "grad_norm": 6.940483570098877,
      "learning_rate": 4.336689478386879e-06,
      "loss": 0.3996,
      "mean_token_accuracy": 0.8612048402428627,
      "num_tokens": 131980237.0,
      "step": 109780
    },
    {
      "entropy": 1.8743787840008737,
      "epoch": 0.3403393939206066,
      "grad_norm": 8.042901039123535,
      "learning_rate": 4.3364919727973655e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8504285082221031,
      "num_tokens": 131991550.0,
      "step": 109790
    },
    {
      "entropy": 1.86217290610075,
      "epoch": 0.3403703930456563,
      "grad_norm": 8.980916023254395,
      "learning_rate": 4.33629449419035e-06,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8460476860404015,
      "num_tokens": 132003560.0,
      "step": 109800
    },
    {
      "entropy": 1.7462153866887093,
      "epoch": 0.340401392170706,
      "grad_norm": 7.179854393005371,
      "learning_rate": 4.336097042559689e-06,
      "loss": 0.3725,
      "mean_token_accuracy": 0.8604865953326225,
      "num_tokens": 132016859.0,
      "step": 109810
    },
    {
      "entropy": 1.8352238804101944,
      "epoch": 0.3404323912957557,
      "grad_norm": 8.293797492980957,
      "learning_rate": 4.3358996178992416e-06,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8555570766329765,
      "num_tokens": 132028207.0,
      "step": 109820
    },
    {
      "entropy": 1.7612206950783729,
      "epoch": 0.34046339042080537,
      "grad_norm": 8.661210060119629,
      "learning_rate": 4.3357022202028685e-06,
      "loss": 0.452,
      "mean_token_accuracy": 0.8550508677959442,
      "num_tokens": 132041782.0,
      "step": 109830
    },
    {
      "entropy": 1.8841624334454536,
      "epoch": 0.3404943895458551,
      "grad_norm": 8.303897857666016,
      "learning_rate": 4.335504849464432e-06,
      "loss": 0.543,
      "mean_token_accuracy": 0.8422963932156563,
      "num_tokens": 132053535.0,
      "step": 109840
    },
    {
      "entropy": 1.7789085522294044,
      "epoch": 0.34052538867090476,
      "grad_norm": 3.7404472827911377,
      "learning_rate": 4.335307505677798e-06,
      "loss": 0.407,
      "mean_token_accuracy": 0.8541888251900673,
      "num_tokens": 132066748.0,
      "step": 109850
    },
    {
      "entropy": 1.8748383790254592,
      "epoch": 0.3405563877959545,
      "grad_norm": 7.891180515289307,
      "learning_rate": 4.335110188836832e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8546398028731346,
      "num_tokens": 132078371.0,
      "step": 109860
    },
    {
      "entropy": 1.8530076041817665,
      "epoch": 0.34058738692100415,
      "grad_norm": 8.61839771270752,
      "learning_rate": 4.334912898935402e-06,
      "loss": 0.4438,
      "mean_token_accuracy": 0.845193374156952,
      "num_tokens": 132090696.0,
      "step": 109870
    },
    {
      "entropy": 1.9376697182655334,
      "epoch": 0.3406183860460538,
      "grad_norm": 10.494202613830566,
      "learning_rate": 4.334715635967379e-06,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8341496884822845,
      "num_tokens": 132101490.0,
      "step": 109880
    },
    {
      "entropy": 1.8309767156839372,
      "epoch": 0.34064938517110355,
      "grad_norm": 8.481719970703125,
      "learning_rate": 4.334518399926636e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.85340007096529,
      "num_tokens": 132114554.0,
      "step": 109890
    },
    {
      "entropy": 1.8747343346476555,
      "epoch": 0.3406803842961532,
      "grad_norm": 9.323307037353516,
      "learning_rate": 4.334321190807049e-06,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8413134127855301,
      "num_tokens": 132126561.0,
      "step": 109900
    },
    {
      "entropy": 1.8665272369980812,
      "epoch": 0.34071138342120294,
      "grad_norm": 9.879739761352539,
      "learning_rate": 4.334124008602491e-06,
      "loss": 0.5596,
      "mean_token_accuracy": 0.8321680203080177,
      "num_tokens": 132137909.0,
      "step": 109910
    },
    {
      "entropy": 1.872728843986988,
      "epoch": 0.3407423825462526,
      "grad_norm": 9.696663856506348,
      "learning_rate": 4.333926853306841e-06,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8571437358856201,
      "num_tokens": 132149808.0,
      "step": 109920
    },
    {
      "entropy": 1.9334780007600785,
      "epoch": 0.34077338167130233,
      "grad_norm": 8.289495468139648,
      "learning_rate": 4.333729724913981e-06,
      "loss": 0.5049,
      "mean_token_accuracy": 0.837024663388729,
      "num_tokens": 132160716.0,
      "step": 109930
    },
    {
      "entropy": 1.879612100124359,
      "epoch": 0.340804380796352,
      "grad_norm": 8.466344833374023,
      "learning_rate": 4.333532623417792e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8417725265026093,
      "num_tokens": 132172002.0,
      "step": 109940
    },
    {
      "entropy": 1.8401907190680504,
      "epoch": 0.3408353799214017,
      "grad_norm": 8.750163078308105,
      "learning_rate": 4.333335548812158e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8564073830842972,
      "num_tokens": 132183522.0,
      "step": 109950
    },
    {
      "entropy": 1.7649420738220214,
      "epoch": 0.3408663790464514,
      "grad_norm": 7.617003917694092,
      "learning_rate": 4.333138501090967e-06,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8601633608341217,
      "num_tokens": 132197109.0,
      "step": 109960
    },
    {
      "entropy": 1.823840895295143,
      "epoch": 0.3408973781715011,
      "grad_norm": 4.131731986999512,
      "learning_rate": 4.332941480248105e-06,
      "loss": 0.4298,
      "mean_token_accuracy": 0.8559944689273834,
      "num_tokens": 132209939.0,
      "step": 109970
    },
    {
      "entropy": 1.8967970684170723,
      "epoch": 0.3409283772965508,
      "grad_norm": 8.258302688598633,
      "learning_rate": 4.332744486277461e-06,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8519051864743232,
      "num_tokens": 132220802.0,
      "step": 109980
    },
    {
      "entropy": 1.8975848436355591,
      "epoch": 0.3409593764216005,
      "grad_norm": 9.49807071685791,
      "learning_rate": 4.332547519172929e-06,
      "loss": 0.469,
      "mean_token_accuracy": 0.8441584467887878,
      "num_tokens": 132232580.0,
      "step": 109990
    },
    {
      "entropy": 1.7971587955951691,
      "epoch": 0.3409903755466502,
      "grad_norm": 8.716933250427246,
      "learning_rate": 4.332350578928402e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8555277913808823,
      "num_tokens": 132244660.0,
      "step": 110000
    },
    {
      "entropy": 1.7943158000707626,
      "epoch": 0.3410213746716999,
      "grad_norm": 9.320921897888184,
      "learning_rate": 4.332153665537777e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8571292623877526,
      "num_tokens": 132257556.0,
      "step": 110010
    },
    {
      "entropy": 1.883206208050251,
      "epoch": 0.3410523737967496,
      "grad_norm": 7.559988021850586,
      "learning_rate": 4.331956778994951e-06,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8475514382123948,
      "num_tokens": 132269837.0,
      "step": 110020
    },
    {
      "entropy": 1.8622760981321336,
      "epoch": 0.3410833729217993,
      "grad_norm": 8.432063102722168,
      "learning_rate": 4.331759919293823e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8502196714282035,
      "num_tokens": 132281620.0,
      "step": 110030
    },
    {
      "entropy": 1.8266623839735985,
      "epoch": 0.34111437204684897,
      "grad_norm": 7.229222774505615,
      "learning_rate": 4.331563086428295e-06,
      "loss": 0.3981,
      "mean_token_accuracy": 0.8582778573036194,
      "num_tokens": 132293944.0,
      "step": 110040
    },
    {
      "entropy": 1.8780689999461173,
      "epoch": 0.3411453711718987,
      "grad_norm": 9.767049789428711,
      "learning_rate": 4.331366280392271e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8368968412280082,
      "num_tokens": 132306054.0,
      "step": 110050
    },
    {
      "entropy": 1.9592034816741943,
      "epoch": 0.34117637029694836,
      "grad_norm": 8.0049467086792,
      "learning_rate": 4.331169501179655e-06,
      "loss": 0.5138,
      "mean_token_accuracy": 0.8449469640851021,
      "num_tokens": 132316729.0,
      "step": 110060
    },
    {
      "entropy": 1.8753671124577522,
      "epoch": 0.3412073694219981,
      "grad_norm": 8.132515907287598,
      "learning_rate": 4.330972748784358e-06,
      "loss": 0.452,
      "mean_token_accuracy": 0.8468053087592124,
      "num_tokens": 132328733.0,
      "step": 110070
    },
    {
      "entropy": 1.9104474276304244,
      "epoch": 0.34123836854704775,
      "grad_norm": 9.560647010803223,
      "learning_rate": 4.330776023200286e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.847663089632988,
      "num_tokens": 132339483.0,
      "step": 110080
    },
    {
      "entropy": 1.7548475116491318,
      "epoch": 0.3412693676720975,
      "grad_norm": 7.8430280685424805,
      "learning_rate": 4.330579324421352e-06,
      "loss": 0.4229,
      "mean_token_accuracy": 0.8698502600193023,
      "num_tokens": 132352754.0,
      "step": 110090
    },
    {
      "entropy": 1.778562317788601,
      "epoch": 0.34130036679714715,
      "grad_norm": 3.4906435012817383,
      "learning_rate": 4.330382652441468e-06,
      "loss": 0.409,
      "mean_token_accuracy": 0.8567349463701248,
      "num_tokens": 132366131.0,
      "step": 110100
    },
    {
      "entropy": 1.8327470771968364,
      "epoch": 0.34133136592219687,
      "grad_norm": 9.07868766784668,
      "learning_rate": 4.33018600725455e-06,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8581578373908997,
      "num_tokens": 132377583.0,
      "step": 110110
    },
    {
      "entropy": 1.8057816326618195,
      "epoch": 0.34136236504724654,
      "grad_norm": 9.56470775604248,
      "learning_rate": 4.329989388854515e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8603610098361969,
      "num_tokens": 132389492.0,
      "step": 110120
    },
    {
      "entropy": 1.8053456172347069,
      "epoch": 0.3413933641722962,
      "grad_norm": 8.362168312072754,
      "learning_rate": 4.3297927972352825e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8586013182997704,
      "num_tokens": 132401915.0,
      "step": 110130
    },
    {
      "entropy": 1.8421172469854354,
      "epoch": 0.34142436329734593,
      "grad_norm": 5.665745735168457,
      "learning_rate": 4.3295962323907735e-06,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8431659877300263,
      "num_tokens": 132414324.0,
      "step": 110140
    },
    {
      "entropy": 1.8274363920092582,
      "epoch": 0.3414553624223956,
      "grad_norm": 4.698655128479004,
      "learning_rate": 4.3293996943149094e-06,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8378838330507279,
      "num_tokens": 132426353.0,
      "step": 110150
    },
    {
      "entropy": 1.8264761924743653,
      "epoch": 0.3414863615474453,
      "grad_norm": 8.485648155212402,
      "learning_rate": 4.329203183001617e-06,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8489476129412651,
      "num_tokens": 132438108.0,
      "step": 110160
    },
    {
      "entropy": 1.7348950892686843,
      "epoch": 0.341517360672495,
      "grad_norm": 3.869995355606079,
      "learning_rate": 4.329006698444822e-06,
      "loss": 0.3746,
      "mean_token_accuracy": 0.8635307684540748,
      "num_tokens": 132452376.0,
      "step": 110170
    },
    {
      "entropy": 1.8900276586413383,
      "epoch": 0.3415483597975447,
      "grad_norm": 7.771602630615234,
      "learning_rate": 4.3288102406384535e-06,
      "loss": 0.5439,
      "mean_token_accuracy": 0.8306577250361442,
      "num_tokens": 132463778.0,
      "step": 110180
    },
    {
      "entropy": 1.816038003563881,
      "epoch": 0.3415793589225944,
      "grad_norm": 3.8563501834869385,
      "learning_rate": 4.3286138095764414e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8512075647711754,
      "num_tokens": 132476109.0,
      "step": 110190
    },
    {
      "entropy": 1.7983648508787156,
      "epoch": 0.3416103580476441,
      "grad_norm": 6.0994873046875,
      "learning_rate": 4.328417405252719e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.857032585144043,
      "num_tokens": 132489219.0,
      "step": 110200
    },
    {
      "entropy": 1.7746062129735947,
      "epoch": 0.3416413571726938,
      "grad_norm": 7.233607769012451,
      "learning_rate": 4.3282210276612215e-06,
      "loss": 0.3889,
      "mean_token_accuracy": 0.8589346572756767,
      "num_tokens": 132501835.0,
      "step": 110210
    },
    {
      "entropy": 1.854676540195942,
      "epoch": 0.3416723562977435,
      "grad_norm": 8.12541389465332,
      "learning_rate": 4.328024676795884e-06,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8524368569254875,
      "num_tokens": 132513519.0,
      "step": 110220
    },
    {
      "entropy": 1.7707249775528908,
      "epoch": 0.3417033554227932,
      "grad_norm": 11.075340270996094,
      "learning_rate": 4.3278283526506465e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8538433343172074,
      "num_tokens": 132526640.0,
      "step": 110230
    },
    {
      "entropy": 1.896279625594616,
      "epoch": 0.3417343545478429,
      "grad_norm": 10.801493644714355,
      "learning_rate": 4.3276320552194465e-06,
      "loss": 0.4733,
      "mean_token_accuracy": 0.8482535511255265,
      "num_tokens": 132538190.0,
      "step": 110240
    },
    {
      "entropy": 1.8617658391594887,
      "epoch": 0.34176535367289257,
      "grad_norm": 7.516333103179932,
      "learning_rate": 4.32743578449623e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8491196766495704,
      "num_tokens": 132550185.0,
      "step": 110250
    },
    {
      "entropy": 1.8173480436205864,
      "epoch": 0.3417963527979423,
      "grad_norm": 8.738038063049316,
      "learning_rate": 4.327239540474937e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.836649663746357,
      "num_tokens": 132563493.0,
      "step": 110260
    },
    {
      "entropy": 1.9029010236263275,
      "epoch": 0.34182735192299196,
      "grad_norm": 8.454696655273438,
      "learning_rate": 4.3270433231495165e-06,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8503329768776894,
      "num_tokens": 132575617.0,
      "step": 110270
    },
    {
      "entropy": 1.8139191955327987,
      "epoch": 0.3418583510480417,
      "grad_norm": 3.524780750274658,
      "learning_rate": 4.326847132513916e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8617829859256745,
      "num_tokens": 132588755.0,
      "step": 110280
    },
    {
      "entropy": 1.7936020240187645,
      "epoch": 0.34188935017309136,
      "grad_norm": 8.44808578491211,
      "learning_rate": 4.326650968562085e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8550214633345604,
      "num_tokens": 132601884.0,
      "step": 110290
    },
    {
      "entropy": 1.8529448464512825,
      "epoch": 0.3419203492981411,
      "grad_norm": 4.374883651733398,
      "learning_rate": 4.3264548312879736e-06,
      "loss": 0.4341,
      "mean_token_accuracy": 0.8531146451830864,
      "num_tokens": 132614584.0,
      "step": 110300
    },
    {
      "entropy": 1.800635115802288,
      "epoch": 0.34195134842319075,
      "grad_norm": 4.555408954620361,
      "learning_rate": 4.326258720685538e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8581522032618523,
      "num_tokens": 132627313.0,
      "step": 110310
    },
    {
      "entropy": 1.7813701510429383,
      "epoch": 0.3419823475482405,
      "grad_norm": 4.661769390106201,
      "learning_rate": 4.326062636748733e-06,
      "loss": 0.3958,
      "mean_token_accuracy": 0.8568063646554946,
      "num_tokens": 132640535.0,
      "step": 110320
    },
    {
      "entropy": 1.8333582818508147,
      "epoch": 0.34201334667329014,
      "grad_norm": 8.417710304260254,
      "learning_rate": 4.325866579471516e-06,
      "loss": 0.4775,
      "mean_token_accuracy": 0.839620991051197,
      "num_tokens": 132652723.0,
      "step": 110330
    },
    {
      "entropy": 1.7189094245433807,
      "epoch": 0.34204434579833987,
      "grad_norm": 6.682080268859863,
      "learning_rate": 4.325670548847847e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.8616839155554772,
      "num_tokens": 132667020.0,
      "step": 110340
    },
    {
      "entropy": 1.9014250546693803,
      "epoch": 0.34207534492338953,
      "grad_norm": 9.291830062866211,
      "learning_rate": 4.325474544871687e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8418972924351692,
      "num_tokens": 132678104.0,
      "step": 110350
    },
    {
      "entropy": 1.8418912634253501,
      "epoch": 0.34210634404843926,
      "grad_norm": 9.592106819152832,
      "learning_rate": 4.325278567536999e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8414492189884186,
      "num_tokens": 132691635.0,
      "step": 110360
    },
    {
      "entropy": 1.8121645241975783,
      "epoch": 0.34213734317348893,
      "grad_norm": 10.184158325195312,
      "learning_rate": 4.325082616837749e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8527585849165916,
      "num_tokens": 132703725.0,
      "step": 110370
    },
    {
      "entropy": 1.8687279969453812,
      "epoch": 0.3421683422985386,
      "grad_norm": 7.617676734924316,
      "learning_rate": 4.324886692767904e-06,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8405532166361809,
      "num_tokens": 132715491.0,
      "step": 110380
    },
    {
      "entropy": 1.8479404971003532,
      "epoch": 0.3421993414235883,
      "grad_norm": 8.61865520477295,
      "learning_rate": 4.324690795321433e-06,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8454162836074829,
      "num_tokens": 132727389.0,
      "step": 110390
    },
    {
      "entropy": 1.8582603454589843,
      "epoch": 0.342230340548638,
      "grad_norm": 8.203682899475098,
      "learning_rate": 4.324494924492305e-06,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8352208867669105,
      "num_tokens": 132739602.0,
      "step": 110400
    },
    {
      "entropy": 1.8665910184383392,
      "epoch": 0.3422613396736877,
      "grad_norm": 8.948098182678223,
      "learning_rate": 4.324299080274496e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8546964704990387,
      "num_tokens": 132751570.0,
      "step": 110410
    },
    {
      "entropy": 1.870258367061615,
      "epoch": 0.3422923387987374,
      "grad_norm": 7.890269756317139,
      "learning_rate": 4.32410326266198e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8475801780819893,
      "num_tokens": 132763398.0,
      "step": 110420
    },
    {
      "entropy": 1.841132828593254,
      "epoch": 0.3423233379237871,
      "grad_norm": 6.834152698516846,
      "learning_rate": 4.3239074716487314e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8546931356191635,
      "num_tokens": 132775491.0,
      "step": 110430
    },
    {
      "entropy": 1.9839861959218978,
      "epoch": 0.3423543370488368,
      "grad_norm": 9.827447891235352,
      "learning_rate": 4.323711707228732e-06,
      "loss": 0.5771,
      "mean_token_accuracy": 0.8222515046596527,
      "num_tokens": 132786260.0,
      "step": 110440
    },
    {
      "entropy": 1.869850617647171,
      "epoch": 0.3423853361738865,
      "grad_norm": 7.602901935577393,
      "learning_rate": 4.323515969395961e-06,
      "loss": 0.487,
      "mean_token_accuracy": 0.849778589606285,
      "num_tokens": 132798632.0,
      "step": 110450
    },
    {
      "entropy": 1.8364197805523872,
      "epoch": 0.34241633529893617,
      "grad_norm": 6.836818695068359,
      "learning_rate": 4.3233202581444e-06,
      "loss": 0.3919,
      "mean_token_accuracy": 0.8704248234629631,
      "num_tokens": 132811082.0,
      "step": 110460
    },
    {
      "entropy": 1.76078250259161,
      "epoch": 0.3424473344239859,
      "grad_norm": 3.4893312454223633,
      "learning_rate": 4.323124573468033e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8747459024190902,
      "num_tokens": 132824398.0,
      "step": 110470
    },
    {
      "entropy": 1.9431472271680832,
      "epoch": 0.34247833354903556,
      "grad_norm": 8.165310859680176,
      "learning_rate": 4.3229289153608484e-06,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8350071504712104,
      "num_tokens": 132835017.0,
      "step": 110480
    },
    {
      "entropy": 1.828757031261921,
      "epoch": 0.3425093326740853,
      "grad_norm": 5.19187593460083,
      "learning_rate": 4.3227332838168335e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8579125210642815,
      "num_tokens": 132847063.0,
      "step": 110490
    },
    {
      "entropy": 1.907851468026638,
      "epoch": 0.34254033179913496,
      "grad_norm": 8.472184181213379,
      "learning_rate": 4.3225376788299765e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8507169172167778,
      "num_tokens": 132858770.0,
      "step": 110500
    },
    {
      "entropy": 1.9092920407652856,
      "epoch": 0.3425713309241847,
      "grad_norm": 8.181653022766113,
      "learning_rate": 4.322342100394272e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8374480813741684,
      "num_tokens": 132869792.0,
      "step": 110510
    },
    {
      "entropy": 1.801915517449379,
      "epoch": 0.34260233004923435,
      "grad_norm": 9.726790428161621,
      "learning_rate": 4.322146548503712e-06,
      "loss": 0.37,
      "mean_token_accuracy": 0.8696004092693329,
      "num_tokens": 132882414.0,
      "step": 110520
    },
    {
      "entropy": 1.871196947991848,
      "epoch": 0.3426333291742841,
      "grad_norm": 8.217884063720703,
      "learning_rate": 4.321951023152293e-06,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8464583426713943,
      "num_tokens": 132894145.0,
      "step": 110530
    },
    {
      "entropy": 1.8100114524364472,
      "epoch": 0.34266432829933374,
      "grad_norm": 9.168874740600586,
      "learning_rate": 4.321755524334014e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8399093985557556,
      "num_tokens": 132907133.0,
      "step": 110540
    },
    {
      "entropy": 1.8548692613840103,
      "epoch": 0.34269532742438347,
      "grad_norm": 8.339598655700684,
      "learning_rate": 4.3215600520428715e-06,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8403585880994797,
      "num_tokens": 132918583.0,
      "step": 110550
    },
    {
      "entropy": 1.8597151398658753,
      "epoch": 0.34272632654943314,
      "grad_norm": 8.382675170898438,
      "learning_rate": 4.3213646062728695e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8416018903255462,
      "num_tokens": 132930506.0,
      "step": 110560
    },
    {
      "entropy": 1.8886392131447791,
      "epoch": 0.34275732567448286,
      "grad_norm": 4.080488204956055,
      "learning_rate": 4.321169187018011e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8549024015665054,
      "num_tokens": 132942640.0,
      "step": 110570
    },
    {
      "entropy": 1.7458569899201393,
      "epoch": 0.34278832479953253,
      "grad_norm": 5.599550724029541,
      "learning_rate": 4.3209737942722985e-06,
      "loss": 0.3801,
      "mean_token_accuracy": 0.8586847379803657,
      "num_tokens": 132956699.0,
      "step": 110580
    },
    {
      "entropy": 1.8726666495203972,
      "epoch": 0.34281932392458225,
      "grad_norm": 7.666844367980957,
      "learning_rate": 4.320778428029743e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8506060406565666,
      "num_tokens": 132967858.0,
      "step": 110590
    },
    {
      "entropy": 1.8701406121253967,
      "epoch": 0.3428503230496319,
      "grad_norm": 9.031076431274414,
      "learning_rate": 4.320583088284352e-06,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8525993511080742,
      "num_tokens": 132979844.0,
      "step": 110600
    },
    {
      "entropy": 1.7444271817803383,
      "epoch": 0.34288132217468165,
      "grad_norm": 3.681898593902588,
      "learning_rate": 4.320387775030135e-06,
      "loss": 0.3891,
      "mean_token_accuracy": 0.8686041355133056,
      "num_tokens": 132993483.0,
      "step": 110610
    },
    {
      "entropy": 1.8315519198775292,
      "epoch": 0.3429123212997313,
      "grad_norm": 13.511602401733398,
      "learning_rate": 4.320192488261108e-06,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8444625332951545,
      "num_tokens": 133006258.0,
      "step": 110620
    },
    {
      "entropy": 1.8297417625784873,
      "epoch": 0.342943320424781,
      "grad_norm": 4.069921016693115,
      "learning_rate": 4.319997227971282e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8498388364911079,
      "num_tokens": 133018876.0,
      "step": 110630
    },
    {
      "entropy": 1.9210706606507302,
      "epoch": 0.3429743195498307,
      "grad_norm": 9.114517211914062,
      "learning_rate": 4.319801994154677e-06,
      "loss": 0.536,
      "mean_token_accuracy": 0.8308271437883377,
      "num_tokens": 133030224.0,
      "step": 110640
    },
    {
      "entropy": 1.8908802688121795,
      "epoch": 0.3430053186748804,
      "grad_norm": 9.932719230651855,
      "learning_rate": 4.319606786805309e-06,
      "loss": 0.529,
      "mean_token_accuracy": 0.8441593378782273,
      "num_tokens": 133041073.0,
      "step": 110650
    },
    {
      "entropy": 1.8904657304286956,
      "epoch": 0.3430363177999301,
      "grad_norm": 7.985291957855225,
      "learning_rate": 4.3194116059172e-06,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8325783342123032,
      "num_tokens": 133052277.0,
      "step": 110660
    },
    {
      "entropy": 1.9574284851551056,
      "epoch": 0.34306731692497977,
      "grad_norm": 7.235466957092285,
      "learning_rate": 4.319216451484371e-06,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8355097994208336,
      "num_tokens": 133064090.0,
      "step": 110670
    },
    {
      "entropy": 1.816710241138935,
      "epoch": 0.3430983160500295,
      "grad_norm": 8.71338939666748,
      "learning_rate": 4.319021323500848e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8443104445934295,
      "num_tokens": 133076199.0,
      "step": 110680
    },
    {
      "entropy": 1.843747340142727,
      "epoch": 0.34312931517507916,
      "grad_norm": 6.804162502288818,
      "learning_rate": 4.318826221960655e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8444552600383759,
      "num_tokens": 133088227.0,
      "step": 110690
    },
    {
      "entropy": 1.8602527409791947,
      "epoch": 0.3431603143001289,
      "grad_norm": 7.936315059661865,
      "learning_rate": 4.318631146857822e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8574521571397782,
      "num_tokens": 133100175.0,
      "step": 110700
    },
    {
      "entropy": 1.8244513690471649,
      "epoch": 0.34319131342517856,
      "grad_norm": 11.78140926361084,
      "learning_rate": 4.318436098186376e-06,
      "loss": 0.416,
      "mean_token_accuracy": 0.8596523448824882,
      "num_tokens": 133111645.0,
      "step": 110710
    },
    {
      "entropy": 1.8955079719424248,
      "epoch": 0.3432223125502283,
      "grad_norm": 5.455853462219238,
      "learning_rate": 4.318241075940353e-06,
      "loss": 0.5204,
      "mean_token_accuracy": 0.8417312771081924,
      "num_tokens": 133122905.0,
      "step": 110720
    },
    {
      "entropy": 1.898301364481449,
      "epoch": 0.34325331167527795,
      "grad_norm": 8.922945976257324,
      "learning_rate": 4.318046080113783e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8424309939146042,
      "num_tokens": 133134095.0,
      "step": 110730
    },
    {
      "entropy": 1.8243930265307426,
      "epoch": 0.3432843108003277,
      "grad_norm": 4.409710884094238,
      "learning_rate": 4.317851110700703e-06,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8556462466716767,
      "num_tokens": 133146666.0,
      "step": 110740
    },
    {
      "entropy": 1.9261058151721955,
      "epoch": 0.34331530992537734,
      "grad_norm": 12.174426078796387,
      "learning_rate": 4.31765616769515e-06,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8526764348149299,
      "num_tokens": 133157255.0,
      "step": 110750
    },
    {
      "entropy": 1.911964663863182,
      "epoch": 0.34334630905042707,
      "grad_norm": 8.915404319763184,
      "learning_rate": 4.317461251091163e-06,
      "loss": 0.4823,
      "mean_token_accuracy": 0.8492590352892876,
      "num_tokens": 133168516.0,
      "step": 110760
    },
    {
      "entropy": 1.8462224438786508,
      "epoch": 0.34337730817547674,
      "grad_norm": 9.527721405029297,
      "learning_rate": 4.317266360882783e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8471042737364769,
      "num_tokens": 133180707.0,
      "step": 110770
    },
    {
      "entropy": 1.821668528020382,
      "epoch": 0.34340830730052646,
      "grad_norm": 3.8620917797088623,
      "learning_rate": 4.3170714970640535e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.859420596063137,
      "num_tokens": 133192497.0,
      "step": 110780
    },
    {
      "entropy": 1.875546859204769,
      "epoch": 0.34343930642557613,
      "grad_norm": 7.777998924255371,
      "learning_rate": 4.3168766596290205e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8493057772517204,
      "num_tokens": 133204389.0,
      "step": 110790
    },
    {
      "entropy": 1.807924547791481,
      "epoch": 0.34347030555062585,
      "grad_norm": 12.520805358886719,
      "learning_rate": 4.31668184857173e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8617511481046677,
      "num_tokens": 133217528.0,
      "step": 110800
    },
    {
      "entropy": 1.8352250516414643,
      "epoch": 0.3435013046756755,
      "grad_norm": 5.211875915527344,
      "learning_rate": 4.31648706388623e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8534446090459824,
      "num_tokens": 133229712.0,
      "step": 110810
    },
    {
      "entropy": 1.8670003831386566,
      "epoch": 0.34353230380072525,
      "grad_norm": 7.83875846862793,
      "learning_rate": 4.316292305566571e-06,
      "loss": 0.4622,
      "mean_token_accuracy": 0.850574080646038,
      "num_tokens": 133241353.0,
      "step": 110820
    },
    {
      "entropy": 1.887353539466858,
      "epoch": 0.3435633029257749,
      "grad_norm": 10.31457233428955,
      "learning_rate": 4.316097573606808e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8490018859505654,
      "num_tokens": 133252486.0,
      "step": 110830
    },
    {
      "entropy": 1.891994397342205,
      "epoch": 0.34359430205082464,
      "grad_norm": 8.610884666442871,
      "learning_rate": 4.315902868000992e-06,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8397418484091759,
      "num_tokens": 133264441.0,
      "step": 110840
    },
    {
      "entropy": 1.8664520055055618,
      "epoch": 0.3436253011758743,
      "grad_norm": 7.726016044616699,
      "learning_rate": 4.3157081887431804e-06,
      "loss": 0.499,
      "mean_token_accuracy": 0.8486996784806251,
      "num_tokens": 133276898.0,
      "step": 110850
    },
    {
      "entropy": 1.884059876203537,
      "epoch": 0.34365630030092403,
      "grad_norm": 4.186914443969727,
      "learning_rate": 4.315513535827431e-06,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8496600687503815,
      "num_tokens": 133288536.0,
      "step": 110860
    },
    {
      "entropy": 1.9221114248037339,
      "epoch": 0.3436872994259737,
      "grad_norm": 9.337282180786133,
      "learning_rate": 4.315318909247805e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.842443785071373,
      "num_tokens": 133299978.0,
      "step": 110870
    },
    {
      "entropy": 1.8485377907752991,
      "epoch": 0.34371829855102337,
      "grad_norm": 9.948966979980469,
      "learning_rate": 4.315124308998364e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8506104618310928,
      "num_tokens": 133311745.0,
      "step": 110880
    },
    {
      "entropy": 1.8648578524589539,
      "epoch": 0.3437492976760731,
      "grad_norm": 7.327975749969482,
      "learning_rate": 4.31492973507317e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8568118900060654,
      "num_tokens": 133323824.0,
      "step": 110890
    },
    {
      "entropy": 1.9047629460692406,
      "epoch": 0.34378029680112276,
      "grad_norm": 8.159683227539062,
      "learning_rate": 4.314735187466291e-06,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8451401859521865,
      "num_tokens": 133335255.0,
      "step": 110900
    },
    {
      "entropy": 1.925184278190136,
      "epoch": 0.3438112959261725,
      "grad_norm": 4.655861854553223,
      "learning_rate": 4.3145406661717925e-06,
      "loss": 0.5484,
      "mean_token_accuracy": 0.826390016078949,
      "num_tokens": 133347385.0,
      "step": 110910
    },
    {
      "entropy": 1.9013187378644942,
      "epoch": 0.34384229505122216,
      "grad_norm": 7.1715922355651855,
      "learning_rate": 4.3143461711837445e-06,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8376141622662544,
      "num_tokens": 133358459.0,
      "step": 110920
    },
    {
      "entropy": 1.9203463315963745,
      "epoch": 0.3438732941762719,
      "grad_norm": 10.447772026062012,
      "learning_rate": 4.314151702496219e-06,
      "loss": 0.5415,
      "mean_token_accuracy": 0.8378437146544456,
      "num_tokens": 133369315.0,
      "step": 110930
    },
    {
      "entropy": 1.926878383755684,
      "epoch": 0.34390429330132155,
      "grad_norm": 7.588080406188965,
      "learning_rate": 4.313957260103287e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8456119611859322,
      "num_tokens": 133379860.0,
      "step": 110940
    },
    {
      "entropy": 1.714067880809307,
      "epoch": 0.3439352924263713,
      "grad_norm": 4.141031265258789,
      "learning_rate": 4.313762843999025e-06,
      "loss": 0.3662,
      "mean_token_accuracy": 0.8552543878555298,
      "num_tokens": 133394752.0,
      "step": 110950
    },
    {
      "entropy": 1.7716796442866325,
      "epoch": 0.34396629155142094,
      "grad_norm": 8.312151908874512,
      "learning_rate": 4.31356845417751e-06,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8612663641571998,
      "num_tokens": 133407887.0,
      "step": 110960
    },
    {
      "entropy": 1.909244528412819,
      "epoch": 0.34399729067647067,
      "grad_norm": 9.186685562133789,
      "learning_rate": 4.31337409063282e-06,
      "loss": 0.5101,
      "mean_token_accuracy": 0.8425434127449989,
      "num_tokens": 133419016.0,
      "step": 110970
    },
    {
      "entropy": 1.831558196246624,
      "epoch": 0.34402828980152034,
      "grad_norm": 7.182348728179932,
      "learning_rate": 4.3131797533590354e-06,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8547964975237846,
      "num_tokens": 133431120.0,
      "step": 110980
    },
    {
      "entropy": 1.923382543027401,
      "epoch": 0.34405928892657006,
      "grad_norm": 7.744353771209717,
      "learning_rate": 4.31298544235024e-06,
      "loss": 0.5213,
      "mean_token_accuracy": 0.8368471801280976,
      "num_tokens": 133442547.0,
      "step": 110990
    },
    {
      "entropy": 1.8427509516477585,
      "epoch": 0.34409028805161973,
      "grad_norm": 8.841486930847168,
      "learning_rate": 4.312791157600516e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8453793004155159,
      "num_tokens": 133455042.0,
      "step": 111000
    },
    {
      "entropy": 1.854433636367321,
      "epoch": 0.34412128717666945,
      "grad_norm": 8.208456039428711,
      "learning_rate": 4.312596899103951e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8534053891897202,
      "num_tokens": 133466808.0,
      "step": 111010
    },
    {
      "entropy": 1.936078244447708,
      "epoch": 0.3441522863017191,
      "grad_norm": 10.055807113647461,
      "learning_rate": 4.312402666854633e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8546758487820625,
      "num_tokens": 133478433.0,
      "step": 111020
    },
    {
      "entropy": 1.8237747445702552,
      "epoch": 0.34418328542676885,
      "grad_norm": 10.912522315979004,
      "learning_rate": 4.312208460846651e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.848297068476677,
      "num_tokens": 133490903.0,
      "step": 111030
    },
    {
      "entropy": 1.8850932896137238,
      "epoch": 0.3442142845518185,
      "grad_norm": 8.489951133728027,
      "learning_rate": 4.312014281074098e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8428701400756836,
      "num_tokens": 133503256.0,
      "step": 111040
    },
    {
      "entropy": 1.80315043926239,
      "epoch": 0.34424528367686824,
      "grad_norm": 4.91754674911499,
      "learning_rate": 4.311820127531066e-06,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8500543773174286,
      "num_tokens": 133515626.0,
      "step": 111050
    },
    {
      "entropy": 1.8291157126426696,
      "epoch": 0.3442762828019179,
      "grad_norm": 8.575767517089844,
      "learning_rate": 4.3116260002116505e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8480419516563416,
      "num_tokens": 133528210.0,
      "step": 111060
    },
    {
      "entropy": 1.8445888727903366,
      "epoch": 0.34430728192696763,
      "grad_norm": 7.410253524780273,
      "learning_rate": 4.3114318991099505e-06,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8509875819087028,
      "num_tokens": 133539348.0,
      "step": 111070
    },
    {
      "entropy": 1.7617506772279738,
      "epoch": 0.3443382810520173,
      "grad_norm": 4.4981770515441895,
      "learning_rate": 4.311237824220064e-06,
      "loss": 0.4237,
      "mean_token_accuracy": 0.8568379059433937,
      "num_tokens": 133552472.0,
      "step": 111080
    },
    {
      "entropy": 1.845915214717388,
      "epoch": 0.344369280177067,
      "grad_norm": 9.03842544555664,
      "learning_rate": 4.311043775536092e-06,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8473999381065369,
      "num_tokens": 133564707.0,
      "step": 111090
    },
    {
      "entropy": 1.685810850560665,
      "epoch": 0.3444002793021167,
      "grad_norm": 4.30578088760376,
      "learning_rate": 4.310849753052138e-06,
      "loss": 0.3662,
      "mean_token_accuracy": 0.864673687517643,
      "num_tokens": 133579259.0,
      "step": 111100
    },
    {
      "entropy": 1.8005015209317208,
      "epoch": 0.3444312784271664,
      "grad_norm": 9.097541809082031,
      "learning_rate": 4.310655756762307e-06,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8453745514154434,
      "num_tokens": 133591953.0,
      "step": 111110
    },
    {
      "entropy": 1.8135478526353837,
      "epoch": 0.3444622775522161,
      "grad_norm": 3.2069530487060547,
      "learning_rate": 4.310461786660705e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8551574990153312,
      "num_tokens": 133604208.0,
      "step": 111120
    },
    {
      "entropy": 1.873196867108345,
      "epoch": 0.34449327667726576,
      "grad_norm": 8.344642639160156,
      "learning_rate": 4.310267842741439e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8539148271083832,
      "num_tokens": 133616056.0,
      "step": 111130
    },
    {
      "entropy": 1.9149486266076565,
      "epoch": 0.3445242758023155,
      "grad_norm": 8.610182762145996,
      "learning_rate": 4.3100739249986225e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8442770466208458,
      "num_tokens": 133627877.0,
      "step": 111140
    },
    {
      "entropy": 1.910437923669815,
      "epoch": 0.34455527492736515,
      "grad_norm": 8.95980167388916,
      "learning_rate": 4.309880033426365e-06,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8403397217392922,
      "num_tokens": 133639187.0,
      "step": 111150
    },
    {
      "entropy": 1.9167566150426865,
      "epoch": 0.3445862740524149,
      "grad_norm": 9.184028625488281,
      "learning_rate": 4.3096861680187815e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8474911168217659,
      "num_tokens": 133650200.0,
      "step": 111160
    },
    {
      "entropy": 1.9442434057593345,
      "epoch": 0.34461727317746454,
      "grad_norm": 9.074106216430664,
      "learning_rate": 4.309492328769988e-06,
      "loss": 0.5261,
      "mean_token_accuracy": 0.8301569879055023,
      "num_tokens": 133661615.0,
      "step": 111170
    },
    {
      "entropy": 1.8889770179986953,
      "epoch": 0.34464827230251427,
      "grad_norm": 11.795031547546387,
      "learning_rate": 4.309298515674102e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8382586434483528,
      "num_tokens": 133672713.0,
      "step": 111180
    },
    {
      "entropy": 1.8732004672288896,
      "epoch": 0.34467927142756394,
      "grad_norm": 7.575119495391846,
      "learning_rate": 4.309104728725243e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8471902221441269,
      "num_tokens": 133683668.0,
      "step": 111190
    },
    {
      "entropy": 1.9257921800017357,
      "epoch": 0.34471027055261366,
      "grad_norm": 7.821493148803711,
      "learning_rate": 4.308910967917533e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8495068654417992,
      "num_tokens": 133694848.0,
      "step": 111200
    },
    {
      "entropy": 1.876696328818798,
      "epoch": 0.34474126967766333,
      "grad_norm": 9.46831226348877,
      "learning_rate": 4.308717233245096e-06,
      "loss": 0.4807,
      "mean_token_accuracy": 0.847463445365429,
      "num_tokens": 133706948.0,
      "step": 111210
    },
    {
      "entropy": 1.9034352615475654,
      "epoch": 0.34477226880271306,
      "grad_norm": 4.813274383544922,
      "learning_rate": 4.3085235247020545e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8571040198206902,
      "num_tokens": 133719003.0,
      "step": 111220
    },
    {
      "entropy": 1.8417223021388054,
      "epoch": 0.3448032679277627,
      "grad_norm": 8.651920318603516,
      "learning_rate": 4.3083298422825375e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8489273890852929,
      "num_tokens": 133731771.0,
      "step": 111230
    },
    {
      "entropy": 1.919362673163414,
      "epoch": 0.34483426705281245,
      "grad_norm": 7.571285724639893,
      "learning_rate": 4.308136185980673e-06,
      "loss": 0.5243,
      "mean_token_accuracy": 0.839247289299965,
      "num_tokens": 133742822.0,
      "step": 111240
    },
    {
      "entropy": 1.8016601279377937,
      "epoch": 0.3448652661778621,
      "grad_norm": 9.703665733337402,
      "learning_rate": 4.307942555790593e-06,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8540796846151352,
      "num_tokens": 133755750.0,
      "step": 111250
    },
    {
      "entropy": 1.7813518926501275,
      "epoch": 0.34489626530291184,
      "grad_norm": 4.28598690032959,
      "learning_rate": 4.3077489517064285e-06,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8613515332341194,
      "num_tokens": 133768622.0,
      "step": 111260
    },
    {
      "entropy": 1.8209292754530906,
      "epoch": 0.3449272644279615,
      "grad_norm": 8.168103218078613,
      "learning_rate": 4.307555373722316e-06,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8652269795536995,
      "num_tokens": 133780564.0,
      "step": 111270
    },
    {
      "entropy": 1.8425551727414131,
      "epoch": 0.34495826355301124,
      "grad_norm": 8.19618034362793,
      "learning_rate": 4.307361821832388e-06,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8445751652121544,
      "num_tokens": 133792677.0,
      "step": 111280
    },
    {
      "entropy": 1.824992810189724,
      "epoch": 0.3449892626780609,
      "grad_norm": 8.843597412109375,
      "learning_rate": 4.307168296030786e-06,
      "loss": 0.4662,
      "mean_token_accuracy": 0.8552077680826187,
      "num_tokens": 133805279.0,
      "step": 111290
    },
    {
      "entropy": 1.9060368582606315,
      "epoch": 0.34502026180311063,
      "grad_norm": 3.7033939361572266,
      "learning_rate": 4.306974796311647e-06,
      "loss": 0.4961,
      "mean_token_accuracy": 0.841067835688591,
      "num_tokens": 133817633.0,
      "step": 111300
    },
    {
      "entropy": 1.864403063058853,
      "epoch": 0.3450512609281603,
      "grad_norm": 7.459976673126221,
      "learning_rate": 4.306781322669116e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8469650357961654,
      "num_tokens": 133829133.0,
      "step": 111310
    },
    {
      "entropy": 1.8203039526939393,
      "epoch": 0.34508226005321,
      "grad_norm": 9.59398365020752,
      "learning_rate": 4.306587875097335e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8534867212176322,
      "num_tokens": 133841674.0,
      "step": 111320
    },
    {
      "entropy": 1.843678180873394,
      "epoch": 0.3451132591782597,
      "grad_norm": 11.09599781036377,
      "learning_rate": 4.306394453590449e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8509141758084298,
      "num_tokens": 133853073.0,
      "step": 111330
    },
    {
      "entropy": 1.837014827132225,
      "epoch": 0.3451442583033094,
      "grad_norm": 7.771301746368408,
      "learning_rate": 4.306201058142605e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8529363244771957,
      "num_tokens": 133865482.0,
      "step": 111340
    },
    {
      "entropy": 1.8316177785396577,
      "epoch": 0.3451752574283591,
      "grad_norm": 3.7739436626434326,
      "learning_rate": 4.3060076887479545e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8522952899336815,
      "num_tokens": 133877076.0,
      "step": 111350
    },
    {
      "entropy": 1.8375495672225952,
      "epoch": 0.34520625655340875,
      "grad_norm": 2.9696900844573975,
      "learning_rate": 4.305814345400645e-06,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8476609244942666,
      "num_tokens": 133889360.0,
      "step": 111360
    },
    {
      "entropy": 1.783223459124565,
      "epoch": 0.3452372556784585,
      "grad_norm": 7.960699081420898,
      "learning_rate": 4.305621028094832e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8435281231999397,
      "num_tokens": 133903252.0,
      "step": 111370
    },
    {
      "entropy": 1.84706342369318,
      "epoch": 0.34526825480350815,
      "grad_norm": 7.111573696136475,
      "learning_rate": 4.305427736824668e-06,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8439237996935844,
      "num_tokens": 133915237.0,
      "step": 111380
    },
    {
      "entropy": 1.7956417188048364,
      "epoch": 0.34529925392855787,
      "grad_norm": 7.185604572296143,
      "learning_rate": 4.305234471584312e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8640512511134147,
      "num_tokens": 133927538.0,
      "step": 111390
    },
    {
      "entropy": 1.8561489313840867,
      "epoch": 0.34533025305360754,
      "grad_norm": 8.284113883972168,
      "learning_rate": 4.3050412323679206e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8489404112100601,
      "num_tokens": 133938940.0,
      "step": 111400
    },
    {
      "entropy": 1.8327260576188564,
      "epoch": 0.34536125217865726,
      "grad_norm": 3.422919750213623,
      "learning_rate": 4.3048480191696544e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8489983096718788,
      "num_tokens": 133951108.0,
      "step": 111410
    },
    {
      "entropy": 1.8617496728897094,
      "epoch": 0.34539225130370693,
      "grad_norm": 7.909846305847168,
      "learning_rate": 4.304654831983675e-06,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8476134285330772,
      "num_tokens": 133963396.0,
      "step": 111420
    },
    {
      "entropy": 1.9265403121709823,
      "epoch": 0.34542325042875666,
      "grad_norm": 7.380890369415283,
      "learning_rate": 4.304461670804146e-06,
      "loss": 0.5297,
      "mean_token_accuracy": 0.841274605691433,
      "num_tokens": 133974112.0,
      "step": 111430
    },
    {
      "entropy": 1.8644670337438582,
      "epoch": 0.3454542495538063,
      "grad_norm": 8.807491302490234,
      "learning_rate": 4.3042685356252335e-06,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8540144443511963,
      "num_tokens": 133985279.0,
      "step": 111440
    },
    {
      "entropy": 1.9059779167175293,
      "epoch": 0.34548524867885605,
      "grad_norm": 7.666346073150635,
      "learning_rate": 4.304075426441105e-06,
      "loss": 0.5396,
      "mean_token_accuracy": 0.8371213942766189,
      "num_tokens": 133996376.0,
      "step": 111450
    },
    {
      "entropy": 1.8472643613815307,
      "epoch": 0.3455162478039057,
      "grad_norm": 7.403988361358643,
      "learning_rate": 4.3038823432459305e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8564211338758468,
      "num_tokens": 134008017.0,
      "step": 111460
    },
    {
      "entropy": 1.8477882623672486,
      "epoch": 0.34554724692895544,
      "grad_norm": 7.693085670471191,
      "learning_rate": 4.30368928603388e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8446563795208931,
      "num_tokens": 134020115.0,
      "step": 111470
    },
    {
      "entropy": 1.8344331562519074,
      "epoch": 0.3455782460540051,
      "grad_norm": 4.255249977111816,
      "learning_rate": 4.303496254799126e-06,
      "loss": 0.465,
      "mean_token_accuracy": 0.8438744112849236,
      "num_tokens": 134032355.0,
      "step": 111480
    },
    {
      "entropy": 1.8852899730205537,
      "epoch": 0.34560924517905484,
      "grad_norm": 9.30275821685791,
      "learning_rate": 4.303303249535845e-06,
      "loss": 0.4825,
      "mean_token_accuracy": 0.8437858000397682,
      "num_tokens": 134043633.0,
      "step": 111490
    },
    {
      "entropy": 1.7422870948910714,
      "epoch": 0.3456402443041045,
      "grad_norm": 7.4847941398620605,
      "learning_rate": 4.3031102702382125e-06,
      "loss": 0.3388,
      "mean_token_accuracy": 0.8664668053388596,
      "num_tokens": 134057736.0,
      "step": 111500
    },
    {
      "entropy": 1.7793061807751656,
      "epoch": 0.34567124342915423,
      "grad_norm": 3.7864532470703125,
      "learning_rate": 4.302917316900407e-06,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8605924814939498,
      "num_tokens": 134070395.0,
      "step": 111510
    },
    {
      "entropy": 1.8568082675337791,
      "epoch": 0.3457022425542039,
      "grad_norm": 11.483708381652832,
      "learning_rate": 4.302724389516609e-06,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8510003834962845,
      "num_tokens": 134082721.0,
      "step": 111520
    },
    {
      "entropy": 1.8132430225610734,
      "epoch": 0.3457332416792536,
      "grad_norm": 2.2380177974700928,
      "learning_rate": 4.302531488081001e-06,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8539024695754052,
      "num_tokens": 134094940.0,
      "step": 111530
    },
    {
      "entropy": 1.7989331185817719,
      "epoch": 0.3457642408043033,
      "grad_norm": 3.9611406326293945,
      "learning_rate": 4.302338612587765e-06,
      "loss": 0.4067,
      "mean_token_accuracy": 0.856696180999279,
      "num_tokens": 134108140.0,
      "step": 111540
    },
    {
      "entropy": 1.85663383603096,
      "epoch": 0.345795239929353,
      "grad_norm": 3.9444892406463623,
      "learning_rate": 4.3021457630310894e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.860635980963707,
      "num_tokens": 134119906.0,
      "step": 111550
    },
    {
      "entropy": 1.8466177895665168,
      "epoch": 0.3458262390544027,
      "grad_norm": 9.456862449645996,
      "learning_rate": 4.30195293940516e-06,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8576301142573357,
      "num_tokens": 134132194.0,
      "step": 111560
    },
    {
      "entropy": 1.8773884311318398,
      "epoch": 0.3458572381794524,
      "grad_norm": 6.836810111999512,
      "learning_rate": 4.301760141704167e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8493657246232033,
      "num_tokens": 134143520.0,
      "step": 111570
    },
    {
      "entropy": 1.9150910288095475,
      "epoch": 0.3458882373045021,
      "grad_norm": 8.402644157409668,
      "learning_rate": 4.301567369922301e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8446568265557289,
      "num_tokens": 134154428.0,
      "step": 111580
    },
    {
      "entropy": 1.911669085919857,
      "epoch": 0.3459192364295518,
      "grad_norm": 8.802173614501953,
      "learning_rate": 4.301374624053757e-06,
      "loss": 0.4807,
      "mean_token_accuracy": 0.849866247177124,
      "num_tokens": 134165966.0,
      "step": 111590
    },
    {
      "entropy": 1.8143538311123848,
      "epoch": 0.34595023555460147,
      "grad_norm": 8.931032180786133,
      "learning_rate": 4.301181904092727e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8532052963972092,
      "num_tokens": 134178567.0,
      "step": 111600
    },
    {
      "entropy": 1.8561271876096725,
      "epoch": 0.34598123467965114,
      "grad_norm": 9.355246543884277,
      "learning_rate": 4.300989210033409e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8413384512066842,
      "num_tokens": 134190432.0,
      "step": 111610
    },
    {
      "entropy": 1.8068694084882737,
      "epoch": 0.34601223380470086,
      "grad_norm": 8.892166137695312,
      "learning_rate": 4.3007965418700015e-06,
      "loss": 0.4047,
      "mean_token_accuracy": 0.8617716804146767,
      "num_tokens": 134202936.0,
      "step": 111620
    },
    {
      "entropy": 1.819658127427101,
      "epoch": 0.34604323292975053,
      "grad_norm": 4.105354309082031,
      "learning_rate": 4.300603899596706e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8501406639814377,
      "num_tokens": 134215290.0,
      "step": 111630
    },
    {
      "entropy": 1.7999876148998737,
      "epoch": 0.34607423205480026,
      "grad_norm": 3.328152894973755,
      "learning_rate": 4.300411283207722e-06,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8598711341619492,
      "num_tokens": 134228492.0,
      "step": 111640
    },
    {
      "entropy": 1.8138612404465675,
      "epoch": 0.3461052311798499,
      "grad_norm": 8.897016525268555,
      "learning_rate": 4.300218692697255e-06,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8562261596322059,
      "num_tokens": 134240367.0,
      "step": 111650
    },
    {
      "entropy": 1.8443049594759942,
      "epoch": 0.34613623030489965,
      "grad_norm": 7.959166049957275,
      "learning_rate": 4.300026128059511e-06,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8506998106837272,
      "num_tokens": 134252476.0,
      "step": 111660
    },
    {
      "entropy": 1.8540190950036048,
      "epoch": 0.3461672294299493,
      "grad_norm": 4.293795585632324,
      "learning_rate": 4.2998335892886964e-06,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8463447079062462,
      "num_tokens": 134264385.0,
      "step": 111670
    },
    {
      "entropy": 1.8193232595920563,
      "epoch": 0.34619822855499904,
      "grad_norm": 8.931365966796875,
      "learning_rate": 4.2996410763790225e-06,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8522950485348701,
      "num_tokens": 134276602.0,
      "step": 111680
    },
    {
      "entropy": 1.8343895703554154,
      "epoch": 0.3462292276800487,
      "grad_norm": 9.969979286193848,
      "learning_rate": 4.2994485893247e-06,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8539774760603904,
      "num_tokens": 134288782.0,
      "step": 111690
    },
    {
      "entropy": 1.8512503013014794,
      "epoch": 0.34626022680509844,
      "grad_norm": 2.8974452018737793,
      "learning_rate": 4.2992561281199405e-06,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8387772515416145,
      "num_tokens": 134301209.0,
      "step": 111700
    },
    {
      "entropy": 1.8505238316953183,
      "epoch": 0.3462912259301481,
      "grad_norm": 3.372145175933838,
      "learning_rate": 4.29906369275896e-06,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8435917362570763,
      "num_tokens": 134313368.0,
      "step": 111710
    },
    {
      "entropy": 1.8507800638675689,
      "epoch": 0.34632222505519783,
      "grad_norm": 7.913898944854736,
      "learning_rate": 4.2988712832359755e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8630237206816673,
      "num_tokens": 134325077.0,
      "step": 111720
    },
    {
      "entropy": 1.7628626979887485,
      "epoch": 0.3463532241802475,
      "grad_norm": 3.487839698791504,
      "learning_rate": 4.2986788995452044e-06,
      "loss": 0.3905,
      "mean_token_accuracy": 0.8562428668141365,
      "num_tokens": 134339036.0,
      "step": 111730
    },
    {
      "entropy": 1.843147909641266,
      "epoch": 0.3463842233052972,
      "grad_norm": 6.786396026611328,
      "learning_rate": 4.298486541680868e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8427472576498986,
      "num_tokens": 134351445.0,
      "step": 111740
    },
    {
      "entropy": 1.807756444811821,
      "epoch": 0.3464152224303469,
      "grad_norm": 5.164430141448975,
      "learning_rate": 4.298294209637186e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.853118185698986,
      "num_tokens": 134365069.0,
      "step": 111750
    },
    {
      "entropy": 1.859677466750145,
      "epoch": 0.3464462215553966,
      "grad_norm": 9.513275146484375,
      "learning_rate": 4.298101903408386e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8469240546226502,
      "num_tokens": 134377457.0,
      "step": 111760
    },
    {
      "entropy": 1.8968846127390862,
      "epoch": 0.3464772206804463,
      "grad_norm": 4.018582344055176,
      "learning_rate": 4.297909622988691e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8519986256957054,
      "num_tokens": 134388907.0,
      "step": 111770
    },
    {
      "entropy": 1.9146791011095048,
      "epoch": 0.346508219805496,
      "grad_norm": 10.912361145019531,
      "learning_rate": 4.297717368372331e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8396982088685035,
      "num_tokens": 134400847.0,
      "step": 111780
    },
    {
      "entropy": 1.9298876509070397,
      "epoch": 0.3465392189305457,
      "grad_norm": 3.7877981662750244,
      "learning_rate": 4.2975251395535315e-06,
      "loss": 0.6157,
      "mean_token_accuracy": 0.8234444096684456,
      "num_tokens": 134412392.0,
      "step": 111790
    },
    {
      "entropy": 1.8450885564088821,
      "epoch": 0.3465702180555954,
      "grad_norm": 9.34116268157959,
      "learning_rate": 4.297332936526527e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8459417134523392,
      "num_tokens": 134424120.0,
      "step": 111800
    },
    {
      "entropy": 1.8955134615302085,
      "epoch": 0.34660121718064507,
      "grad_norm": 8.82641315460205,
      "learning_rate": 4.297140759285549e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.852502977848053,
      "num_tokens": 134435233.0,
      "step": 111810
    },
    {
      "entropy": 1.91280208081007,
      "epoch": 0.3466322163056948,
      "grad_norm": 7.556041240692139,
      "learning_rate": 4.296948607824833e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8468595892190933,
      "num_tokens": 134446764.0,
      "step": 111820
    },
    {
      "entropy": 1.7918421171605587,
      "epoch": 0.34666321543074446,
      "grad_norm": 8.250411033630371,
      "learning_rate": 4.296756482138616e-06,
      "loss": 0.4091,
      "mean_token_accuracy": 0.8426727816462517,
      "num_tokens": 134460455.0,
      "step": 111830
    },
    {
      "entropy": 1.8062309354543686,
      "epoch": 0.3466942145557942,
      "grad_norm": 8.455044746398926,
      "learning_rate": 4.2965643822211335e-06,
      "loss": 0.4249,
      "mean_token_accuracy": 0.8552202418446541,
      "num_tokens": 134473028.0,
      "step": 111840
    },
    {
      "entropy": 1.8862827733159064,
      "epoch": 0.34672521368084386,
      "grad_norm": 9.205554962158203,
      "learning_rate": 4.2963723080666284e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8518849700689316,
      "num_tokens": 134484150.0,
      "step": 111850
    },
    {
      "entropy": 1.874978469312191,
      "epoch": 0.3467562128058935,
      "grad_norm": 7.733843803405762,
      "learning_rate": 4.2961802596693425e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8496404230594635,
      "num_tokens": 134495656.0,
      "step": 111860
    },
    {
      "entropy": 1.7182648986577989,
      "epoch": 0.34678721193094325,
      "grad_norm": 3.7245407104492188,
      "learning_rate": 4.295988237023518e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.8628142789006233,
      "num_tokens": 134509703.0,
      "step": 111870
    },
    {
      "entropy": 1.8488290548324584,
      "epoch": 0.3468182110559929,
      "grad_norm": 9.35959529876709,
      "learning_rate": 4.295796240123402e-06,
      "loss": 0.485,
      "mean_token_accuracy": 0.8483680263161659,
      "num_tokens": 134522262.0,
      "step": 111880
    },
    {
      "entropy": 1.7668012008070946,
      "epoch": 0.34684921018104264,
      "grad_norm": 6.6548662185668945,
      "learning_rate": 4.2956042689632414e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8533818453550339,
      "num_tokens": 134534783.0,
      "step": 111890
    },
    {
      "entropy": 1.8939444333314897,
      "epoch": 0.3468802093060923,
      "grad_norm": 3.5856080055236816,
      "learning_rate": 4.295412323537284e-06,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8458682015538216,
      "num_tokens": 134545993.0,
      "step": 111900
    },
    {
      "entropy": 1.8058057472109794,
      "epoch": 0.34691120843114204,
      "grad_norm": 8.451927185058594,
      "learning_rate": 4.295220403839784e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8482358440756798,
      "num_tokens": 134558581.0,
      "step": 111910
    },
    {
      "entropy": 1.92232054322958,
      "epoch": 0.3469422075561917,
      "grad_norm": 9.112445831298828,
      "learning_rate": 4.29502850986499e-06,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8385997459292411,
      "num_tokens": 134570176.0,
      "step": 111920
    },
    {
      "entropy": 1.8965173974633216,
      "epoch": 0.34697320668124143,
      "grad_norm": 8.362327575683594,
      "learning_rate": 4.294836641607161e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8484830498695374,
      "num_tokens": 134581749.0,
      "step": 111930
    },
    {
      "entropy": 1.8018444642424583,
      "epoch": 0.3470042058062911,
      "grad_norm": 7.452921390533447,
      "learning_rate": 4.294644799060549e-06,
      "loss": 0.3927,
      "mean_token_accuracy": 0.861353749036789,
      "num_tokens": 134595028.0,
      "step": 111940
    },
    {
      "entropy": 1.8184639766812325,
      "epoch": 0.3470352049313408,
      "grad_norm": 6.740150451660156,
      "learning_rate": 4.2944529822194155e-06,
      "loss": 0.3941,
      "mean_token_accuracy": 0.8679483845829964,
      "num_tokens": 134607592.0,
      "step": 111950
    },
    {
      "entropy": 1.8601836189627647,
      "epoch": 0.3470662040563905,
      "grad_norm": 7.418848991394043,
      "learning_rate": 4.294261191078018e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.847012946009636,
      "num_tokens": 134619234.0,
      "step": 111960
    },
    {
      "entropy": 1.8946197971701622,
      "epoch": 0.3470972031814402,
      "grad_norm": 10.559837341308594,
      "learning_rate": 4.29406942563062e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8531604185700417,
      "num_tokens": 134630843.0,
      "step": 111970
    },
    {
      "entropy": 1.8593143001198769,
      "epoch": 0.3471282023064899,
      "grad_norm": 9.078370094299316,
      "learning_rate": 4.293877685871484e-06,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8558160826563835,
      "num_tokens": 134642923.0,
      "step": 111980
    },
    {
      "entropy": 1.8398296535015106,
      "epoch": 0.3471592014315396,
      "grad_norm": 7.763385772705078,
      "learning_rate": 4.293685971794876e-06,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8360516101121902,
      "num_tokens": 134655091.0,
      "step": 111990
    },
    {
      "entropy": 1.893106034398079,
      "epoch": 0.3471902005565893,
      "grad_norm": 9.48266315460205,
      "learning_rate": 4.293494283395064e-06,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8328988835215568,
      "num_tokens": 134666427.0,
      "step": 112000
    },
    {
      "entropy": 1.9177181079983712,
      "epoch": 0.347221199681639,
      "grad_norm": 7.473196029663086,
      "learning_rate": 4.293302620666314e-06,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8428963899612427,
      "num_tokens": 134678226.0,
      "step": 112010
    },
    {
      "entropy": 1.903357920050621,
      "epoch": 0.34725219880668867,
      "grad_norm": 8.634038925170898,
      "learning_rate": 4.293110983602899e-06,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8499746888875961,
      "num_tokens": 134688516.0,
      "step": 112020
    },
    {
      "entropy": 1.9164575353264808,
      "epoch": 0.3472831979317384,
      "grad_norm": 7.6614789962768555,
      "learning_rate": 4.29291937219909e-06,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8393423616886139,
      "num_tokens": 134699742.0,
      "step": 112030
    },
    {
      "entropy": 1.8141165480017662,
      "epoch": 0.34731419705678807,
      "grad_norm": 3.817478656768799,
      "learning_rate": 4.292727786449164e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8489150524139404,
      "num_tokens": 134713148.0,
      "step": 112040
    },
    {
      "entropy": 1.9213459312915802,
      "epoch": 0.3473451961818378,
      "grad_norm": 10.819091796875,
      "learning_rate": 4.292536226347394e-06,
      "loss": 0.5138,
      "mean_token_accuracy": 0.8452482402324677,
      "num_tokens": 134723649.0,
      "step": 112050
    },
    {
      "entropy": 1.8041829138994216,
      "epoch": 0.34737619530688746,
      "grad_norm": 2.2280683517456055,
      "learning_rate": 4.29234469188806e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8564143598079681,
      "num_tokens": 134736057.0,
      "step": 112060
    },
    {
      "entropy": 1.8665026307106019,
      "epoch": 0.3474071944319372,
      "grad_norm": 8.091485023498535,
      "learning_rate": 4.2921531830654395e-06,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8456338688731193,
      "num_tokens": 134746830.0,
      "step": 112070
    },
    {
      "entropy": 1.818691223859787,
      "epoch": 0.34743819355698685,
      "grad_norm": 9.684453964233398,
      "learning_rate": 4.291961699873817e-06,
      "loss": 0.4977,
      "mean_token_accuracy": 0.8366310462355614,
      "num_tokens": 134759170.0,
      "step": 112080
    },
    {
      "entropy": 1.8601401686668395,
      "epoch": 0.3474691926820366,
      "grad_norm": 8.760641098022461,
      "learning_rate": 4.291770242307472e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8496808990836143,
      "num_tokens": 134770727.0,
      "step": 112090
    },
    {
      "entropy": 1.782160858809948,
      "epoch": 0.34750019180708624,
      "grad_norm": 3.916487455368042,
      "learning_rate": 4.291578810360692e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.859810471534729,
      "num_tokens": 134783191.0,
      "step": 112100
    },
    {
      "entropy": 1.8653858050704002,
      "epoch": 0.3475311909321359,
      "grad_norm": 3.788681983947754,
      "learning_rate": 4.291387404027763e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8446184307336807,
      "num_tokens": 134794470.0,
      "step": 112110
    },
    {
      "entropy": 1.816742117702961,
      "epoch": 0.34756219005718564,
      "grad_norm": 3.5885984897613525,
      "learning_rate": 4.2911960233029745e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8555623233318329,
      "num_tokens": 134806055.0,
      "step": 112120
    },
    {
      "entropy": 1.817148557305336,
      "epoch": 0.3475931891822353,
      "grad_norm": 9.98969841003418,
      "learning_rate": 4.291004668180616e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8531207337975502,
      "num_tokens": 134817396.0,
      "step": 112130
    },
    {
      "entropy": 1.916732382774353,
      "epoch": 0.34762418830728503,
      "grad_norm": 9.270211219787598,
      "learning_rate": 4.290813338654979e-06,
      "loss": 0.5333,
      "mean_token_accuracy": 0.8368429586291313,
      "num_tokens": 134828575.0,
      "step": 112140
    },
    {
      "entropy": 1.8983269765973092,
      "epoch": 0.3476551874323347,
      "grad_norm": 8.712531089782715,
      "learning_rate": 4.2906220347203585e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8402269974350929,
      "num_tokens": 134840282.0,
      "step": 112150
    },
    {
      "entropy": 1.8531476333737373,
      "epoch": 0.3476861865573844,
      "grad_norm": 8.996459007263184,
      "learning_rate": 4.290430756371049e-06,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8387810304760933,
      "num_tokens": 134852480.0,
      "step": 112160
    },
    {
      "entropy": 1.9212437227368355,
      "epoch": 0.3477171856824341,
      "grad_norm": 7.855039119720459,
      "learning_rate": 4.290239503601349e-06,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8299373790621758,
      "num_tokens": 134864007.0,
      "step": 112170
    },
    {
      "entropy": 1.8462037175893784,
      "epoch": 0.3477481848074838,
      "grad_norm": 8.122658729553223,
      "learning_rate": 4.290048276405558e-06,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8515694797039032,
      "num_tokens": 134874756.0,
      "step": 112180
    },
    {
      "entropy": 1.8744586423039435,
      "epoch": 0.3477791839325335,
      "grad_norm": 8.11306381225586,
      "learning_rate": 4.289857074777977e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8434885591268539,
      "num_tokens": 134886296.0,
      "step": 112190
    },
    {
      "entropy": 1.8067267432808876,
      "epoch": 0.3478101830575832,
      "grad_norm": 4.103103160858154,
      "learning_rate": 4.289665898712908e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8561098381876946,
      "num_tokens": 134898467.0,
      "step": 112200
    },
    {
      "entropy": 1.9343467682600022,
      "epoch": 0.3478411821826329,
      "grad_norm": 6.8563923835754395,
      "learning_rate": 4.289474748204655e-06,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8526286870241165,
      "num_tokens": 134909189.0,
      "step": 112210
    },
    {
      "entropy": 1.8394694164395333,
      "epoch": 0.3478721813076826,
      "grad_norm": 12.269657135009766,
      "learning_rate": 4.289283623247527e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8569044172763824,
      "num_tokens": 134920985.0,
      "step": 112220
    },
    {
      "entropy": 1.7924383148550986,
      "epoch": 0.3479031804327323,
      "grad_norm": 3.9235920906066895,
      "learning_rate": 4.289092523835829e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8516478061676025,
      "num_tokens": 134934467.0,
      "step": 112230
    },
    {
      "entropy": 1.9101425260305405,
      "epoch": 0.347934179557782,
      "grad_norm": 8.595767974853516,
      "learning_rate": 4.288901449963873e-06,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8431054592132569,
      "num_tokens": 134945129.0,
      "step": 112240
    },
    {
      "entropy": 1.8743219137191773,
      "epoch": 0.34796517868283167,
      "grad_norm": 8.251272201538086,
      "learning_rate": 4.288710401625969e-06,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8410423472523689,
      "num_tokens": 134957099.0,
      "step": 112250
    },
    {
      "entropy": 1.8412858352065087,
      "epoch": 0.3479961778078814,
      "grad_norm": 7.280477046966553,
      "learning_rate": 4.2885193788164325e-06,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8605325534939766,
      "num_tokens": 134969309.0,
      "step": 112260
    },
    {
      "entropy": 1.8739237666130066,
      "epoch": 0.34802717693293106,
      "grad_norm": 7.483038902282715,
      "learning_rate": 4.288328381529578e-06,
      "loss": 0.4976,
      "mean_token_accuracy": 0.8409060567617417,
      "num_tokens": 134980937.0,
      "step": 112270
    },
    {
      "entropy": 1.7966715544462204,
      "epoch": 0.3480581760579808,
      "grad_norm": 3.68406081199646,
      "learning_rate": 4.288137409759721e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8603166654706002,
      "num_tokens": 134993822.0,
      "step": 112280
    },
    {
      "entropy": 1.9044744417071342,
      "epoch": 0.34808917518303045,
      "grad_norm": 9.015130043029785,
      "learning_rate": 4.287946463501182e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8462210029363633,
      "num_tokens": 135005423.0,
      "step": 112290
    },
    {
      "entropy": 1.8514644920825958,
      "epoch": 0.3481201743080802,
      "grad_norm": 4.599631309509277,
      "learning_rate": 4.287755542748281e-06,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8454376772046089,
      "num_tokens": 135017386.0,
      "step": 112300
    },
    {
      "entropy": 1.8641838699579238,
      "epoch": 0.34815117343312985,
      "grad_norm": 9.161344528198242,
      "learning_rate": 4.287564647495341e-06,
      "loss": 0.4695,
      "mean_token_accuracy": 0.841671122610569,
      "num_tokens": 135029362.0,
      "step": 112310
    },
    {
      "entropy": 1.8425541408360004,
      "epoch": 0.34818217255817957,
      "grad_norm": 8.259307861328125,
      "learning_rate": 4.287373777736684e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8496177807450295,
      "num_tokens": 135042307.0,
      "step": 112320
    },
    {
      "entropy": 1.8292279705405234,
      "epoch": 0.34821317168322924,
      "grad_norm": 7.943572044372559,
      "learning_rate": 4.287182933466639e-06,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8518629640340805,
      "num_tokens": 135054362.0,
      "step": 112330
    },
    {
      "entropy": 1.7942394033074378,
      "epoch": 0.34824417080827896,
      "grad_norm": 3.5839719772338867,
      "learning_rate": 4.286992114679531e-06,
      "loss": 0.3978,
      "mean_token_accuracy": 0.858057115972042,
      "num_tokens": 135067150.0,
      "step": 112340
    },
    {
      "entropy": 1.9045852065086364,
      "epoch": 0.34827516993332863,
      "grad_norm": 7.259175777435303,
      "learning_rate": 4.286801321369691e-06,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8487705394625664,
      "num_tokens": 135078189.0,
      "step": 112350
    },
    {
      "entropy": 1.8571503296494485,
      "epoch": 0.3483061690583783,
      "grad_norm": 9.252874374389648,
      "learning_rate": 4.286610553531448e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8501718282699585,
      "num_tokens": 135089770.0,
      "step": 112360
    },
    {
      "entropy": 1.8037226751446724,
      "epoch": 0.348337168183428,
      "grad_norm": 7.917973041534424,
      "learning_rate": 4.286419811159137e-06,
      "loss": 0.485,
      "mean_token_accuracy": 0.8473261296749115,
      "num_tokens": 135102849.0,
      "step": 112370
    },
    {
      "entropy": 1.8692201957106591,
      "epoch": 0.3483681673084777,
      "grad_norm": 4.2562575340271,
      "learning_rate": 4.286229094247093e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8484464436769485,
      "num_tokens": 135114862.0,
      "step": 112380
    },
    {
      "entropy": 1.7813385412096978,
      "epoch": 0.3483991664335274,
      "grad_norm": 8.083758354187012,
      "learning_rate": 4.28603840278965e-06,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8533005833625793,
      "num_tokens": 135127643.0,
      "step": 112390
    },
    {
      "entropy": 1.7049074426293374,
      "epoch": 0.3484301655585771,
      "grad_norm": 4.437235355377197,
      "learning_rate": 4.285847736781148e-06,
      "loss": 0.3249,
      "mean_token_accuracy": 0.871871954202652,
      "num_tokens": 135141468.0,
      "step": 112400
    },
    {
      "entropy": 1.8399584576487542,
      "epoch": 0.3484611646836268,
      "grad_norm": 7.485554218292236,
      "learning_rate": 4.285657096215928e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.850048904120922,
      "num_tokens": 135153861.0,
      "step": 112410
    },
    {
      "entropy": 1.8359861955046655,
      "epoch": 0.3484921638086765,
      "grad_norm": 4.102065563201904,
      "learning_rate": 4.285466481088329e-06,
      "loss": 0.5005,
      "mean_token_accuracy": 0.8472228407859802,
      "num_tokens": 135166609.0,
      "step": 112420
    },
    {
      "entropy": 1.8454686462879182,
      "epoch": 0.3485231629337262,
      "grad_norm": 7.877338409423828,
      "learning_rate": 4.2852758913926965e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.849373996257782,
      "num_tokens": 135179113.0,
      "step": 112430
    },
    {
      "entropy": 1.8713324323296547,
      "epoch": 0.3485541620587759,
      "grad_norm": 9.064533233642578,
      "learning_rate": 4.285085327123374e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8380694881081581,
      "num_tokens": 135191714.0,
      "step": 112440
    },
    {
      "entropy": 1.8374795719981194,
      "epoch": 0.3485851611838256,
      "grad_norm": 8.24960994720459,
      "learning_rate": 4.284894788274712e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8538482531905174,
      "num_tokens": 135203618.0,
      "step": 112450
    },
    {
      "entropy": 1.9125969141721726,
      "epoch": 0.34861616030887527,
      "grad_norm": 7.1157002449035645,
      "learning_rate": 4.284704274841055e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8587391778826714,
      "num_tokens": 135214093.0,
      "step": 112460
    },
    {
      "entropy": 1.7782380178570747,
      "epoch": 0.348647159433925,
      "grad_norm": 6.913045406341553,
      "learning_rate": 4.284513786816757e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.865745086967945,
      "num_tokens": 135227964.0,
      "step": 112470
    },
    {
      "entropy": 1.8960969477891922,
      "epoch": 0.34867815855897466,
      "grad_norm": 8.845172882080078,
      "learning_rate": 4.284323324196168e-06,
      "loss": 0.5353,
      "mean_token_accuracy": 0.8316770166158676,
      "num_tokens": 135238940.0,
      "step": 112480
    },
    {
      "entropy": 1.8000614181160928,
      "epoch": 0.3487091576840244,
      "grad_norm": 9.656888008117676,
      "learning_rate": 4.284132886973643e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8554758876562119,
      "num_tokens": 135251553.0,
      "step": 112490
    },
    {
      "entropy": 1.8248675152659417,
      "epoch": 0.34874015680907405,
      "grad_norm": 8.482026100158691,
      "learning_rate": 4.283942475143537e-06,
      "loss": 0.4344,
      "mean_token_accuracy": 0.8560914367437362,
      "num_tokens": 135263125.0,
      "step": 112500
    },
    {
      "entropy": 1.830253078043461,
      "epoch": 0.3487711559341238,
      "grad_norm": 7.03892183303833,
      "learning_rate": 4.283752088700209e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8495120465755462,
      "num_tokens": 135275581.0,
      "step": 112510
    },
    {
      "entropy": 1.840060842782259,
      "epoch": 0.34880215505917345,
      "grad_norm": 8.70134449005127,
      "learning_rate": 4.283561727638018e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8449655011296272,
      "num_tokens": 135287871.0,
      "step": 112520
    },
    {
      "entropy": 1.7875398755073548,
      "epoch": 0.34883315418422317,
      "grad_norm": 8.763019561767578,
      "learning_rate": 4.283371391951324e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.846840138733387,
      "num_tokens": 135300518.0,
      "step": 112530
    },
    {
      "entropy": 1.906669056415558,
      "epoch": 0.34886415330927284,
      "grad_norm": 8.401066780090332,
      "learning_rate": 4.2831810816344906e-06,
      "loss": 0.5297,
      "mean_token_accuracy": 0.8414530888199806,
      "num_tokens": 135311327.0,
      "step": 112540
    },
    {
      "entropy": 1.847528837621212,
      "epoch": 0.34889515243432256,
      "grad_norm": 9.117242813110352,
      "learning_rate": 4.282990796681881e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8574508696794509,
      "num_tokens": 135322556.0,
      "step": 112550
    },
    {
      "entropy": 1.8085899502038956,
      "epoch": 0.34892615155937223,
      "grad_norm": 8.368361473083496,
      "learning_rate": 4.282800537087866e-06,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8587646409869194,
      "num_tokens": 135335243.0,
      "step": 112560
    },
    {
      "entropy": 1.8577345311641693,
      "epoch": 0.34895715068442196,
      "grad_norm": 8.698378562927246,
      "learning_rate": 4.282610302846807e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8500236317515373,
      "num_tokens": 135346757.0,
      "step": 112570
    },
    {
      "entropy": 1.8168217539787292,
      "epoch": 0.3489881498094716,
      "grad_norm": 9.063081741333008,
      "learning_rate": 4.2824200939530796e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8485912069678306,
      "num_tokens": 135359508.0,
      "step": 112580
    },
    {
      "entropy": 1.8818160638213157,
      "epoch": 0.34901914893452135,
      "grad_norm": 9.063243865966797,
      "learning_rate": 4.282229910401052e-06,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8384617939591408,
      "num_tokens": 135371143.0,
      "step": 112590
    },
    {
      "entropy": 1.908516588807106,
      "epoch": 0.349050148059571,
      "grad_norm": 7.1072282791137695,
      "learning_rate": 4.282039752185099e-06,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8510788649320602,
      "num_tokens": 135382304.0,
      "step": 112600
    },
    {
      "entropy": 1.9243893340229987,
      "epoch": 0.3490811471846207,
      "grad_norm": 8.661938667297363,
      "learning_rate": 4.2818496192995955e-06,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8423077017068863,
      "num_tokens": 135393382.0,
      "step": 112610
    },
    {
      "entropy": 1.9478312745690345,
      "epoch": 0.3491121463096704,
      "grad_norm": 8.19333267211914,
      "learning_rate": 4.281659511738918e-06,
      "loss": 0.545,
      "mean_token_accuracy": 0.8402326017618179,
      "num_tokens": 135404664.0,
      "step": 112620
    },
    {
      "entropy": 1.9101561456918716,
      "epoch": 0.3491431454347201,
      "grad_norm": 11.05017375946045,
      "learning_rate": 4.281469429497445e-06,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8389782354235649,
      "num_tokens": 135416146.0,
      "step": 112630
    },
    {
      "entropy": 1.8043391317129136,
      "epoch": 0.3491741445597698,
      "grad_norm": 9.59979248046875,
      "learning_rate": 4.281279372569557e-06,
      "loss": 0.4637,
      "mean_token_accuracy": 0.8521313741803169,
      "num_tokens": 135428371.0,
      "step": 112640
    },
    {
      "entropy": 1.822686144709587,
      "epoch": 0.3492051436848195,
      "grad_norm": 9.630875587463379,
      "learning_rate": 4.281089340949636e-06,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8514738500118255,
      "num_tokens": 135440601.0,
      "step": 112650
    },
    {
      "entropy": 1.844109094142914,
      "epoch": 0.3492361428098692,
      "grad_norm": 3.6397085189819336,
      "learning_rate": 4.280899334632067e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8515435203909874,
      "num_tokens": 135452180.0,
      "step": 112660
    },
    {
      "entropy": 1.7884308710694312,
      "epoch": 0.34926714193491887,
      "grad_norm": 8.593812942504883,
      "learning_rate": 4.280709353611234e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8537843570113182,
      "num_tokens": 135465795.0,
      "step": 112670
    },
    {
      "entropy": 1.7283026084303856,
      "epoch": 0.3492981410599686,
      "grad_norm": 8.442497253417969,
      "learning_rate": 4.280519397881524e-06,
      "loss": 0.3646,
      "mean_token_accuracy": 0.8640974462032318,
      "num_tokens": 135478919.0,
      "step": 112680
    },
    {
      "entropy": 1.8432815566658973,
      "epoch": 0.34932914018501826,
      "grad_norm": 8.06016731262207,
      "learning_rate": 4.280329467437327e-06,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8606173783540726,
      "num_tokens": 135490424.0,
      "step": 112690
    },
    {
      "entropy": 1.821248809993267,
      "epoch": 0.349360139310068,
      "grad_norm": 7.616511821746826,
      "learning_rate": 4.280139562273034e-06,
      "loss": 0.4417,
      "mean_token_accuracy": 0.847284109890461,
      "num_tokens": 135503303.0,
      "step": 112700
    },
    {
      "entropy": 1.8346032842993736,
      "epoch": 0.34939113843511765,
      "grad_norm": 8.359513282775879,
      "learning_rate": 4.279949682383039e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8445788651704789,
      "num_tokens": 135514959.0,
      "step": 112710
    },
    {
      "entropy": 1.8606849431991577,
      "epoch": 0.3494221375601674,
      "grad_norm": 9.360651016235352,
      "learning_rate": 4.279759827761733e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8447454944252968,
      "num_tokens": 135525909.0,
      "step": 112720
    },
    {
      "entropy": 1.8562923952937127,
      "epoch": 0.34945313668521705,
      "grad_norm": 7.43868350982666,
      "learning_rate": 4.279569998403512e-06,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8330693423748017,
      "num_tokens": 135537774.0,
      "step": 112730
    },
    {
      "entropy": 1.7586598336696624,
      "epoch": 0.34948413581026677,
      "grad_norm": 4.424074649810791,
      "learning_rate": 4.279380194302777e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8576010316610336,
      "num_tokens": 135550831.0,
      "step": 112740
    },
    {
      "entropy": 1.7588186517357827,
      "epoch": 0.34951513493531644,
      "grad_norm": 8.668641090393066,
      "learning_rate": 4.279190415453926e-06,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8586042076349258,
      "num_tokens": 135564325.0,
      "step": 112750
    },
    {
      "entropy": 1.814384798705578,
      "epoch": 0.34954613406036616,
      "grad_norm": 8.124059677124023,
      "learning_rate": 4.279000661851359e-06,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8556953683495522,
      "num_tokens": 135576775.0,
      "step": 112760
    },
    {
      "entropy": 1.8137044474482535,
      "epoch": 0.34957713318541583,
      "grad_norm": 9.291952133178711,
      "learning_rate": 4.278810933489481e-06,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8497354611754417,
      "num_tokens": 135589390.0,
      "step": 112770
    },
    {
      "entropy": 1.8097498089075088,
      "epoch": 0.34960813231046556,
      "grad_norm": 7.762622356414795,
      "learning_rate": 4.278621230362695e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8576091334223748,
      "num_tokens": 135602123.0,
      "step": 112780
    },
    {
      "entropy": 1.9097516283392906,
      "epoch": 0.3496391314355152,
      "grad_norm": 8.101143836975098,
      "learning_rate": 4.278431552465409e-06,
      "loss": 0.522,
      "mean_token_accuracy": 0.8374461770057678,
      "num_tokens": 135613911.0,
      "step": 112790
    },
    {
      "entropy": 1.8759268373250961,
      "epoch": 0.34967013056056495,
      "grad_norm": 8.917213439941406,
      "learning_rate": 4.27824189979203e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8517908856272698,
      "num_tokens": 135625727.0,
      "step": 112800
    },
    {
      "entropy": 1.8602227076888085,
      "epoch": 0.3497011296856146,
      "grad_norm": 5.823835372924805,
      "learning_rate": 4.278052272336967e-06,
      "loss": 0.4806,
      "mean_token_accuracy": 0.8478166356682777,
      "num_tokens": 135637869.0,
      "step": 112810
    },
    {
      "entropy": 1.921775433421135,
      "epoch": 0.34973212881066434,
      "grad_norm": 10.301258087158203,
      "learning_rate": 4.2778626700946335e-06,
      "loss": 0.5161,
      "mean_token_accuracy": 0.8499249458312989,
      "num_tokens": 135648405.0,
      "step": 112820
    },
    {
      "entropy": 1.877712282538414,
      "epoch": 0.349763127935714,
      "grad_norm": 7.107439994812012,
      "learning_rate": 4.2776730930594425e-06,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8331371307373047,
      "num_tokens": 135660457.0,
      "step": 112830
    },
    {
      "entropy": 1.83882287889719,
      "epoch": 0.34979412706076374,
      "grad_norm": 12.444573402404785,
      "learning_rate": 4.277483541225809e-06,
      "loss": 0.5465,
      "mean_token_accuracy": 0.8394935175776481,
      "num_tokens": 135672611.0,
      "step": 112840
    },
    {
      "entropy": 1.8228483900427819,
      "epoch": 0.3498251261858134,
      "grad_norm": 7.8412933349609375,
      "learning_rate": 4.27729401458815e-06,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8546199440956116,
      "num_tokens": 135684136.0,
      "step": 112850
    },
    {
      "entropy": 1.8391767397522927,
      "epoch": 0.3498561253108631,
      "grad_norm": 8.248656272888184,
      "learning_rate": 4.2771045131408834e-06,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8546992540359497,
      "num_tokens": 135696190.0,
      "step": 112860
    },
    {
      "entropy": 1.8091487675905227,
      "epoch": 0.3498871244359128,
      "grad_norm": 3.6982574462890625,
      "learning_rate": 4.2769150368784295e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8531347304582596,
      "num_tokens": 135708890.0,
      "step": 112870
    },
    {
      "entropy": 1.7924383908510209,
      "epoch": 0.34991812356096247,
      "grad_norm": 8.61406135559082,
      "learning_rate": 4.2767255857952115e-06,
      "loss": 0.4149,
      "mean_token_accuracy": 0.8567244291305542,
      "num_tokens": 135721688.0,
      "step": 112880
    },
    {
      "entropy": 1.850319354236126,
      "epoch": 0.3499491226860122,
      "grad_norm": 7.656332015991211,
      "learning_rate": 4.2765361598856534e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8434787318110466,
      "num_tokens": 135733821.0,
      "step": 112890
    },
    {
      "entropy": 1.8255574628710747,
      "epoch": 0.34998012181106186,
      "grad_norm": 3.7060792446136475,
      "learning_rate": 4.276346759144178e-06,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8535738855600357,
      "num_tokens": 135745431.0,
      "step": 112900
    },
    {
      "entropy": 1.8570487424731255,
      "epoch": 0.3500111209361116,
      "grad_norm": 8.941734313964844,
      "learning_rate": 4.276157383565215e-06,
      "loss": 0.488,
      "mean_token_accuracy": 0.8438050165772438,
      "num_tokens": 135757174.0,
      "step": 112910
    },
    {
      "entropy": 1.683679609745741,
      "epoch": 0.35004212006116125,
      "grad_norm": 2.426884174346924,
      "learning_rate": 4.2759680331431915e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.8656910866498947,
      "num_tokens": 135772071.0,
      "step": 112920
    },
    {
      "entropy": 1.8811020210385323,
      "epoch": 0.350073119186211,
      "grad_norm": 9.89200210571289,
      "learning_rate": 4.275778707872541e-06,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8509863570332528,
      "num_tokens": 135783363.0,
      "step": 112930
    },
    {
      "entropy": 1.8929930627346039,
      "epoch": 0.35010411831126065,
      "grad_norm": 8.002644538879395,
      "learning_rate": 4.2755894077476925e-06,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8419990673661232,
      "num_tokens": 135794454.0,
      "step": 112940
    },
    {
      "entropy": 1.8019443243741988,
      "epoch": 0.3501351174363104,
      "grad_norm": 7.989790916442871,
      "learning_rate": 4.275400132763083e-06,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8515826135873794,
      "num_tokens": 135806472.0,
      "step": 112950
    },
    {
      "entropy": 1.8456357419490814,
      "epoch": 0.35016611656136004,
      "grad_norm": 4.0877556800842285,
      "learning_rate": 4.275210882913148e-06,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8436346635222435,
      "num_tokens": 135818253.0,
      "step": 112960
    },
    {
      "entropy": 1.8550100848078728,
      "epoch": 0.35019711568640977,
      "grad_norm": 9.118851661682129,
      "learning_rate": 4.275021658192323e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8482651144266129,
      "num_tokens": 135829973.0,
      "step": 112970
    },
    {
      "entropy": 1.8620183497667313,
      "epoch": 0.35022811481145943,
      "grad_norm": 12.924306869506836,
      "learning_rate": 4.274832458595049e-06,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8375033006072045,
      "num_tokens": 135840885.0,
      "step": 112980
    },
    {
      "entropy": 1.7792857438325882,
      "epoch": 0.35025911393650916,
      "grad_norm": 7.036866664886475,
      "learning_rate": 4.274643284115767e-06,
      "loss": 0.4054,
      "mean_token_accuracy": 0.848195219039917,
      "num_tokens": 135853598.0,
      "step": 112990
    },
    {
      "entropy": 1.8005570635199546,
      "epoch": 0.3502901130615588,
      "grad_norm": 8.605463027954102,
      "learning_rate": 4.274454134748919e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8483079835772515,
      "num_tokens": 135866796.0,
      "step": 113000
    },
    {
      "entropy": 1.867513844370842,
      "epoch": 0.35032111218660855,
      "grad_norm": 8.543618202209473,
      "learning_rate": 4.2742650104889496e-06,
      "loss": 0.4927,
      "mean_token_accuracy": 0.8421450614929199,
      "num_tokens": 135878432.0,
      "step": 113010
    },
    {
      "entropy": 1.7870407178997993,
      "epoch": 0.3503521113116582,
      "grad_norm": 8.509590148925781,
      "learning_rate": 4.274075911330306e-06,
      "loss": 0.437,
      "mean_token_accuracy": 0.8518823325634003,
      "num_tokens": 135891001.0,
      "step": 113020
    },
    {
      "entropy": 1.8838384568691253,
      "epoch": 0.35038311043670795,
      "grad_norm": 8.782331466674805,
      "learning_rate": 4.273886837267435e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8412837624549866,
      "num_tokens": 135902263.0,
      "step": 113030
    },
    {
      "entropy": 1.7536726333200932,
      "epoch": 0.3504141095617576,
      "grad_norm": 4.015054702758789,
      "learning_rate": 4.2736977882947855e-06,
      "loss": 0.3985,
      "mean_token_accuracy": 0.8617945462465286,
      "num_tokens": 135915709.0,
      "step": 113040
    },
    {
      "entropy": 1.8604958072304725,
      "epoch": 0.35044510868680734,
      "grad_norm": 7.668582916259766,
      "learning_rate": 4.273508764406811e-06,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8354014694690705,
      "num_tokens": 135928153.0,
      "step": 113050
    },
    {
      "entropy": 1.7598976030945779,
      "epoch": 0.350476107811857,
      "grad_norm": 10.459297180175781,
      "learning_rate": 4.273319765597964e-06,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8666064769029618,
      "num_tokens": 135940924.0,
      "step": 113060
    },
    {
      "entropy": 1.925863367319107,
      "epoch": 0.35050710693690673,
      "grad_norm": 4.1394853591918945,
      "learning_rate": 4.273130791862697e-06,
      "loss": 0.5217,
      "mean_token_accuracy": 0.839854308962822,
      "num_tokens": 135952237.0,
      "step": 113070
    },
    {
      "entropy": 1.8460562735795976,
      "epoch": 0.3505381060619564,
      "grad_norm": 8.900185585021973,
      "learning_rate": 4.27294184319547e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8506451055407525,
      "num_tokens": 135964374.0,
      "step": 113080
    },
    {
      "entropy": 1.8739062339067458,
      "epoch": 0.35056910518700607,
      "grad_norm": Infinity,
      "learning_rate": 4.272752919590739e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8531906709074975,
      "num_tokens": 135974841.0,
      "step": 113090
    },
    {
      "entropy": 1.9152967289090157,
      "epoch": 0.3506001043120558,
      "grad_norm": 9.287792205810547,
      "learning_rate": 4.272564021042964e-06,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8334060996770859,
      "num_tokens": 135986242.0,
      "step": 113100
    },
    {
      "entropy": 1.779229535162449,
      "epoch": 0.35063110343710546,
      "grad_norm": 8.452899932861328,
      "learning_rate": 4.272375147546608e-06,
      "loss": 0.3765,
      "mean_token_accuracy": 0.8608766749501229,
      "num_tokens": 135998810.0,
      "step": 113110
    },
    {
      "entropy": 1.9012344628572464,
      "epoch": 0.3506621025621552,
      "grad_norm": 8.408564567565918,
      "learning_rate": 4.272186299096133e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8417745620012284,
      "num_tokens": 136009754.0,
      "step": 113120
    },
    {
      "entropy": 1.8403314396739006,
      "epoch": 0.35069310168720486,
      "grad_norm": 10.097549438476562,
      "learning_rate": 4.271997475686004e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8540255069732666,
      "num_tokens": 136021001.0,
      "step": 113130
    },
    {
      "entropy": 1.8598705619573592,
      "epoch": 0.3507241008122546,
      "grad_norm": 10.163576126098633,
      "learning_rate": 4.2718086773106895e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8600896030664444,
      "num_tokens": 136032506.0,
      "step": 113140
    },
    {
      "entropy": 1.8539843708276749,
      "epoch": 0.35075509993730425,
      "grad_norm": 9.255050659179688,
      "learning_rate": 4.271619903964656e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8301028832793236,
      "num_tokens": 136044950.0,
      "step": 113150
    },
    {
      "entropy": 1.7757720202207565,
      "epoch": 0.350786099062354,
      "grad_norm": 9.801958084106445,
      "learning_rate": 4.271431155642374e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8526146367192269,
      "num_tokens": 136057364.0,
      "step": 113160
    },
    {
      "entropy": 1.818256576359272,
      "epoch": 0.35081709818740364,
      "grad_norm": 8.34134578704834,
      "learning_rate": 4.271242432338316e-06,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8509904339909553,
      "num_tokens": 136069897.0,
      "step": 113170
    },
    {
      "entropy": 1.7901548087596892,
      "epoch": 0.35084809731245337,
      "grad_norm": 10.111919403076172,
      "learning_rate": 4.271053734046957e-06,
      "loss": 0.428,
      "mean_token_accuracy": 0.8516396522521973,
      "num_tokens": 136081662.0,
      "step": 113180
    },
    {
      "entropy": 1.8030595824122428,
      "epoch": 0.35087909643750304,
      "grad_norm": 8.223822593688965,
      "learning_rate": 4.270865060762769e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8596414580941201,
      "num_tokens": 136094054.0,
      "step": 113190
    },
    {
      "entropy": 1.8742952436208724,
      "epoch": 0.35091009556255276,
      "grad_norm": 7.283956050872803,
      "learning_rate": 4.270676412480232e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8499576464295387,
      "num_tokens": 136104584.0,
      "step": 113200
    },
    {
      "entropy": 1.8488594472408295,
      "epoch": 0.35094109468760243,
      "grad_norm": 4.605078220367432,
      "learning_rate": 4.270487789193823e-06,
      "loss": 0.5479,
      "mean_token_accuracy": 0.829779888689518,
      "num_tokens": 136116396.0,
      "step": 113210
    },
    {
      "entropy": 1.7764787435531617,
      "epoch": 0.35097209381265215,
      "grad_norm": 4.046032905578613,
      "learning_rate": 4.270299190898024e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8495251774787903,
      "num_tokens": 136129027.0,
      "step": 113220
    },
    {
      "entropy": 1.8619268134236335,
      "epoch": 0.3510030929377018,
      "grad_norm": 7.183846473693848,
      "learning_rate": 4.2701106175873156e-06,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8402719959616661,
      "num_tokens": 136141038.0,
      "step": 113230
    },
    {
      "entropy": 1.8102800309658051,
      "epoch": 0.35103409206275155,
      "grad_norm": 3.5368010997772217,
      "learning_rate": 4.2699220692561825e-06,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8623706936836243,
      "num_tokens": 136153341.0,
      "step": 113240
    },
    {
      "entropy": 1.9274267673492431,
      "epoch": 0.3510650911878012,
      "grad_norm": 6.329407215118408,
      "learning_rate": 4.26973354589911e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8464403167366982,
      "num_tokens": 136164572.0,
      "step": 113250
    },
    {
      "entropy": 1.859284907579422,
      "epoch": 0.35109609031285094,
      "grad_norm": 9.37317943572998,
      "learning_rate": 4.2695450475105855e-06,
      "loss": 0.5012,
      "mean_token_accuracy": 0.8453105211257934,
      "num_tokens": 136176202.0,
      "step": 113260
    },
    {
      "entropy": 1.8950539276003837,
      "epoch": 0.3511270894379006,
      "grad_norm": 8.385150909423828,
      "learning_rate": 4.269356574085098e-06,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8488032355904579,
      "num_tokens": 136187307.0,
      "step": 113270
    },
    {
      "entropy": 1.844504640996456,
      "epoch": 0.35115808856295033,
      "grad_norm": 3.8115599155426025,
      "learning_rate": 4.269168125617139e-06,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8605356469750405,
      "num_tokens": 136198798.0,
      "step": 113280
    },
    {
      "entropy": 1.8304932430386542,
      "epoch": 0.351189087688,
      "grad_norm": 6.604245185852051,
      "learning_rate": 4.2689797021012e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8543907523155212,
      "num_tokens": 136211146.0,
      "step": 113290
    },
    {
      "entropy": 1.7669322073459626,
      "epoch": 0.3512200868130497,
      "grad_norm": 9.710320472717285,
      "learning_rate": 4.268791303531774e-06,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8585976973176003,
      "num_tokens": 136223754.0,
      "step": 113300
    },
    {
      "entropy": 1.8966532975435257,
      "epoch": 0.3512510859380994,
      "grad_norm": 7.919690132141113,
      "learning_rate": 4.268602929903359e-06,
      "loss": 0.5226,
      "mean_token_accuracy": 0.8372878462076188,
      "num_tokens": 136234793.0,
      "step": 113310
    },
    {
      "entropy": 1.8168953225016593,
      "epoch": 0.3512820850631491,
      "grad_norm": 9.95252513885498,
      "learning_rate": 4.26841458121045e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.8458817571401596,
      "num_tokens": 136247165.0,
      "step": 113320
    },
    {
      "entropy": 1.8963631451129914,
      "epoch": 0.3513130841881988,
      "grad_norm": 8.91224193572998,
      "learning_rate": 4.26822625744755e-06,
      "loss": 0.5239,
      "mean_token_accuracy": 0.83456601947546,
      "num_tokens": 136259113.0,
      "step": 113330
    },
    {
      "entropy": 1.7711772859096526,
      "epoch": 0.35134408331324846,
      "grad_norm": 8.007668495178223,
      "learning_rate": 4.268037958609155e-06,
      "loss": 0.3735,
      "mean_token_accuracy": 0.8632699027657509,
      "num_tokens": 136272310.0,
      "step": 113340
    },
    {
      "entropy": 1.9127058327198028,
      "epoch": 0.3513750824382982,
      "grad_norm": 8.543416023254395,
      "learning_rate": 4.267849684689771e-06,
      "loss": 0.5151,
      "mean_token_accuracy": 0.8420788779854774,
      "num_tokens": 136283081.0,
      "step": 113350
    },
    {
      "entropy": 1.8306966736912726,
      "epoch": 0.35140608156334785,
      "grad_norm": 7.711248397827148,
      "learning_rate": 4.267661435683903e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8500320836901665,
      "num_tokens": 136295002.0,
      "step": 113360
    },
    {
      "entropy": 1.9049995988607407,
      "epoch": 0.3514370806883976,
      "grad_norm": 9.507410049438477,
      "learning_rate": 4.267473211586053e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8459584578871727,
      "num_tokens": 136305515.0,
      "step": 113370
    },
    {
      "entropy": 1.91897811293602,
      "epoch": 0.35146807981344724,
      "grad_norm": 8.550996780395508,
      "learning_rate": 4.267285012390732e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8397913321852684,
      "num_tokens": 136316379.0,
      "step": 113380
    },
    {
      "entropy": 1.837944434583187,
      "epoch": 0.35149907893849697,
      "grad_norm": 9.132309913635254,
      "learning_rate": 4.267096838092449e-06,
      "loss": 0.4261,
      "mean_token_accuracy": 0.8522770449519157,
      "num_tokens": 136328473.0,
      "step": 113390
    },
    {
      "entropy": 1.8238879337906837,
      "epoch": 0.35153007806354664,
      "grad_norm": 8.676041603088379,
      "learning_rate": 4.266908688685714e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8456574112176896,
      "num_tokens": 136340510.0,
      "step": 113400
    },
    {
      "entropy": 1.8029537141323089,
      "epoch": 0.35156107718859636,
      "grad_norm": 8.039926528930664,
      "learning_rate": 4.26672056416504e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8499104723334312,
      "num_tokens": 136352921.0,
      "step": 113410
    },
    {
      "entropy": 1.8220291912555695,
      "epoch": 0.35159207631364603,
      "grad_norm": 3.9269118309020996,
      "learning_rate": 4.2665324645249425e-06,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8560845017433166,
      "num_tokens": 136365546.0,
      "step": 113420
    },
    {
      "entropy": 1.8378960967063904,
      "epoch": 0.35162307543869575,
      "grad_norm": 7.433810234069824,
      "learning_rate": 4.266344389759935e-06,
      "loss": 0.4517,
      "mean_token_accuracy": 0.8511560037732124,
      "num_tokens": 136377851.0,
      "step": 113430
    },
    {
      "entropy": 1.838972160220146,
      "epoch": 0.3516540745637454,
      "grad_norm": 8.42202091217041,
      "learning_rate": 4.2661563398645395e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8631669625639915,
      "num_tokens": 136389500.0,
      "step": 113440
    },
    {
      "entropy": 1.9216330140829085,
      "epoch": 0.35168507368879515,
      "grad_norm": 8.987174987792969,
      "learning_rate": 4.2659683148332716e-06,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8396774441003799,
      "num_tokens": 136400087.0,
      "step": 113450
    },
    {
      "entropy": 1.814877037703991,
      "epoch": 0.3517160728138448,
      "grad_norm": 4.2126688957214355,
      "learning_rate": 4.265780314660655e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8484560027718544,
      "num_tokens": 136412750.0,
      "step": 113460
    },
    {
      "entropy": 1.675320317596197,
      "epoch": 0.35174707193889454,
      "grad_norm": 7.873843669891357,
      "learning_rate": 4.265592339341211e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.861806321144104,
      "num_tokens": 136428096.0,
      "step": 113470
    },
    {
      "entropy": 1.7690242916345595,
      "epoch": 0.3517780710639442,
      "grad_norm": 8.084175109863281,
      "learning_rate": 4.265404388869465e-06,
      "loss": 0.4054,
      "mean_token_accuracy": 0.8658703714609146,
      "num_tokens": 136440575.0,
      "step": 113480
    },
    {
      "entropy": 1.853781743347645,
      "epoch": 0.35180907018899393,
      "grad_norm": 3.831251621246338,
      "learning_rate": 4.265216463239944e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8527114719152451,
      "num_tokens": 136452228.0,
      "step": 113490
    },
    {
      "entropy": 1.7241961359977722,
      "epoch": 0.3518400693140436,
      "grad_norm": 3.6539437770843506,
      "learning_rate": 4.265028562447174e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8586121633648872,
      "num_tokens": 136465487.0,
      "step": 113500
    },
    {
      "entropy": 1.763987348973751,
      "epoch": 0.3518710684390933,
      "grad_norm": 9.397258758544922,
      "learning_rate": 4.264840686485687e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8535782873630524,
      "num_tokens": 136477731.0,
      "step": 113510
    },
    {
      "entropy": 1.821959725022316,
      "epoch": 0.351902067564143,
      "grad_norm": 3.157595634460449,
      "learning_rate": 4.264652835350013e-06,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8425436571240426,
      "num_tokens": 136489410.0,
      "step": 113520
    },
    {
      "entropy": 1.8831856340169906,
      "epoch": 0.3519330666891927,
      "grad_norm": 6.910709857940674,
      "learning_rate": 4.264465009034684e-06,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8598052114248276,
      "num_tokens": 136500743.0,
      "step": 113530
    },
    {
      "entropy": 1.9352434664964675,
      "epoch": 0.3519640658142424,
      "grad_norm": 11.542353630065918,
      "learning_rate": 4.264277207534237e-06,
      "loss": 0.5419,
      "mean_token_accuracy": 0.8231759518384933,
      "num_tokens": 136511423.0,
      "step": 113540
    },
    {
      "entropy": 1.719076856970787,
      "epoch": 0.3519950649392921,
      "grad_norm": 8.80350112915039,
      "learning_rate": 4.264089430843207e-06,
      "loss": 0.3627,
      "mean_token_accuracy": 0.8623046159744263,
      "num_tokens": 136525514.0,
      "step": 113550
    },
    {
      "entropy": 1.8064083874225616,
      "epoch": 0.3520260640643418,
      "grad_norm": 9.935619354248047,
      "learning_rate": 4.263901678956134e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.85721445530653,
      "num_tokens": 136537364.0,
      "step": 113560
    },
    {
      "entropy": 1.696343556046486,
      "epoch": 0.3520570631893915,
      "grad_norm": 3.945061445236206,
      "learning_rate": 4.263713951867554e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8635579511523247,
      "num_tokens": 136551460.0,
      "step": 113570
    },
    {
      "entropy": 1.8621020391583443,
      "epoch": 0.3520880623144412,
      "grad_norm": 10.784354209899902,
      "learning_rate": 4.263526249572011e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8448983728885651,
      "num_tokens": 136562725.0,
      "step": 113580
    },
    {
      "entropy": 1.8034757658839227,
      "epoch": 0.35211906143949084,
      "grad_norm": 7.975634574890137,
      "learning_rate": 4.263338572064049e-06,
      "loss": 0.4426,
      "mean_token_accuracy": 0.8480996668338776,
      "num_tokens": 136574947.0,
      "step": 113590
    },
    {
      "entropy": 1.7639338225126266,
      "epoch": 0.35215006056454057,
      "grad_norm": 2.267158031463623,
      "learning_rate": 4.263150919338211e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8544442996382713,
      "num_tokens": 136587812.0,
      "step": 113600
    },
    {
      "entropy": 1.7771714597940445,
      "epoch": 0.35218105968959024,
      "grad_norm": 4.19359016418457,
      "learning_rate": 4.262963291389045e-06,
      "loss": 0.4269,
      "mean_token_accuracy": 0.851497220993042,
      "num_tokens": 136601096.0,
      "step": 113610
    },
    {
      "entropy": 1.8166782572865485,
      "epoch": 0.35221205881463996,
      "grad_norm": 4.370350360870361,
      "learning_rate": 4.262775688211097e-06,
      "loss": 0.4615,
      "mean_token_accuracy": 0.8466724216938019,
      "num_tokens": 136613550.0,
      "step": 113620
    },
    {
      "entropy": 1.8014537960290908,
      "epoch": 0.35224305793968963,
      "grad_norm": 7.764684677124023,
      "learning_rate": 4.262588109798919e-06,
      "loss": 0.4062,
      "mean_token_accuracy": 0.8574507102370262,
      "num_tokens": 136626300.0,
      "step": 113630
    },
    {
      "entropy": 1.7783242926001548,
      "epoch": 0.35227405706473935,
      "grad_norm": 7.457787036895752,
      "learning_rate": 4.262400556147062e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8619847133755684,
      "num_tokens": 136638878.0,
      "step": 113640
    },
    {
      "entropy": 1.7466833159327506,
      "epoch": 0.352305056189789,
      "grad_norm": 4.234191417694092,
      "learning_rate": 4.2622130272500775e-06,
      "loss": 0.4237,
      "mean_token_accuracy": 0.8527040392160415,
      "num_tokens": 136652119.0,
      "step": 113650
    },
    {
      "entropy": 1.8685536488890648,
      "epoch": 0.35233605531483875,
      "grad_norm": 8.825045585632324,
      "learning_rate": 4.262025523102523e-06,
      "loss": 0.5118,
      "mean_token_accuracy": 0.8355319261550903,
      "num_tokens": 136663383.0,
      "step": 113660
    },
    {
      "entropy": 1.88068146109581,
      "epoch": 0.3523670544398884,
      "grad_norm": 10.0654296875,
      "learning_rate": 4.261838043698953e-06,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8337692707777024,
      "num_tokens": 136674512.0,
      "step": 113670
    },
    {
      "entropy": 1.8553010821342468,
      "epoch": 0.35239805356493814,
      "grad_norm": 8.771764755249023,
      "learning_rate": 4.2616505890339275e-06,
      "loss": 0.4976,
      "mean_token_accuracy": 0.8464409559965134,
      "num_tokens": 136686081.0,
      "step": 113680
    },
    {
      "entropy": 1.8780170038342476,
      "epoch": 0.3524290526899878,
      "grad_norm": 7.942883014678955,
      "learning_rate": 4.261463159102005e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8412024825811386,
      "num_tokens": 136697359.0,
      "step": 113690
    },
    {
      "entropy": 1.82506482899189,
      "epoch": 0.35246005181503753,
      "grad_norm": 8.61766242980957,
      "learning_rate": 4.261275753897748e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8486872628331185,
      "num_tokens": 136709869.0,
      "step": 113700
    },
    {
      "entropy": 1.8322357684373856,
      "epoch": 0.3524910509400872,
      "grad_norm": 3.983318328857422,
      "learning_rate": 4.261088373415719e-06,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8400922358036041,
      "num_tokens": 136721994.0,
      "step": 113710
    },
    {
      "entropy": 1.8604241654276847,
      "epoch": 0.3525220500651369,
      "grad_norm": 7.393711090087891,
      "learning_rate": 4.260901017650485e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8476014077663422,
      "num_tokens": 136732793.0,
      "step": 113720
    },
    {
      "entropy": 1.8291943043470382,
      "epoch": 0.3525530491901866,
      "grad_norm": 8.577530860900879,
      "learning_rate": 4.26071368659661e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8571088537573814,
      "num_tokens": 136744850.0,
      "step": 113730
    },
    {
      "entropy": 1.810156986117363,
      "epoch": 0.3525840483152363,
      "grad_norm": 7.944915294647217,
      "learning_rate": 4.260526380248662e-06,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8527490749955178,
      "num_tokens": 136756757.0,
      "step": 113740
    },
    {
      "entropy": 1.8879317745566369,
      "epoch": 0.352615047440286,
      "grad_norm": 9.205636978149414,
      "learning_rate": 4.260339098601214e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8410181164741516,
      "num_tokens": 136768280.0,
      "step": 113750
    },
    {
      "entropy": 1.8330311551690102,
      "epoch": 0.3526460465653357,
      "grad_norm": 4.422030448913574,
      "learning_rate": 4.2601518416488344e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8590439051389694,
      "num_tokens": 136779826.0,
      "step": 113760
    },
    {
      "entropy": 1.8848232328891754,
      "epoch": 0.3526770456903854,
      "grad_norm": 8.074309349060059,
      "learning_rate": 4.259964609386099e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8494037300348282,
      "num_tokens": 136791347.0,
      "step": 113770
    },
    {
      "entropy": 1.862524962425232,
      "epoch": 0.3527080448154351,
      "grad_norm": 9.903372764587402,
      "learning_rate": 4.2597774018075815e-06,
      "loss": 0.4728,
      "mean_token_accuracy": 0.8448254838585854,
      "num_tokens": 136803506.0,
      "step": 113780
    },
    {
      "entropy": 1.787963542342186,
      "epoch": 0.3527390439404848,
      "grad_norm": 9.371294975280762,
      "learning_rate": 4.259590218907858e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8571085095405578,
      "num_tokens": 136816790.0,
      "step": 113790
    },
    {
      "entropy": 1.8950905337929727,
      "epoch": 0.3527700430655345,
      "grad_norm": 4.0718159675598145,
      "learning_rate": 4.259403060681509e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8508754774928093,
      "num_tokens": 136827775.0,
      "step": 113800
    },
    {
      "entropy": 1.8704029634594916,
      "epoch": 0.35280104219058417,
      "grad_norm": 8.028072357177734,
      "learning_rate": 4.259215927123112e-06,
      "loss": 0.477,
      "mean_token_accuracy": 0.8462028250098228,
      "num_tokens": 136839928.0,
      "step": 113810
    },
    {
      "entropy": 1.7776376932859421,
      "epoch": 0.3528320413156339,
      "grad_norm": 7.8810529708862305,
      "learning_rate": 4.25902881822725e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8655279219150543,
      "num_tokens": 136852948.0,
      "step": 113820
    },
    {
      "entropy": 1.886681966483593,
      "epoch": 0.35286304044068356,
      "grad_norm": 9.677278518676758,
      "learning_rate": 4.2588417339885055e-06,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8440056949853897,
      "num_tokens": 136864663.0,
      "step": 113830
    },
    {
      "entropy": 1.860078476369381,
      "epoch": 0.35289403956573323,
      "grad_norm": 7.314884185791016,
      "learning_rate": 4.258654674401464e-06,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8441386088728905,
      "num_tokens": 136875913.0,
      "step": 113840
    },
    {
      "entropy": 1.856418649852276,
      "epoch": 0.35292503869078296,
      "grad_norm": 6.926784515380859,
      "learning_rate": 4.258467639460713e-06,
      "loss": 0.4665,
      "mean_token_accuracy": 0.8466035217046738,
      "num_tokens": 136888046.0,
      "step": 113850
    },
    {
      "entropy": 1.8510633751749992,
      "epoch": 0.3529560378158326,
      "grad_norm": 8.287582397460938,
      "learning_rate": 4.258280629160838e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8363550141453743,
      "num_tokens": 136900496.0,
      "step": 113860
    },
    {
      "entropy": 1.8378751188516618,
      "epoch": 0.35298703694088235,
      "grad_norm": 7.225609302520752,
      "learning_rate": 4.258093643496433e-06,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8513509348034859,
      "num_tokens": 136912943.0,
      "step": 113870
    },
    {
      "entropy": 1.8136658892035484,
      "epoch": 0.353018036065932,
      "grad_norm": 8.231322288513184,
      "learning_rate": 4.257906682462087e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8516188710927963,
      "num_tokens": 136925555.0,
      "step": 113880
    },
    {
      "entropy": 1.8374749287962913,
      "epoch": 0.35304903519098174,
      "grad_norm": 7.933830738067627,
      "learning_rate": 4.257719746052393e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8512130409479142,
      "num_tokens": 136937326.0,
      "step": 113890
    },
    {
      "entropy": 1.8850289553403854,
      "epoch": 0.3530800343160314,
      "grad_norm": 9.631787300109863,
      "learning_rate": 4.257532834261947e-06,
      "loss": 0.5034,
      "mean_token_accuracy": 0.8466032981872559,
      "num_tokens": 136947938.0,
      "step": 113900
    },
    {
      "entropy": 1.774746771156788,
      "epoch": 0.35311103344108113,
      "grad_norm": 3.474881172180176,
      "learning_rate": 4.257345947085346e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8617874652147293,
      "num_tokens": 136960564.0,
      "step": 113910
    },
    {
      "entropy": 1.8160261258482933,
      "epoch": 0.3531420325661308,
      "grad_norm": 8.455029487609863,
      "learning_rate": 4.257159084517186e-06,
      "loss": 0.4386,
      "mean_token_accuracy": 0.8601742401719094,
      "num_tokens": 136972540.0,
      "step": 113920
    },
    {
      "entropy": 1.8941467106342316,
      "epoch": 0.35317303169118053,
      "grad_norm": 4.235901355743408,
      "learning_rate": 4.25697224655207e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8378585115075111,
      "num_tokens": 136984312.0,
      "step": 113930
    },
    {
      "entropy": 1.8767295882105828,
      "epoch": 0.3532040308162302,
      "grad_norm": 9.515031814575195,
      "learning_rate": 4.256785433184598e-06,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8369233354926109,
      "num_tokens": 136995625.0,
      "step": 113940
    },
    {
      "entropy": 1.81492610424757,
      "epoch": 0.3532350299412799,
      "grad_norm": 4.470407485961914,
      "learning_rate": 4.256598644409373e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8580437764525414,
      "num_tokens": 137008213.0,
      "step": 113950
    },
    {
      "entropy": 1.849051834642887,
      "epoch": 0.3532660290663296,
      "grad_norm": 7.72014045715332,
      "learning_rate": 4.2564118802210006e-06,
      "loss": 0.4815,
      "mean_token_accuracy": 0.8522540688514709,
      "num_tokens": 137020112.0,
      "step": 113960
    },
    {
      "entropy": 1.8805166974663734,
      "epoch": 0.3532970281913793,
      "grad_norm": 4.944484710693359,
      "learning_rate": 4.256225140614086e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8436195507645607,
      "num_tokens": 137031603.0,
      "step": 113970
    },
    {
      "entropy": 1.874251839518547,
      "epoch": 0.353328027316429,
      "grad_norm": 8.50593090057373,
      "learning_rate": 4.25603842558324e-06,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8450019791722297,
      "num_tokens": 137043366.0,
      "step": 113980
    },
    {
      "entropy": 1.8801655188202857,
      "epoch": 0.3533590264414787,
      "grad_norm": 8.969855308532715,
      "learning_rate": 4.25585173512307e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8407104894518852,
      "num_tokens": 137054687.0,
      "step": 113990
    },
    {
      "entropy": 1.8593452662229537,
      "epoch": 0.3533900255665284,
      "grad_norm": 8.071375846862793,
      "learning_rate": 4.255665069228188e-06,
      "loss": 0.4094,
      "mean_token_accuracy": 0.8621448844671249,
      "num_tokens": 137066396.0,
      "step": 114000
    },
    {
      "entropy": 1.8180650487542152,
      "epoch": 0.3534210246915781,
      "grad_norm": 8.56116771697998,
      "learning_rate": 4.255478427893209e-06,
      "loss": 0.4487,
      "mean_token_accuracy": 0.8478880062699318,
      "num_tokens": 137077708.0,
      "step": 114010
    },
    {
      "entropy": 1.815793578326702,
      "epoch": 0.35345202381662777,
      "grad_norm": 9.843114852905273,
      "learning_rate": 4.255291811112745e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8555766463279724,
      "num_tokens": 137090459.0,
      "step": 114020
    },
    {
      "entropy": 1.8414412125945092,
      "epoch": 0.3534830229416775,
      "grad_norm": 7.240721225738525,
      "learning_rate": 4.255105218881416e-06,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8408783197402954,
      "num_tokens": 137101852.0,
      "step": 114030
    },
    {
      "entropy": 1.8683375775814057,
      "epoch": 0.35351402206672716,
      "grad_norm": 7.452057361602783,
      "learning_rate": 4.2549186511938356e-06,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8472641706466675,
      "num_tokens": 137113521.0,
      "step": 114040
    },
    {
      "entropy": 1.9232849180698395,
      "epoch": 0.3535450211917769,
      "grad_norm": 8.561758995056152,
      "learning_rate": 4.254732108044627e-06,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8424664407968521,
      "num_tokens": 137124822.0,
      "step": 114050
    },
    {
      "entropy": 1.8383488565683366,
      "epoch": 0.35357602031682656,
      "grad_norm": 10.501875877380371,
      "learning_rate": 4.254545589428411e-06,
      "loss": 0.4763,
      "mean_token_accuracy": 0.836837200820446,
      "num_tokens": 137136758.0,
      "step": 114060
    },
    {
      "entropy": 1.873065246641636,
      "epoch": 0.3536070194418763,
      "grad_norm": 9.702674865722656,
      "learning_rate": 4.254359095339811e-06,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8503431528806686,
      "num_tokens": 137148091.0,
      "step": 114070
    },
    {
      "entropy": 1.8993804574012756,
      "epoch": 0.35363801856692595,
      "grad_norm": 7.872958183288574,
      "learning_rate": 4.254172625773451e-06,
      "loss": 0.469,
      "mean_token_accuracy": 0.849186685681343,
      "num_tokens": 137159064.0,
      "step": 114080
    },
    {
      "entropy": 1.9097682803869247,
      "epoch": 0.3536690176919756,
      "grad_norm": 8.155599594116211,
      "learning_rate": 4.253986180723957e-06,
      "loss": 0.4959,
      "mean_token_accuracy": 0.8419617161154747,
      "num_tokens": 137170143.0,
      "step": 114090
    },
    {
      "entropy": 1.8655310958623885,
      "epoch": 0.35370001681702534,
      "grad_norm": 10.33460807800293,
      "learning_rate": 4.253799760185957e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8362808287143707,
      "num_tokens": 137181569.0,
      "step": 114100
    },
    {
      "entropy": 1.7499642267823219,
      "epoch": 0.353731015942075,
      "grad_norm": 3.961226224899292,
      "learning_rate": 4.253613364154082e-06,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8528372913599014,
      "num_tokens": 137194428.0,
      "step": 114110
    },
    {
      "entropy": 1.8577033221721648,
      "epoch": 0.35376201506712474,
      "grad_norm": 8.127702713012695,
      "learning_rate": 4.2534269926229625e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8494561642408371,
      "num_tokens": 137206496.0,
      "step": 114120
    },
    {
      "entropy": 1.950427946448326,
      "epoch": 0.3537930141921744,
      "grad_norm": 7.7237043380737305,
      "learning_rate": 4.253240645587232e-06,
      "loss": 0.5027,
      "mean_token_accuracy": 0.847864530980587,
      "num_tokens": 137217616.0,
      "step": 114130
    },
    {
      "entropy": 1.8073005706071854,
      "epoch": 0.35382401331722413,
      "grad_norm": 8.634578704833984,
      "learning_rate": 4.253054323041525e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8615914180874824,
      "num_tokens": 137229842.0,
      "step": 114140
    },
    {
      "entropy": 1.9044538155198096,
      "epoch": 0.3538550124422738,
      "grad_norm": 10.229740142822266,
      "learning_rate": 4.252868024980477e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.840241327881813,
      "num_tokens": 137241062.0,
      "step": 114150
    },
    {
      "entropy": 1.7597592651844025,
      "epoch": 0.3538860115673235,
      "grad_norm": 9.922557830810547,
      "learning_rate": 4.252681751398727e-06,
      "loss": 0.3942,
      "mean_token_accuracy": 0.8564444318413734,
      "num_tokens": 137254220.0,
      "step": 114160
    },
    {
      "entropy": 1.9016540557146073,
      "epoch": 0.3539170106923732,
      "grad_norm": 10.562433242797852,
      "learning_rate": 4.252495502290913e-06,
      "loss": 0.5396,
      "mean_token_accuracy": 0.832036292552948,
      "num_tokens": 137266517.0,
      "step": 114170
    },
    {
      "entropy": 1.9116821497678758,
      "epoch": 0.3539480098174229,
      "grad_norm": 8.220035552978516,
      "learning_rate": 4.252309277651677e-06,
      "loss": 0.511,
      "mean_token_accuracy": 0.8397183999419212,
      "num_tokens": 137277540.0,
      "step": 114180
    },
    {
      "entropy": 1.8361477851867676,
      "epoch": 0.3539790089424726,
      "grad_norm": 8.83745288848877,
      "learning_rate": 4.252123077475664e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8428474217653275,
      "num_tokens": 137289242.0,
      "step": 114190
    },
    {
      "entropy": 1.8084876164793968,
      "epoch": 0.3540100080675223,
      "grad_norm": 7.766282081604004,
      "learning_rate": 4.251936901757515e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8553501293063164,
      "num_tokens": 137301999.0,
      "step": 114200
    },
    {
      "entropy": 1.8107165440917015,
      "epoch": 0.354041007192572,
      "grad_norm": 4.7136993408203125,
      "learning_rate": 4.251750750491878e-06,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8484808370471001,
      "num_tokens": 137314968.0,
      "step": 114210
    },
    {
      "entropy": 1.94393610060215,
      "epoch": 0.3540720063176217,
      "grad_norm": 9.238444328308105,
      "learning_rate": 4.2515646236734e-06,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8350920498371124,
      "num_tokens": 137325889.0,
      "step": 114220
    },
    {
      "entropy": 1.782192386686802,
      "epoch": 0.35410300544267137,
      "grad_norm": 9.513643264770508,
      "learning_rate": 4.251378521296731e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8538872867822647,
      "num_tokens": 137338800.0,
      "step": 114230
    },
    {
      "entropy": 1.830018326640129,
      "epoch": 0.3541340045677211,
      "grad_norm": 7.8421759605407715,
      "learning_rate": 4.251192443356523e-06,
      "loss": 0.468,
      "mean_token_accuracy": 0.843048295378685,
      "num_tokens": 137350877.0,
      "step": 114240
    },
    {
      "entropy": 1.8395485177636146,
      "epoch": 0.35416500369277076,
      "grad_norm": 8.42922592163086,
      "learning_rate": 4.251006389847427e-06,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8559403479099273,
      "num_tokens": 137362938.0,
      "step": 114250
    },
    {
      "entropy": 1.8116194292902947,
      "epoch": 0.3541960028178205,
      "grad_norm": 4.232943534851074,
      "learning_rate": 4.250820360764097e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.847702045738697,
      "num_tokens": 137375890.0,
      "step": 114260
    },
    {
      "entropy": 1.699189220368862,
      "epoch": 0.35422700194287016,
      "grad_norm": 6.99352502822876,
      "learning_rate": 4.25063435610119e-06,
      "loss": 0.3176,
      "mean_token_accuracy": 0.8732634499669075,
      "num_tokens": 137389460.0,
      "step": 114270
    },
    {
      "entropy": 1.7898903474211694,
      "epoch": 0.3542580010679199,
      "grad_norm": 9.566137313842773,
      "learning_rate": 4.250448375853365e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8538245901465416,
      "num_tokens": 137402759.0,
      "step": 114280
    },
    {
      "entropy": 1.8570916399359703,
      "epoch": 0.35428900019296955,
      "grad_norm": 3.8215625286102295,
      "learning_rate": 4.250262420015279e-06,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8420146137475968,
      "num_tokens": 137414428.0,
      "step": 114290
    },
    {
      "entropy": 1.8959661558270455,
      "epoch": 0.3543199993180193,
      "grad_norm": 7.246954441070557,
      "learning_rate": 4.250076488581593e-06,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8415970131754875,
      "num_tokens": 137425829.0,
      "step": 114300
    },
    {
      "entropy": 1.8812722265720367,
      "epoch": 0.35435099844306894,
      "grad_norm": 7.860531330108643,
      "learning_rate": 4.24989058154697e-06,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8447592079639434,
      "num_tokens": 137437268.0,
      "step": 114310
    },
    {
      "entropy": 1.8524739906191825,
      "epoch": 0.35438199756811867,
      "grad_norm": 8.29738712310791,
      "learning_rate": 4.2497046989060754e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8458985432982444,
      "num_tokens": 137449877.0,
      "step": 114320
    },
    {
      "entropy": 1.8565015509724616,
      "epoch": 0.35441299669316834,
      "grad_norm": 8.345024108886719,
      "learning_rate": 4.2495188406535735e-06,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8513003289699554,
      "num_tokens": 137461527.0,
      "step": 114330
    },
    {
      "entropy": 1.8561938509345055,
      "epoch": 0.354443995818218,
      "grad_norm": 9.380087852478027,
      "learning_rate": 4.249333006784131e-06,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8521606922149658,
      "num_tokens": 137473760.0,
      "step": 114340
    },
    {
      "entropy": 1.8369782000780106,
      "epoch": 0.35447499494326773,
      "grad_norm": 8.772756576538086,
      "learning_rate": 4.249147197292419e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8478844106197357,
      "num_tokens": 137485545.0,
      "step": 114350
    },
    {
      "entropy": 1.9263944923877716,
      "epoch": 0.3545059940683174,
      "grad_norm": 7.329461574554443,
      "learning_rate": 4.248961412173107e-06,
      "loss": 0.5178,
      "mean_token_accuracy": 0.8424314111471176,
      "num_tokens": 137496566.0,
      "step": 114360
    },
    {
      "entropy": 1.8391064286231995,
      "epoch": 0.3545369931933671,
      "grad_norm": 8.21514892578125,
      "learning_rate": 4.2487756514208675e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8564809501171112,
      "num_tokens": 137508491.0,
      "step": 114370
    },
    {
      "entropy": 1.8533219590783119,
      "epoch": 0.3545679923184168,
      "grad_norm": 8.260367393493652,
      "learning_rate": 4.248589915030374e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8540527150034904,
      "num_tokens": 137520622.0,
      "step": 114380
    },
    {
      "entropy": 1.9417014926671983,
      "epoch": 0.3545989914434665,
      "grad_norm": 8.334314346313477,
      "learning_rate": 4.248404202996303e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8378186166286469,
      "num_tokens": 137531376.0,
      "step": 114390
    },
    {
      "entropy": 1.8331923067569733,
      "epoch": 0.3546299905685162,
      "grad_norm": 9.672205924987793,
      "learning_rate": 4.248218515313331e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8539971187710762,
      "num_tokens": 137543888.0,
      "step": 114400
    },
    {
      "entropy": 1.8398851856589318,
      "epoch": 0.3546609896935659,
      "grad_norm": 8.222822189331055,
      "learning_rate": 4.248032851976136e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8429914176464081,
      "num_tokens": 137555582.0,
      "step": 114410
    },
    {
      "entropy": 1.817541065812111,
      "epoch": 0.3546919888186156,
      "grad_norm": 7.350632190704346,
      "learning_rate": 4.2478472129794004e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8614602595567703,
      "num_tokens": 137567761.0,
      "step": 114420
    },
    {
      "entropy": 1.8555574625730515,
      "epoch": 0.3547229879436653,
      "grad_norm": 4.468617916107178,
      "learning_rate": 4.247661598317806e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8440041437745094,
      "num_tokens": 137579603.0,
      "step": 114430
    },
    {
      "entropy": 1.8921550765633584,
      "epoch": 0.35475398706871497,
      "grad_norm": 9.684183120727539,
      "learning_rate": 4.247476007986034e-06,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8349121674895287,
      "num_tokens": 137591460.0,
      "step": 114440
    },
    {
      "entropy": 1.9489626735448837,
      "epoch": 0.3547849861937647,
      "grad_norm": 7.472236633300781,
      "learning_rate": 4.247290441978772e-06,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8391944885253906,
      "num_tokens": 137602126.0,
      "step": 114450
    },
    {
      "entropy": 1.8100664362311363,
      "epoch": 0.35481598531881436,
      "grad_norm": 7.485744476318359,
      "learning_rate": 4.247104900290708e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8578644886612892,
      "num_tokens": 137614832.0,
      "step": 114460
    },
    {
      "entropy": 1.8753397777676581,
      "epoch": 0.3548469844438641,
      "grad_norm": 7.755117893218994,
      "learning_rate": 4.246919382916528e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8451044067740441,
      "num_tokens": 137626735.0,
      "step": 114470
    },
    {
      "entropy": 1.851782961189747,
      "epoch": 0.35487798356891376,
      "grad_norm": 8.35659408569336,
      "learning_rate": 4.2467338898509225e-06,
      "loss": 0.456,
      "mean_token_accuracy": 0.8530733227729798,
      "num_tokens": 137638707.0,
      "step": 114480
    },
    {
      "entropy": 1.8803992569446564,
      "epoch": 0.3549089826939635,
      "grad_norm": 9.96238899230957,
      "learning_rate": 4.2465484210885845e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.8485779255628586,
      "num_tokens": 137650395.0,
      "step": 114490
    },
    {
      "entropy": 1.8529548197984695,
      "epoch": 0.35493998181901315,
      "grad_norm": 7.451599597930908,
      "learning_rate": 4.2463629766242074e-06,
      "loss": 0.4776,
      "mean_token_accuracy": 0.8496254831552505,
      "num_tokens": 137661248.0,
      "step": 114500
    },
    {
      "entropy": 1.8691770792007447,
      "epoch": 0.3549709809440629,
      "grad_norm": 8.7528657913208,
      "learning_rate": 4.246177556452486e-06,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8437319055199624,
      "num_tokens": 137672985.0,
      "step": 114510
    },
    {
      "entropy": 1.8243165865540505,
      "epoch": 0.35500198006911254,
      "grad_norm": 7.32008171081543,
      "learning_rate": 4.245992160568117e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.841618250310421,
      "num_tokens": 137684795.0,
      "step": 114520
    },
    {
      "entropy": 1.8710746631026267,
      "epoch": 0.35503297919416227,
      "grad_norm": 8.433467864990234,
      "learning_rate": 4.245806788965798e-06,
      "loss": 0.5185,
      "mean_token_accuracy": 0.8461954742670059,
      "num_tokens": 137696877.0,
      "step": 114530
    },
    {
      "entropy": 1.9102421700954437,
      "epoch": 0.35506397831921194,
      "grad_norm": 8.911306381225586,
      "learning_rate": 4.245621441640229e-06,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8445306703448295,
      "num_tokens": 137708323.0,
      "step": 114540
    },
    {
      "entropy": 1.704357886314392,
      "epoch": 0.35509497744426166,
      "grad_norm": 5.612151145935059,
      "learning_rate": 4.245436118586114e-06,
      "loss": 0.3443,
      "mean_token_accuracy": 0.8622689947485924,
      "num_tokens": 137722546.0,
      "step": 114550
    },
    {
      "entropy": 1.8894066840410233,
      "epoch": 0.35512597656931133,
      "grad_norm": 3.8784425258636475,
      "learning_rate": 4.245250819798153e-06,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8490568831562996,
      "num_tokens": 137733876.0,
      "step": 114560
    },
    {
      "entropy": 1.7194712564349175,
      "epoch": 0.355156975694361,
      "grad_norm": 9.469301223754883,
      "learning_rate": 4.2450655452710515e-06,
      "loss": 0.4478,
      "mean_token_accuracy": 0.8465792283415794,
      "num_tokens": 137747364.0,
      "step": 114570
    },
    {
      "entropy": 1.7580340534448624,
      "epoch": 0.3551879748194107,
      "grad_norm": 8.43166732788086,
      "learning_rate": 4.244880294999517e-06,
      "loss": 0.4017,
      "mean_token_accuracy": 0.8530274391174316,
      "num_tokens": 137760169.0,
      "step": 114580
    },
    {
      "entropy": 1.798786661028862,
      "epoch": 0.3552189739444604,
      "grad_norm": 3.7261557579040527,
      "learning_rate": 4.2446950689782575e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8502959847450257,
      "num_tokens": 137772888.0,
      "step": 114590
    },
    {
      "entropy": 1.782324093580246,
      "epoch": 0.3552499730695101,
      "grad_norm": 3.953874349594116,
      "learning_rate": 4.244509867201982e-06,
      "loss": 0.397,
      "mean_token_accuracy": 0.8616974368691445,
      "num_tokens": 137786038.0,
      "step": 114600
    },
    {
      "entropy": 1.8710459485650062,
      "epoch": 0.3552809721945598,
      "grad_norm": 7.7761101722717285,
      "learning_rate": 4.244324689665401e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8453879669308663,
      "num_tokens": 137797978.0,
      "step": 114610
    },
    {
      "entropy": 1.8599458813667298,
      "epoch": 0.3553119713196095,
      "grad_norm": 7.9371747970581055,
      "learning_rate": 4.244139536363229e-06,
      "loss": 0.4237,
      "mean_token_accuracy": 0.8475207060575485,
      "num_tokens": 137810153.0,
      "step": 114620
    },
    {
      "entropy": 1.8876738995313644,
      "epoch": 0.3553429704446592,
      "grad_norm": 9.171344757080078,
      "learning_rate": 4.24395440729018e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.8501772657036781,
      "num_tokens": 137822116.0,
      "step": 114630
    },
    {
      "entropy": 1.91156704723835,
      "epoch": 0.3553739695697089,
      "grad_norm": 8.98309326171875,
      "learning_rate": 4.2437693024409685e-06,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8459622815251351,
      "num_tokens": 137833476.0,
      "step": 114640
    },
    {
      "entropy": 1.855041791498661,
      "epoch": 0.35540496869475857,
      "grad_norm": 7.716914176940918,
      "learning_rate": 4.243584221810315e-06,
      "loss": 0.4479,
      "mean_token_accuracy": 0.8591550663113594,
      "num_tokens": 137845518.0,
      "step": 114650
    },
    {
      "entropy": 1.9073261603713036,
      "epoch": 0.3554359678198083,
      "grad_norm": 7.420741558074951,
      "learning_rate": 4.243399165392936e-06,
      "loss": 0.5075,
      "mean_token_accuracy": 0.842662687599659,
      "num_tokens": 137856959.0,
      "step": 114660
    },
    {
      "entropy": 1.9126224562525749,
      "epoch": 0.35546696694485796,
      "grad_norm": 8.25358772277832,
      "learning_rate": 4.243214133183554e-06,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8389646530151367,
      "num_tokens": 137867845.0,
      "step": 114670
    },
    {
      "entropy": 1.908442348241806,
      "epoch": 0.3554979660699077,
      "grad_norm": 3.6674983501434326,
      "learning_rate": 4.243029125176892e-06,
      "loss": 0.4823,
      "mean_token_accuracy": 0.8443290576338768,
      "num_tokens": 137879689.0,
      "step": 114680
    },
    {
      "entropy": 1.7568098783493042,
      "epoch": 0.35552896519495736,
      "grad_norm": 3.361020088195801,
      "learning_rate": 4.242844141367674e-06,
      "loss": 0.389,
      "mean_token_accuracy": 0.8680830702185631,
      "num_tokens": 137892854.0,
      "step": 114690
    },
    {
      "entropy": 1.8590233013033868,
      "epoch": 0.3555599643200071,
      "grad_norm": 8.2853422164917,
      "learning_rate": 4.242659181750625e-06,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8496688306331635,
      "num_tokens": 137904751.0,
      "step": 114700
    },
    {
      "entropy": 1.8025938093662262,
      "epoch": 0.35559096344505675,
      "grad_norm": 6.7893595695495605,
      "learning_rate": 4.242474246320471e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8490458041429519,
      "num_tokens": 137917185.0,
      "step": 114710
    },
    {
      "entropy": 1.8439732417464256,
      "epoch": 0.3556219625701065,
      "grad_norm": 4.053720951080322,
      "learning_rate": 4.2422893350719436e-06,
      "loss": 0.455,
      "mean_token_accuracy": 0.8445042937994003,
      "num_tokens": 137929018.0,
      "step": 114720
    },
    {
      "entropy": 1.8289795368909836,
      "epoch": 0.35565296169515614,
      "grad_norm": 7.735437393188477,
      "learning_rate": 4.2421044479997735e-06,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8526295185089111,
      "num_tokens": 137941602.0,
      "step": 114730
    },
    {
      "entropy": 1.7583549529314042,
      "epoch": 0.35568396082020587,
      "grad_norm": 9.589922904968262,
      "learning_rate": 4.24191958509869e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8651691839098931,
      "num_tokens": 137954921.0,
      "step": 114740
    },
    {
      "entropy": 1.848554477095604,
      "epoch": 0.35571495994525554,
      "grad_norm": 8.365562438964844,
      "learning_rate": 4.24173474636343e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8425380006432533,
      "num_tokens": 137965949.0,
      "step": 114750
    },
    {
      "entropy": 1.9080881744623184,
      "epoch": 0.35574595907030526,
      "grad_norm": 8.790996551513672,
      "learning_rate": 4.241549931788727e-06,
      "loss": 0.507,
      "mean_token_accuracy": 0.8438135430216789,
      "num_tokens": 137976920.0,
      "step": 114760
    },
    {
      "entropy": 1.9018929034471512,
      "epoch": 0.35577695819535493,
      "grad_norm": 8.04598617553711,
      "learning_rate": 4.2413651413693185e-06,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8404053092002869,
      "num_tokens": 137988280.0,
      "step": 114770
    },
    {
      "entropy": 1.8129638865590096,
      "epoch": 0.35580795732040466,
      "grad_norm": 3.7647767066955566,
      "learning_rate": 4.241180375099945e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8564304232597351,
      "num_tokens": 138001123.0,
      "step": 114780
    },
    {
      "entropy": 1.821030892431736,
      "epoch": 0.3558389564454543,
      "grad_norm": 3.7905266284942627,
      "learning_rate": 4.240995632975342e-06,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8544037476181984,
      "num_tokens": 138012902.0,
      "step": 114790
    },
    {
      "entropy": 1.8608674079179763,
      "epoch": 0.35586995557050405,
      "grad_norm": 8.616358757019043,
      "learning_rate": 4.240810914990257e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8417997151613236,
      "num_tokens": 138025108.0,
      "step": 114800
    },
    {
      "entropy": 1.8812236726284026,
      "epoch": 0.3559009546955537,
      "grad_norm": 8.76230239868164,
      "learning_rate": 4.240626221139429e-06,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8435042083263398,
      "num_tokens": 138037279.0,
      "step": 114810
    },
    {
      "entropy": 1.7783361107110978,
      "epoch": 0.3559319538206034,
      "grad_norm": 3.4879698753356934,
      "learning_rate": 4.240441551417605e-06,
      "loss": 0.3984,
      "mean_token_accuracy": 0.8596335366368294,
      "num_tokens": 138050823.0,
      "step": 114820
    },
    {
      "entropy": 1.7827171936631203,
      "epoch": 0.3559629529456531,
      "grad_norm": 3.9974160194396973,
      "learning_rate": 4.240256905819533e-06,
      "loss": 0.3823,
      "mean_token_accuracy": 0.8612391158938408,
      "num_tokens": 138064142.0,
      "step": 114830
    },
    {
      "entropy": 1.8113099455833435,
      "epoch": 0.3559939520707028,
      "grad_norm": 7.9462666511535645,
      "learning_rate": 4.2400722843399585e-06,
      "loss": 0.4029,
      "mean_token_accuracy": 0.8514809906482697,
      "num_tokens": 138076703.0,
      "step": 114840
    },
    {
      "entropy": 1.818100643157959,
      "epoch": 0.3560249511957525,
      "grad_norm": 8.276450157165527,
      "learning_rate": 4.2398876869736325e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8574807778000831,
      "num_tokens": 138088592.0,
      "step": 114850
    },
    {
      "entropy": 1.899727213382721,
      "epoch": 0.3560559503208022,
      "grad_norm": 8.389613151550293,
      "learning_rate": 4.239703113715307e-06,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8366639405488968,
      "num_tokens": 138099369.0,
      "step": 114860
    },
    {
      "entropy": 1.8642989337444305,
      "epoch": 0.3560869494458519,
      "grad_norm": 9.097145080566406,
      "learning_rate": 4.239518564559734e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8484190404415131,
      "num_tokens": 138111261.0,
      "step": 114870
    },
    {
      "entropy": 1.8235100641846658,
      "epoch": 0.35611794857090157,
      "grad_norm": 8.840439796447754,
      "learning_rate": 4.239334039501668e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8622523456811905,
      "num_tokens": 138124180.0,
      "step": 114880
    },
    {
      "entropy": 1.8383688524365425,
      "epoch": 0.3561489476959513,
      "grad_norm": 5.670360088348389,
      "learning_rate": 4.2391495385358675e-06,
      "loss": 0.4512,
      "mean_token_accuracy": 0.8513874992728233,
      "num_tokens": 138136327.0,
      "step": 114890
    },
    {
      "entropy": 1.8067172005772592,
      "epoch": 0.35617994682100096,
      "grad_norm": 9.207831382751465,
      "learning_rate": 4.238965061657087e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8570899114012718,
      "num_tokens": 138148730.0,
      "step": 114900
    },
    {
      "entropy": 1.917934286594391,
      "epoch": 0.3562109459460507,
      "grad_norm": 8.583123207092285,
      "learning_rate": 4.238780608860088e-06,
      "loss": 0.5284,
      "mean_token_accuracy": 0.8423309728503228,
      "num_tokens": 138159657.0,
      "step": 114910
    },
    {
      "entropy": 1.8137140288949012,
      "epoch": 0.35624194507110035,
      "grad_norm": 7.209733009338379,
      "learning_rate": 4.238596180139632e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8488231867551803,
      "num_tokens": 138172455.0,
      "step": 114920
    },
    {
      "entropy": 1.8409941777586938,
      "epoch": 0.3562729441961501,
      "grad_norm": 7.371655464172363,
      "learning_rate": 4.238411775490481e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8461980625987053,
      "num_tokens": 138184789.0,
      "step": 114930
    },
    {
      "entropy": 1.8432543560862542,
      "epoch": 0.35630394332119975,
      "grad_norm": 8.352115631103516,
      "learning_rate": 4.238227394907398e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8550651222467422,
      "num_tokens": 138196321.0,
      "step": 114940
    },
    {
      "entropy": 1.880720390379429,
      "epoch": 0.35633494244624947,
      "grad_norm": 7.84929084777832,
      "learning_rate": 4.23804303838515e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.8451797798275947,
      "num_tokens": 138207824.0,
      "step": 114950
    },
    {
      "entropy": 1.8087940603494643,
      "epoch": 0.35636594157129914,
      "grad_norm": 9.440030097961426,
      "learning_rate": 4.237858705918504e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.8487761154770851,
      "num_tokens": 138220985.0,
      "step": 114960
    },
    {
      "entropy": 1.875385396182537,
      "epoch": 0.35639694069634886,
      "grad_norm": 7.9317474365234375,
      "learning_rate": 4.23767439750223e-06,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8596486851572991,
      "num_tokens": 138232460.0,
      "step": 114970
    },
    {
      "entropy": 1.7997772373259067,
      "epoch": 0.35642793982139853,
      "grad_norm": 4.375933647155762,
      "learning_rate": 4.237490113131097e-06,
      "loss": 0.4026,
      "mean_token_accuracy": 0.8406259372830391,
      "num_tokens": 138246226.0,
      "step": 114980
    },
    {
      "entropy": 1.9087285965681076,
      "epoch": 0.35645893894644826,
      "grad_norm": 9.480292320251465,
      "learning_rate": 4.237305852799878e-06,
      "loss": 0.5354,
      "mean_token_accuracy": 0.8353115454316139,
      "num_tokens": 138258064.0,
      "step": 114990
    },
    {
      "entropy": 1.8310803532600404,
      "epoch": 0.3564899380714979,
      "grad_norm": 5.65321683883667,
      "learning_rate": 4.237121616503348e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8590890899300575,
      "num_tokens": 138271607.0,
      "step": 115000
    },
    {
      "entropy": 1.727584820985794,
      "epoch": 0.35652093719654765,
      "grad_norm": 4.688484191894531,
      "learning_rate": 4.2369374042362805e-06,
      "loss": 0.3905,
      "mean_token_accuracy": 0.8516164928674698,
      "num_tokens": 138286322.0,
      "step": 115010
    },
    {
      "entropy": 1.946674033999443,
      "epoch": 0.3565519363215973,
      "grad_norm": 7.794990062713623,
      "learning_rate": 4.236753215993452e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8492588356137276,
      "num_tokens": 138297078.0,
      "step": 115020
    },
    {
      "entropy": 1.7960212633013726,
      "epoch": 0.35658293544664704,
      "grad_norm": 4.166678428649902,
      "learning_rate": 4.236569051769643e-06,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8482924640178681,
      "num_tokens": 138310205.0,
      "step": 115030
    },
    {
      "entropy": 1.8630914211273193,
      "epoch": 0.3566139345716967,
      "grad_norm": 7.438714504241943,
      "learning_rate": 4.236384911559633e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8541238859295845,
      "num_tokens": 138322264.0,
      "step": 115040
    },
    {
      "entropy": 1.8478296250104904,
      "epoch": 0.35664493369674644,
      "grad_norm": 3.994792938232422,
      "learning_rate": 4.236200795358203e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.853163267672062,
      "num_tokens": 138334432.0,
      "step": 115050
    },
    {
      "entropy": 1.840542307496071,
      "epoch": 0.3566759328217961,
      "grad_norm": 7.899695873260498,
      "learning_rate": 4.2360167031601366e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8437621504068374,
      "num_tokens": 138346236.0,
      "step": 115060
    },
    {
      "entropy": 1.9013837978243828,
      "epoch": 0.3567069319468458,
      "grad_norm": 8.970711708068848,
      "learning_rate": 4.235832634960219e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8377508997917176,
      "num_tokens": 138357432.0,
      "step": 115070
    },
    {
      "entropy": 1.8663888260722161,
      "epoch": 0.3567379310718955,
      "grad_norm": 8.907432556152344,
      "learning_rate": 4.235648590753237e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8535327434539794,
      "num_tokens": 138369028.0,
      "step": 115080
    },
    {
      "entropy": 1.8116501927375794,
      "epoch": 0.35676893019694517,
      "grad_norm": 7.725051403045654,
      "learning_rate": 4.235464570533978e-06,
      "loss": 0.3945,
      "mean_token_accuracy": 0.8590543121099472,
      "num_tokens": 138380629.0,
      "step": 115090
    },
    {
      "entropy": 1.8142176762223243,
      "epoch": 0.3567999293219949,
      "grad_norm": 10.798834800720215,
      "learning_rate": 4.2352805742972315e-06,
      "loss": 0.4315,
      "mean_token_accuracy": 0.848982447385788,
      "num_tokens": 138393683.0,
      "step": 115100
    },
    {
      "entropy": 1.8765118628740312,
      "epoch": 0.35683092844704456,
      "grad_norm": 11.132994651794434,
      "learning_rate": 4.23509660203779e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8448266923427582,
      "num_tokens": 138405992.0,
      "step": 115110
    },
    {
      "entropy": 1.895536944270134,
      "epoch": 0.3568619275720943,
      "grad_norm": 7.782128810882568,
      "learning_rate": 4.234912653750445e-06,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8532333001494408,
      "num_tokens": 138417451.0,
      "step": 115120
    },
    {
      "entropy": 1.851093652844429,
      "epoch": 0.35689292669714395,
      "grad_norm": 3.747323751449585,
      "learning_rate": 4.23472872942999e-06,
      "loss": 0.51,
      "mean_token_accuracy": 0.8384227573871612,
      "num_tokens": 138428567.0,
      "step": 115130
    },
    {
      "entropy": 1.9262410998344421,
      "epoch": 0.3569239258221937,
      "grad_norm": 8.404274940490723,
      "learning_rate": 4.234544829071223e-06,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8368908122181893,
      "num_tokens": 138439819.0,
      "step": 115140
    },
    {
      "entropy": 1.8113437429070474,
      "epoch": 0.35695492494724335,
      "grad_norm": 8.569499969482422,
      "learning_rate": 4.234360952668942e-06,
      "loss": 0.4249,
      "mean_token_accuracy": 0.8553946748375892,
      "num_tokens": 138452192.0,
      "step": 115150
    },
    {
      "entropy": 1.8376051411032677,
      "epoch": 0.35698592407229307,
      "grad_norm": 8.602355003356934,
      "learning_rate": 4.2341771002179445e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8601582199335098,
      "num_tokens": 138464939.0,
      "step": 115160
    },
    {
      "entropy": 1.7371099531650542,
      "epoch": 0.35701692319734274,
      "grad_norm": 8.904439926147461,
      "learning_rate": 4.233993271713032e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.8645968466997147,
      "num_tokens": 138477993.0,
      "step": 115170
    },
    {
      "entropy": 1.8617297038435936,
      "epoch": 0.35704792232239246,
      "grad_norm": 7.874594211578369,
      "learning_rate": 4.233809467149005e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.852948172390461,
      "num_tokens": 138489670.0,
      "step": 115180
    },
    {
      "entropy": 1.8260050147771836,
      "epoch": 0.35707892144744213,
      "grad_norm": 3.6491997241973877,
      "learning_rate": 4.23362568652067e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8457119166851044,
      "num_tokens": 138501968.0,
      "step": 115190
    },
    {
      "entropy": 1.84189365953207,
      "epoch": 0.35710992057249186,
      "grad_norm": 8.175023078918457,
      "learning_rate": 4.2334419298228315e-06,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8577545329928398,
      "num_tokens": 138514376.0,
      "step": 115200
    },
    {
      "entropy": 1.8731770426034928,
      "epoch": 0.3571409196975415,
      "grad_norm": 9.102734565734863,
      "learning_rate": 4.2332581970502965e-06,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8443191021680831,
      "num_tokens": 138526194.0,
      "step": 115210
    },
    {
      "entropy": 1.8527789831161499,
      "epoch": 0.35717191882259125,
      "grad_norm": 7.554600238800049,
      "learning_rate": 4.233074488197873e-06,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8490279093384743,
      "num_tokens": 138538253.0,
      "step": 115220
    },
    {
      "entropy": 1.9107401639223098,
      "epoch": 0.3572029179476409,
      "grad_norm": 9.52925968170166,
      "learning_rate": 4.232890803260372e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8357208788394928,
      "num_tokens": 138549714.0,
      "step": 115230
    },
    {
      "entropy": 1.857765756547451,
      "epoch": 0.35723391707269064,
      "grad_norm": 8.061179161071777,
      "learning_rate": 4.232707142232605e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8525500863790512,
      "num_tokens": 138561400.0,
      "step": 115240
    },
    {
      "entropy": 1.852206926047802,
      "epoch": 0.3572649161977403,
      "grad_norm": 2.4236626625061035,
      "learning_rate": 4.232523505109386e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.854184539616108,
      "num_tokens": 138573545.0,
      "step": 115250
    },
    {
      "entropy": 1.9055633306503297,
      "epoch": 0.35729591532279004,
      "grad_norm": 8.553743362426758,
      "learning_rate": 4.232339891885528e-06,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8542478173971176,
      "num_tokens": 138584725.0,
      "step": 115260
    },
    {
      "entropy": 1.917627716064453,
      "epoch": 0.3573269144478397,
      "grad_norm": 8.1434326171875,
      "learning_rate": 4.23215630255585e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8403222784399986,
      "num_tokens": 138596131.0,
      "step": 115270
    },
    {
      "entropy": 1.7823843270540238,
      "epoch": 0.35735791357288943,
      "grad_norm": 8.837855339050293,
      "learning_rate": 4.2319727371151685e-06,
      "loss": 0.449,
      "mean_token_accuracy": 0.8501473888754845,
      "num_tokens": 138609457.0,
      "step": 115280
    },
    {
      "entropy": 1.802987203001976,
      "epoch": 0.3573889126979391,
      "grad_norm": 8.615238189697266,
      "learning_rate": 4.231789195558304e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8549714654684066,
      "num_tokens": 138621968.0,
      "step": 115290
    },
    {
      "entropy": 1.7702410399913788,
      "epoch": 0.3574199118229888,
      "grad_norm": 9.068475723266602,
      "learning_rate": 4.231605677880076e-06,
      "loss": 0.417,
      "mean_token_accuracy": 0.8504804894328117,
      "num_tokens": 138635487.0,
      "step": 115300
    },
    {
      "entropy": 1.7608641982078552,
      "epoch": 0.3574509109480385,
      "grad_norm": 3.600853204727173,
      "learning_rate": 4.2314221840753095e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8591883108019829,
      "num_tokens": 138649773.0,
      "step": 115310
    },
    {
      "entropy": 1.845149078965187,
      "epoch": 0.35748191007308816,
      "grad_norm": 8.9186372756958,
      "learning_rate": 4.231238714138827e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8526797413825988,
      "num_tokens": 138661604.0,
      "step": 115320
    },
    {
      "entropy": 1.7300834864377976,
      "epoch": 0.3575129091981379,
      "grad_norm": 3.3661861419677734,
      "learning_rate": 4.231055268065456e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8640945121645928,
      "num_tokens": 138675068.0,
      "step": 115330
    },
    {
      "entropy": 1.7648856535553932,
      "epoch": 0.35754390832318755,
      "grad_norm": 7.938429832458496,
      "learning_rate": 4.230871845850023e-06,
      "loss": 0.3813,
      "mean_token_accuracy": 0.8690492272377014,
      "num_tokens": 138687865.0,
      "step": 115340
    },
    {
      "entropy": 1.8311346605420113,
      "epoch": 0.3575749074482373,
      "grad_norm": 8.160514831542969,
      "learning_rate": 4.230688447487358e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8510325223207473,
      "num_tokens": 138700240.0,
      "step": 115350
    },
    {
      "entropy": 1.8927728280425071,
      "epoch": 0.35760590657328695,
      "grad_norm": 7.9980926513671875,
      "learning_rate": 4.230505072972291e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8445928290486335,
      "num_tokens": 138711415.0,
      "step": 115360
    },
    {
      "entropy": 1.9151074886322021,
      "epoch": 0.35763690569833667,
      "grad_norm": 8.005682945251465,
      "learning_rate": 4.230321722299654e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.859444510936737,
      "num_tokens": 138722459.0,
      "step": 115370
    },
    {
      "entropy": 1.8123747482895851,
      "epoch": 0.35766790482338634,
      "grad_norm": 3.6913387775421143,
      "learning_rate": 4.23013839546428e-06,
      "loss": 0.4163,
      "mean_token_accuracy": 0.8595644026994705,
      "num_tokens": 138734633.0,
      "step": 115380
    },
    {
      "entropy": 1.8173100531101227,
      "epoch": 0.35769890394843606,
      "grad_norm": 4.180068016052246,
      "learning_rate": 4.2299550924610065e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8400069773197174,
      "num_tokens": 138747367.0,
      "step": 115390
    },
    {
      "entropy": 1.9122267067432404,
      "epoch": 0.35772990307348573,
      "grad_norm": 8.708627700805664,
      "learning_rate": 4.229771813284669e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8578638061881065,
      "num_tokens": 138758011.0,
      "step": 115400
    },
    {
      "entropy": 1.8789207234978675,
      "epoch": 0.35776090219853546,
      "grad_norm": 8.253546714782715,
      "learning_rate": 4.229588557930106e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8481252580881119,
      "num_tokens": 138770045.0,
      "step": 115410
    },
    {
      "entropy": 1.8555074393749238,
      "epoch": 0.3577919013235851,
      "grad_norm": 9.735289573669434,
      "learning_rate": 4.229405326392158e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.850931105017662,
      "num_tokens": 138782324.0,
      "step": 115420
    },
    {
      "entropy": 1.857953730225563,
      "epoch": 0.35782290044863485,
      "grad_norm": 3.539726495742798,
      "learning_rate": 4.229222118665667e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8570713594555854,
      "num_tokens": 138793921.0,
      "step": 115430
    },
    {
      "entropy": 1.8376942738890647,
      "epoch": 0.3578538995736845,
      "grad_norm": 9.80213451385498,
      "learning_rate": 4.229038934745475e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8545270070433617,
      "num_tokens": 138806022.0,
      "step": 115440
    },
    {
      "entropy": 1.803570680320263,
      "epoch": 0.35788489869873424,
      "grad_norm": 9.964187622070312,
      "learning_rate": 4.228855774626427e-06,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8454685240983963,
      "num_tokens": 138818578.0,
      "step": 115450
    },
    {
      "entropy": 1.7595643281936646,
      "epoch": 0.3579158978237839,
      "grad_norm": 8.82363224029541,
      "learning_rate": 4.22867263830337e-06,
      "loss": 0.4198,
      "mean_token_accuracy": 0.8574300840497017,
      "num_tokens": 138831318.0,
      "step": 115460
    },
    {
      "entropy": 1.8091361120343208,
      "epoch": 0.35794689694883364,
      "grad_norm": 7.342787742614746,
      "learning_rate": 4.228489525771151e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8519743844866753,
      "num_tokens": 138843358.0,
      "step": 115470
    },
    {
      "entropy": 1.878666016459465,
      "epoch": 0.3579778960738833,
      "grad_norm": 8.650647163391113,
      "learning_rate": 4.22830643702462e-06,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8367117345333099,
      "num_tokens": 138854574.0,
      "step": 115480
    },
    {
      "entropy": 1.7836987793445587,
      "epoch": 0.35800889519893303,
      "grad_norm": 4.929512023925781,
      "learning_rate": 4.228123372058628e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8506701856851577,
      "num_tokens": 138867932.0,
      "step": 115490
    },
    {
      "entropy": 1.841169884800911,
      "epoch": 0.3580398943239827,
      "grad_norm": 8.031444549560547,
      "learning_rate": 4.227940330868028e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8463330894708634,
      "num_tokens": 138880884.0,
      "step": 115500
    },
    {
      "entropy": 1.8855138301849366,
      "epoch": 0.3580708934490324,
      "grad_norm": 7.73654317855835,
      "learning_rate": 4.227757313447673e-06,
      "loss": 0.5016,
      "mean_token_accuracy": 0.8478637263178825,
      "num_tokens": 138891775.0,
      "step": 115510
    },
    {
      "entropy": 1.8171266317367554,
      "epoch": 0.3581018925740821,
      "grad_norm": 4.404027462005615,
      "learning_rate": 4.2275743197924185e-06,
      "loss": 0.4329,
      "mean_token_accuracy": 0.8575972750782966,
      "num_tokens": 138904411.0,
      "step": 115520
    },
    {
      "entropy": 1.8890212267637252,
      "epoch": 0.3581328916991318,
      "grad_norm": 8.792807579040527,
      "learning_rate": 4.227391349897123e-06,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8435036420822144,
      "num_tokens": 138917174.0,
      "step": 115530
    },
    {
      "entropy": 1.8580753847956657,
      "epoch": 0.3581638908241815,
      "grad_norm": 7.962569236755371,
      "learning_rate": 4.227208403756644e-06,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8595779970288276,
      "num_tokens": 138928817.0,
      "step": 115540
    },
    {
      "entropy": 1.79182361215353,
      "epoch": 0.3581948899492312,
      "grad_norm": 8.742132186889648,
      "learning_rate": 4.227025481365844e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8503703057765961,
      "num_tokens": 138941364.0,
      "step": 115550
    },
    {
      "entropy": 1.9465411305427551,
      "epoch": 0.3582258890742809,
      "grad_norm": 9.272126197814941,
      "learning_rate": 4.226842582719583e-06,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8437400907278061,
      "num_tokens": 138952471.0,
      "step": 115560
    },
    {
      "entropy": 1.8139133349061012,
      "epoch": 0.35825688819933055,
      "grad_norm": 3.899977684020996,
      "learning_rate": 4.226659707812723e-06,
      "loss": 0.3716,
      "mean_token_accuracy": 0.8707442149519921,
      "num_tokens": 138965676.0,
      "step": 115570
    },
    {
      "entropy": 1.8158785462379456,
      "epoch": 0.35828788732438027,
      "grad_norm": 9.818337440490723,
      "learning_rate": 4.2264768566401325e-06,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8463491559028625,
      "num_tokens": 138977871.0,
      "step": 115580
    },
    {
      "entropy": 1.764494700729847,
      "epoch": 0.35831888644942994,
      "grad_norm": 8.205857276916504,
      "learning_rate": 4.226294029196676e-06,
      "loss": 0.3989,
      "mean_token_accuracy": 0.8628710359334946,
      "num_tokens": 138991007.0,
      "step": 115590
    },
    {
      "entropy": 1.8400152049958707,
      "epoch": 0.35834988557447967,
      "grad_norm": 6.771747589111328,
      "learning_rate": 4.226111225477222e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8476285025477409,
      "num_tokens": 139002964.0,
      "step": 115600
    },
    {
      "entropy": 1.899264845252037,
      "epoch": 0.35838088469952933,
      "grad_norm": 9.301812171936035,
      "learning_rate": 4.225928445476641e-06,
      "loss": 0.5283,
      "mean_token_accuracy": 0.8411499798297882,
      "num_tokens": 139014497.0,
      "step": 115610
    },
    {
      "entropy": 1.861107437312603,
      "epoch": 0.35841188382457906,
      "grad_norm": 3.8199284076690674,
      "learning_rate": 4.2257456891898015e-06,
      "loss": 0.518,
      "mean_token_accuracy": 0.8456135064363479,
      "num_tokens": 139026732.0,
      "step": 115620
    },
    {
      "entropy": 1.786482220888138,
      "epoch": 0.3584428829496287,
      "grad_norm": 7.765440940856934,
      "learning_rate": 4.2255629566115795e-06,
      "loss": 0.3853,
      "mean_token_accuracy": 0.8641697198152543,
      "num_tokens": 139039189.0,
      "step": 115630
    },
    {
      "entropy": 1.8395205929875373,
      "epoch": 0.35847388207467845,
      "grad_norm": 9.376654624938965,
      "learning_rate": 4.225380247736847e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.8413670718669891,
      "num_tokens": 139052067.0,
      "step": 115640
    },
    {
      "entropy": 1.9465155333280564,
      "epoch": 0.3585048811997281,
      "grad_norm": 8.243704795837402,
      "learning_rate": 4.225197562560482e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8416017279028892,
      "num_tokens": 139062764.0,
      "step": 115650
    },
    {
      "entropy": 1.842594537138939,
      "epoch": 0.35853588032477784,
      "grad_norm": 4.1204681396484375,
      "learning_rate": 4.22501490107736e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8348633021116256,
      "num_tokens": 139075044.0,
      "step": 115660
    },
    {
      "entropy": 1.8628608852624893,
      "epoch": 0.3585668794498275,
      "grad_norm": 7.374316215515137,
      "learning_rate": 4.2248322632823606e-06,
      "loss": 0.4423,
      "mean_token_accuracy": 0.8515262484550477,
      "num_tokens": 139086635.0,
      "step": 115670
    },
    {
      "entropy": 1.8487340614199639,
      "epoch": 0.35859787857487724,
      "grad_norm": 8.635334014892578,
      "learning_rate": 4.224649649170366e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8458402872085571,
      "num_tokens": 139098547.0,
      "step": 115680
    },
    {
      "entropy": 1.9221700012683869,
      "epoch": 0.3586288776999269,
      "grad_norm": 6.838761806488037,
      "learning_rate": 4.224467058736255e-06,
      "loss": 0.5657,
      "mean_token_accuracy": 0.8330105841159821,
      "num_tokens": 139109818.0,
      "step": 115690
    },
    {
      "entropy": 1.8932587698101997,
      "epoch": 0.35865987682497663,
      "grad_norm": 9.387928009033203,
      "learning_rate": 4.224284491974914e-06,
      "loss": 0.5134,
      "mean_token_accuracy": 0.8380125910043716,
      "num_tokens": 139121059.0,
      "step": 115700
    },
    {
      "entropy": 1.9230019852519036,
      "epoch": 0.3586908759500263,
      "grad_norm": 8.831636428833008,
      "learning_rate": 4.224101948881227e-06,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8391887709498406,
      "num_tokens": 139132889.0,
      "step": 115710
    },
    {
      "entropy": 1.8802688077092171,
      "epoch": 0.358721875075076,
      "grad_norm": 3.8806021213531494,
      "learning_rate": 4.22391942945008e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8473272129893303,
      "num_tokens": 139145205.0,
      "step": 115720
    },
    {
      "entropy": 1.9340951204299928,
      "epoch": 0.3587528742001257,
      "grad_norm": 8.25346851348877,
      "learning_rate": 4.2237369336763625e-06,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8459508880972862,
      "num_tokens": 139156135.0,
      "step": 115730
    },
    {
      "entropy": 1.900227214396,
      "epoch": 0.3587838733251754,
      "grad_norm": 8.245087623596191,
      "learning_rate": 4.223554461554964e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8462023451924324,
      "num_tokens": 139167831.0,
      "step": 115740
    },
    {
      "entropy": 1.8334951400756836,
      "epoch": 0.3588148724502251,
      "grad_norm": 4.438302040100098,
      "learning_rate": 4.223372013080776e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8457862615585328,
      "num_tokens": 139180382.0,
      "step": 115750
    },
    {
      "entropy": 1.8711395308375358,
      "epoch": 0.3588458715752748,
      "grad_norm": 8.4141845703125,
      "learning_rate": 4.223189588248691e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8420770734548568,
      "num_tokens": 139192007.0,
      "step": 115760
    },
    {
      "entropy": 1.892673698067665,
      "epoch": 0.3588768707003245,
      "grad_norm": 8.64199161529541,
      "learning_rate": 4.223007187053604e-06,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8470587819814682,
      "num_tokens": 139203626.0,
      "step": 115770
    },
    {
      "entropy": 1.8397964909672737,
      "epoch": 0.3589078698253742,
      "grad_norm": 8.469186782836914,
      "learning_rate": 4.222824809490409e-06,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8453930094838142,
      "num_tokens": 139215766.0,
      "step": 115780
    },
    {
      "entropy": 1.932980051636696,
      "epoch": 0.3589388689504239,
      "grad_norm": 6.354795455932617,
      "learning_rate": 4.2226424555540065e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8456053122878074,
      "num_tokens": 139227681.0,
      "step": 115790
    },
    {
      "entropy": 1.8004991948604583,
      "epoch": 0.3589698680754736,
      "grad_norm": 6.467407703399658,
      "learning_rate": 4.2224601252392935e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8630038425326347,
      "num_tokens": 139239994.0,
      "step": 115800
    },
    {
      "entropy": 1.8661329224705696,
      "epoch": 0.35900086720052327,
      "grad_norm": 4.90010929107666,
      "learning_rate": 4.222277818541172e-06,
      "loss": 0.498,
      "mean_token_accuracy": 0.8436091437935829,
      "num_tokens": 139251297.0,
      "step": 115810
    },
    {
      "entropy": 1.8228035122156143,
      "epoch": 0.35903186632557293,
      "grad_norm": 6.741776943206787,
      "learning_rate": 4.2220955354545435e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8479675367474556,
      "num_tokens": 139263722.0,
      "step": 115820
    },
    {
      "entropy": 1.796131867170334,
      "epoch": 0.35906286545062266,
      "grad_norm": 6.297611713409424,
      "learning_rate": 4.221913275974311e-06,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8649747148156166,
      "num_tokens": 139276144.0,
      "step": 115830
    },
    {
      "entropy": 1.873625811934471,
      "epoch": 0.35909386457567233,
      "grad_norm": 7.727682590484619,
      "learning_rate": 4.221731040095381e-06,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8530295416712761,
      "num_tokens": 139287760.0,
      "step": 115840
    },
    {
      "entropy": 1.8328562811017037,
      "epoch": 0.35912486370072205,
      "grad_norm": 8.941142082214355,
      "learning_rate": 4.221548827812659e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.8467715248465538,
      "num_tokens": 139300146.0,
      "step": 115850
    },
    {
      "entropy": 1.8431931123137475,
      "epoch": 0.3591558628257717,
      "grad_norm": 8.686592102050781,
      "learning_rate": 4.221366639121054e-06,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8481633648276329,
      "num_tokens": 139312174.0,
      "step": 115860
    },
    {
      "entropy": 1.7692224755883217,
      "epoch": 0.35918686195082145,
      "grad_norm": 7.942878723144531,
      "learning_rate": 4.221184474015477e-06,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8561850890517235,
      "num_tokens": 139325247.0,
      "step": 115870
    },
    {
      "entropy": 1.8758103132247925,
      "epoch": 0.3592178610758711,
      "grad_norm": 8.962884902954102,
      "learning_rate": 4.221002332490837e-06,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8550149217247963,
      "num_tokens": 139336232.0,
      "step": 115880
    },
    {
      "entropy": 1.7588933199644088,
      "epoch": 0.35924886020092084,
      "grad_norm": 4.053327560424805,
      "learning_rate": 4.220820214542049e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.8556968107819557,
      "num_tokens": 139350251.0,
      "step": 115890
    },
    {
      "entropy": 1.8734589472413063,
      "epoch": 0.3592798593259705,
      "grad_norm": 9.165533065795898,
      "learning_rate": 4.220638120164026e-06,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8462863549590111,
      "num_tokens": 139361764.0,
      "step": 115900
    },
    {
      "entropy": 1.8266181737184524,
      "epoch": 0.35931085845102023,
      "grad_norm": 7.9356255531311035,
      "learning_rate": 4.220456049351685e-06,
      "loss": 0.45,
      "mean_token_accuracy": 0.8596974804997444,
      "num_tokens": 139373578.0,
      "step": 115910
    },
    {
      "entropy": 1.8083546802401542,
      "epoch": 0.3593418575760699,
      "grad_norm": 3.8307013511657715,
      "learning_rate": 4.220274002099943e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.849444305896759,
      "num_tokens": 139386017.0,
      "step": 115920
    },
    {
      "entropy": 1.8357669502496718,
      "epoch": 0.3593728567011196,
      "grad_norm": 3.281771421432495,
      "learning_rate": 4.220091978403719e-06,
      "loss": 0.5271,
      "mean_token_accuracy": 0.848252309858799,
      "num_tokens": 139398172.0,
      "step": 115930
    },
    {
      "entropy": 1.818470537662506,
      "epoch": 0.3594038558261693,
      "grad_norm": 2.5742039680480957,
      "learning_rate": 4.219909978257934e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8499492272734642,
      "num_tokens": 139410862.0,
      "step": 115940
    },
    {
      "entropy": 1.8105147659778595,
      "epoch": 0.359434854951219,
      "grad_norm": 7.754059791564941,
      "learning_rate": 4.219728001657508e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.8644904047250748,
      "num_tokens": 139422867.0,
      "step": 115950
    },
    {
      "entropy": 1.8963597849011422,
      "epoch": 0.3594658540762687,
      "grad_norm": 9.430159568786621,
      "learning_rate": 4.219546048597369e-06,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8429613292217255,
      "num_tokens": 139433810.0,
      "step": 115960
    },
    {
      "entropy": 1.8394680365920066,
      "epoch": 0.3594968532013184,
      "grad_norm": 3.8290538787841797,
      "learning_rate": 4.219364119072439e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8480119571089745,
      "num_tokens": 139446229.0,
      "step": 115970
    },
    {
      "entropy": 1.8272418528795242,
      "epoch": 0.3595278523263681,
      "grad_norm": 7.950870990753174,
      "learning_rate": 4.219182213077646e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8526568979024887,
      "num_tokens": 139458579.0,
      "step": 115980
    },
    {
      "entropy": 1.887735801935196,
      "epoch": 0.3595588514514178,
      "grad_norm": 3.933110237121582,
      "learning_rate": 4.219000330607917e-06,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8449479952454567,
      "num_tokens": 139470657.0,
      "step": 115990
    },
    {
      "entropy": 1.9549269363284112,
      "epoch": 0.3595898505764675,
      "grad_norm": 9.912589073181152,
      "learning_rate": 4.218818471658183e-06,
      "loss": 0.4899,
      "mean_token_accuracy": 0.841597031056881,
      "num_tokens": 139481544.0,
      "step": 116000
    },
    {
      "entropy": 1.904000848531723,
      "epoch": 0.3596208497015172,
      "grad_norm": 6.567648887634277,
      "learning_rate": 4.218636636223375e-06,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8507451817393303,
      "num_tokens": 139492483.0,
      "step": 116010
    },
    {
      "entropy": 1.9398685723543168,
      "epoch": 0.35965184882656687,
      "grad_norm": 4.198793411254883,
      "learning_rate": 4.218454824298425e-06,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8451578319072723,
      "num_tokens": 139504258.0,
      "step": 116020
    },
    {
      "entropy": 1.9017282500863075,
      "epoch": 0.3596828479516166,
      "grad_norm": 7.781552314758301,
      "learning_rate": 4.218273035878269e-06,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8435791626572609,
      "num_tokens": 139515473.0,
      "step": 116030
    },
    {
      "entropy": 1.9017712712287902,
      "epoch": 0.35971384707666626,
      "grad_norm": 8.259926795959473,
      "learning_rate": 4.218091270957841e-06,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8393255040049553,
      "num_tokens": 139526776.0,
      "step": 116040
    },
    {
      "entropy": 1.8337740018963813,
      "epoch": 0.359744846201716,
      "grad_norm": 9.133225440979004,
      "learning_rate": 4.21790952953208e-06,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8523485392332077,
      "num_tokens": 139539083.0,
      "step": 116050
    },
    {
      "entropy": 1.7753349527716638,
      "epoch": 0.35977584532676565,
      "grad_norm": 7.402058124542236,
      "learning_rate": 4.217727811595925e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.847336596250534,
      "num_tokens": 139552085.0,
      "step": 116060
    },
    {
      "entropy": 1.8660298094153405,
      "epoch": 0.3598068444518153,
      "grad_norm": 7.4676337242126465,
      "learning_rate": 4.217546117144314e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8392193883657455,
      "num_tokens": 139564096.0,
      "step": 116070
    },
    {
      "entropy": 1.8619601786136628,
      "epoch": 0.35983784357686505,
      "grad_norm": 9.07433795928955,
      "learning_rate": 4.217364446172193e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.839654740691185,
      "num_tokens": 139576414.0,
      "step": 116080
    },
    {
      "entropy": 1.9037289932370185,
      "epoch": 0.3598688427019147,
      "grad_norm": 8.008439064025879,
      "learning_rate": 4.217182798674502e-06,
      "loss": 0.4998,
      "mean_token_accuracy": 0.8401681199669838,
      "num_tokens": 139587784.0,
      "step": 116090
    },
    {
      "entropy": 1.8978828579187392,
      "epoch": 0.35989984182696444,
      "grad_norm": 8.578210830688477,
      "learning_rate": 4.2170011746461886e-06,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8408385306596756,
      "num_tokens": 139599171.0,
      "step": 116100
    },
    {
      "entropy": 1.84117241948843,
      "epoch": 0.3599308409520141,
      "grad_norm": 3.883363962173462,
      "learning_rate": 4.216819574082197e-06,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8472616419196128,
      "num_tokens": 139611729.0,
      "step": 116110
    },
    {
      "entropy": 1.8370385035872459,
      "epoch": 0.35996184007706383,
      "grad_norm": 4.151669979095459,
      "learning_rate": 4.2166379969774775e-06,
      "loss": 0.4473,
      "mean_token_accuracy": 0.855852136015892,
      "num_tokens": 139623044.0,
      "step": 116120
    },
    {
      "entropy": 1.8387348473072052,
      "epoch": 0.3599928392021135,
      "grad_norm": 8.288407325744629,
      "learning_rate": 4.216456443326979e-06,
      "loss": 0.4327,
      "mean_token_accuracy": 0.8595815673470497,
      "num_tokens": 139634713.0,
      "step": 116130
    },
    {
      "entropy": 1.9357994318008422,
      "epoch": 0.3600238383271632,
      "grad_norm": 10.54449462890625,
      "learning_rate": 4.216274913125652e-06,
      "loss": 0.5153,
      "mean_token_accuracy": 0.8542236015200615,
      "num_tokens": 139645451.0,
      "step": 116140
    },
    {
      "entropy": 1.8762347102165222,
      "epoch": 0.3600548374522129,
      "grad_norm": 8.778923034667969,
      "learning_rate": 4.216093406368449e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8462166652083397,
      "num_tokens": 139657312.0,
      "step": 116150
    },
    {
      "entropy": 1.8865218967199326,
      "epoch": 0.3600858365772626,
      "grad_norm": 7.668148040771484,
      "learning_rate": 4.2159119230503255e-06,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8440534695982933,
      "num_tokens": 139668202.0,
      "step": 116160
    },
    {
      "entropy": 1.8358159840106965,
      "epoch": 0.3601168357023123,
      "grad_norm": 8.311078071594238,
      "learning_rate": 4.215730463166237e-06,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8523998752236366,
      "num_tokens": 139680084.0,
      "step": 116170
    },
    {
      "entropy": 1.925374338030815,
      "epoch": 0.360147834827362,
      "grad_norm": 8.066094398498535,
      "learning_rate": 4.21554902671114e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8462306708097458,
      "num_tokens": 139690862.0,
      "step": 116180
    },
    {
      "entropy": 1.8480087623000145,
      "epoch": 0.3601788339524117,
      "grad_norm": 4.514954090118408,
      "learning_rate": 4.2153676136799934e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8582058116793633,
      "num_tokens": 139703017.0,
      "step": 116190
    },
    {
      "entropy": 1.7714590035378932,
      "epoch": 0.3602098330774614,
      "grad_norm": 2.609532356262207,
      "learning_rate": 4.215186224067758e-06,
      "loss": 0.3956,
      "mean_token_accuracy": 0.8530898556113243,
      "num_tokens": 139716890.0,
      "step": 116200
    },
    {
      "entropy": 1.9455319941043854,
      "epoch": 0.3602408322025111,
      "grad_norm": 7.525179862976074,
      "learning_rate": 4.215004857869394e-06,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8397414952516555,
      "num_tokens": 139727863.0,
      "step": 116210
    },
    {
      "entropy": 1.766502921283245,
      "epoch": 0.3602718313275608,
      "grad_norm": 8.821333885192871,
      "learning_rate": 4.214823515079867e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8470347732305527,
      "num_tokens": 139740638.0,
      "step": 116220
    },
    {
      "entropy": 1.8291141256690024,
      "epoch": 0.36030283045261047,
      "grad_norm": 4.927423000335693,
      "learning_rate": 4.214642195694141e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.8610336139798165,
      "num_tokens": 139753345.0,
      "step": 116230
    },
    {
      "entropy": 1.8271966442465781,
      "epoch": 0.3603338295776602,
      "grad_norm": 9.124969482421875,
      "learning_rate": 4.214460899707181e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8442619845271111,
      "num_tokens": 139765345.0,
      "step": 116240
    },
    {
      "entropy": 1.8535583779215812,
      "epoch": 0.36036482870270986,
      "grad_norm": 8.4541597366333,
      "learning_rate": 4.214279627113957e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8529452204704284,
      "num_tokens": 139777146.0,
      "step": 116250
    },
    {
      "entropy": 1.8414417818188666,
      "epoch": 0.3603958278277596,
      "grad_norm": 3.8833229541778564,
      "learning_rate": 4.214098377909436e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.852535355091095,
      "num_tokens": 139789410.0,
      "step": 116260
    },
    {
      "entropy": 1.8427911520004272,
      "epoch": 0.36042682695280925,
      "grad_norm": 3.8969404697418213,
      "learning_rate": 4.213917152088591e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8565546840429306,
      "num_tokens": 139801439.0,
      "step": 116270
    },
    {
      "entropy": 1.7546941101551057,
      "epoch": 0.360457826077859,
      "grad_norm": 8.13198184967041,
      "learning_rate": 4.2137359496463936e-06,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8690331175923347,
      "num_tokens": 139814865.0,
      "step": 116280
    },
    {
      "entropy": 1.7768574252724647,
      "epoch": 0.36048882520290865,
      "grad_norm": 3.7505130767822266,
      "learning_rate": 4.213554770577818e-06,
      "loss": 0.3773,
      "mean_token_accuracy": 0.86257985830307,
      "num_tokens": 139828042.0,
      "step": 116290
    },
    {
      "entropy": 1.8828290045261382,
      "epoch": 0.3605198243279583,
      "grad_norm": 6.885351181030273,
      "learning_rate": 4.213373614877838e-06,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8605077177286148,
      "num_tokens": 139838856.0,
      "step": 116300
    },
    {
      "entropy": 1.7729797944426537,
      "epoch": 0.36055082345300804,
      "grad_norm": 9.330283164978027,
      "learning_rate": 4.213192482541433e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.8671557918190956,
      "num_tokens": 139852638.0,
      "step": 116310
    },
    {
      "entropy": 1.9487037807703018,
      "epoch": 0.3605818225780577,
      "grad_norm": 7.210131645202637,
      "learning_rate": 4.21301137356358e-06,
      "loss": 0.5066,
      "mean_token_accuracy": 0.8386420547962189,
      "num_tokens": 139863146.0,
      "step": 116320
    },
    {
      "entropy": 1.9695353388786316,
      "epoch": 0.36061282170310743,
      "grad_norm": 9.178292274475098,
      "learning_rate": 4.21283028793926e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8453469514846802,
      "num_tokens": 139874014.0,
      "step": 116330
    },
    {
      "entropy": 1.8018761366605758,
      "epoch": 0.3606438208281571,
      "grad_norm": 7.328160285949707,
      "learning_rate": 4.212649225663452e-06,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8488622546195984,
      "num_tokens": 139886434.0,
      "step": 116340
    },
    {
      "entropy": 1.9044070094823837,
      "epoch": 0.3606748199532068,
      "grad_norm": 7.514301776885986,
      "learning_rate": 4.212468186731141e-06,
      "loss": 0.5155,
      "mean_token_accuracy": 0.8392976179718972,
      "num_tokens": 139897996.0,
      "step": 116350
    },
    {
      "entropy": 1.8434274643659592,
      "epoch": 0.3607058190782565,
      "grad_norm": 8.972174644470215,
      "learning_rate": 4.212287171137313e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8550772503018379,
      "num_tokens": 139910313.0,
      "step": 116360
    },
    {
      "entropy": 1.803036929666996,
      "epoch": 0.3607368182033062,
      "grad_norm": 8.392457008361816,
      "learning_rate": 4.212106178876951e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8607822299003601,
      "num_tokens": 139922530.0,
      "step": 116370
    },
    {
      "entropy": 1.8568144857883453,
      "epoch": 0.3607678173283559,
      "grad_norm": 3.606626510620117,
      "learning_rate": 4.211925209945044e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8569135695695878,
      "num_tokens": 139933805.0,
      "step": 116380
    },
    {
      "entropy": 1.9102680534124374,
      "epoch": 0.3607988164534056,
      "grad_norm": 8.40206241607666,
      "learning_rate": 4.21174426433658e-06,
      "loss": 0.5215,
      "mean_token_accuracy": 0.8408423826098442,
      "num_tokens": 139945504.0,
      "step": 116390
    },
    {
      "entropy": 1.8889756113290788,
      "epoch": 0.3608298155784553,
      "grad_norm": 7.147616386413574,
      "learning_rate": 4.21156334204655e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8580622985959053,
      "num_tokens": 139956323.0,
      "step": 116400
    },
    {
      "entropy": 1.7778198927640916,
      "epoch": 0.360860814703505,
      "grad_norm": 8.543440818786621,
      "learning_rate": 4.211382443069949e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.8652505591511727,
      "num_tokens": 139968634.0,
      "step": 116410
    },
    {
      "entropy": 1.877928839623928,
      "epoch": 0.3608918138285547,
      "grad_norm": 2.7431368827819824,
      "learning_rate": 4.2112015674017645e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8467942878603936,
      "num_tokens": 139981044.0,
      "step": 116420
    },
    {
      "entropy": 1.8687953114509583,
      "epoch": 0.3609228129536044,
      "grad_norm": 8.985494613647461,
      "learning_rate": 4.211020715036995e-06,
      "loss": 0.5351,
      "mean_token_accuracy": 0.844055813550949,
      "num_tokens": 139992856.0,
      "step": 116430
    },
    {
      "entropy": 1.8689031660556794,
      "epoch": 0.36095381207865407,
      "grad_norm": 9.813769340515137,
      "learning_rate": 4.210839885970638e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8516811087727547,
      "num_tokens": 140003419.0,
      "step": 116440
    },
    {
      "entropy": 1.8415801107883454,
      "epoch": 0.3609848112037038,
      "grad_norm": 3.827608346939087,
      "learning_rate": 4.210659080197691e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.8390990689396858,
      "num_tokens": 140014798.0,
      "step": 116450
    },
    {
      "entropy": 1.847266887128353,
      "epoch": 0.36101581032875346,
      "grad_norm": 8.154986381530762,
      "learning_rate": 4.210478297713152e-06,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8583290934562683,
      "num_tokens": 140026777.0,
      "step": 116460
    },
    {
      "entropy": 1.872428523004055,
      "epoch": 0.3610468094538032,
      "grad_norm": 9.429468154907227,
      "learning_rate": 4.210297538512023e-06,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8427754417061806,
      "num_tokens": 140038193.0,
      "step": 116470
    },
    {
      "entropy": 1.815330520272255,
      "epoch": 0.36107780857885285,
      "grad_norm": 7.44341516494751,
      "learning_rate": 4.210116802589307e-06,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8459246262907982,
      "num_tokens": 140049896.0,
      "step": 116480
    },
    {
      "entropy": 1.80218645632267,
      "epoch": 0.3611088077039026,
      "grad_norm": 4.414290428161621,
      "learning_rate": 4.209936089940008e-06,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8553975149989128,
      "num_tokens": 140062012.0,
      "step": 116490
    },
    {
      "entropy": 1.8422569379210472,
      "epoch": 0.36113980682895225,
      "grad_norm": 7.908370494842529,
      "learning_rate": 4.20975540055913e-06,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8528754249215126,
      "num_tokens": 140074154.0,
      "step": 116500
    },
    {
      "entropy": 1.7738358169794082,
      "epoch": 0.36117080595400197,
      "grad_norm": 7.086633682250977,
      "learning_rate": 4.2095747344416815e-06,
      "loss": 0.3905,
      "mean_token_accuracy": 0.8605238318443298,
      "num_tokens": 140086516.0,
      "step": 116510
    },
    {
      "entropy": 1.8560999408364296,
      "epoch": 0.36120180507905164,
      "grad_norm": 8.640069007873535,
      "learning_rate": 4.209394091582671e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.8428186282515526,
      "num_tokens": 140098344.0,
      "step": 116520
    },
    {
      "entropy": 1.8515635520219802,
      "epoch": 0.36123280420410137,
      "grad_norm": 9.717535972595215,
      "learning_rate": 4.209213471977109e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.846831089258194,
      "num_tokens": 140110320.0,
      "step": 116530
    },
    {
      "entropy": 1.8465173259377479,
      "epoch": 0.36126380332915103,
      "grad_norm": 7.840369701385498,
      "learning_rate": 4.209032875620006e-06,
      "loss": 0.4722,
      "mean_token_accuracy": 0.8543838858604431,
      "num_tokens": 140122072.0,
      "step": 116540
    },
    {
      "entropy": 1.793773990869522,
      "epoch": 0.3612948024542007,
      "grad_norm": 7.085291385650635,
      "learning_rate": 4.2088523025063745e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8551607668399811,
      "num_tokens": 140134814.0,
      "step": 116550
    },
    {
      "entropy": 1.8004359588027001,
      "epoch": 0.3613258015792504,
      "grad_norm": 2.7206969261169434,
      "learning_rate": 4.208671752631231e-06,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8510082602500916,
      "num_tokens": 140147230.0,
      "step": 116560
    },
    {
      "entropy": 1.838956043124199,
      "epoch": 0.3613568007043001,
      "grad_norm": 4.567168235778809,
      "learning_rate": 4.20849122598959e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8527957811951637,
      "num_tokens": 140159657.0,
      "step": 116570
    },
    {
      "entropy": 1.8510146543383599,
      "epoch": 0.3613877998293498,
      "grad_norm": 8.271007537841797,
      "learning_rate": 4.208310722576469e-06,
      "loss": 0.4629,
      "mean_token_accuracy": 0.8527455732226372,
      "num_tokens": 140171588.0,
      "step": 116580
    },
    {
      "entropy": 1.817806077003479,
      "epoch": 0.3614187989543995,
      "grad_norm": 8.602225303649902,
      "learning_rate": 4.208130242386889e-06,
      "loss": 0.4285,
      "mean_token_accuracy": 0.8533978506922721,
      "num_tokens": 140184012.0,
      "step": 116590
    },
    {
      "entropy": 1.8294128969311714,
      "epoch": 0.3614497980794492,
      "grad_norm": 7.451015949249268,
      "learning_rate": 4.207949785415868e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8549151733517647,
      "num_tokens": 140196072.0,
      "step": 116600
    },
    {
      "entropy": 1.9123011171817779,
      "epoch": 0.3614807972044989,
      "grad_norm": 10.21450138092041,
      "learning_rate": 4.207769351658429e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8501939132809639,
      "num_tokens": 140206777.0,
      "step": 116610
    },
    {
      "entropy": 1.898598951101303,
      "epoch": 0.3615117963295486,
      "grad_norm": 7.918070316314697,
      "learning_rate": 4.2075889411095965e-06,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8529366075992584,
      "num_tokens": 140217957.0,
      "step": 116620
    },
    {
      "entropy": 1.871064305305481,
      "epoch": 0.3615427954545983,
      "grad_norm": 3.4128713607788086,
      "learning_rate": 4.2074085537643945e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8430089339613914,
      "num_tokens": 140229754.0,
      "step": 116630
    },
    {
      "entropy": 1.8388782501220704,
      "epoch": 0.361573794579648,
      "grad_norm": 3.9377996921539307,
      "learning_rate": 4.207228189617849e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.855095773935318,
      "num_tokens": 140242273.0,
      "step": 116640
    },
    {
      "entropy": 1.7982016265392304,
      "epoch": 0.36160479370469767,
      "grad_norm": 4.408384323120117,
      "learning_rate": 4.20704784866499e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8601232185959816,
      "num_tokens": 140255003.0,
      "step": 116650
    },
    {
      "entropy": 1.822973382472992,
      "epoch": 0.3616357928297474,
      "grad_norm": 9.451082229614258,
      "learning_rate": 4.206867530900845e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8590276911854744,
      "num_tokens": 140267479.0,
      "step": 116660
    },
    {
      "entropy": 1.7897161670029162,
      "epoch": 0.36166679195479706,
      "grad_norm": 8.063048362731934,
      "learning_rate": 4.206687236320445e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8569493010640145,
      "num_tokens": 140280536.0,
      "step": 116670
    },
    {
      "entropy": 1.8633152276277543,
      "epoch": 0.3616977910798468,
      "grad_norm": 7.586259365081787,
      "learning_rate": 4.206506964918824e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.84778813123703,
      "num_tokens": 140292788.0,
      "step": 116680
    },
    {
      "entropy": 1.9379972368478775,
      "epoch": 0.36172879020489646,
      "grad_norm": 8.198206901550293,
      "learning_rate": 4.206326716691015e-06,
      "loss": 0.5271,
      "mean_token_accuracy": 0.8326962515711784,
      "num_tokens": 140303768.0,
      "step": 116690
    },
    {
      "entropy": 1.8760272949934005,
      "epoch": 0.3617597893299462,
      "grad_norm": 7.511582851409912,
      "learning_rate": 4.206146491632053e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8527100086212158,
      "num_tokens": 140314927.0,
      "step": 116700
    },
    {
      "entropy": 1.8044881626963616,
      "epoch": 0.36179078845499585,
      "grad_norm": 7.948654651641846,
      "learning_rate": 4.205966289736976e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.8563151493668556,
      "num_tokens": 140327161.0,
      "step": 116710
    },
    {
      "entropy": 1.9257992446422576,
      "epoch": 0.3618217875800456,
      "grad_norm": 9.45364761352539,
      "learning_rate": 4.205786111000822e-06,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8418830558657646,
      "num_tokens": 140337921.0,
      "step": 116720
    },
    {
      "entropy": 1.8222204566001892,
      "epoch": 0.36185278670509524,
      "grad_norm": 3.7886269092559814,
      "learning_rate": 4.2056059554186305e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8379009708762168,
      "num_tokens": 140350153.0,
      "step": 116730
    },
    {
      "entropy": 1.8184516310691834,
      "epoch": 0.36188378583014497,
      "grad_norm": 4.151296138763428,
      "learning_rate": 4.2054258229854435e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8527412965893746,
      "num_tokens": 140363079.0,
      "step": 116740
    },
    {
      "entropy": 1.8538252338767052,
      "epoch": 0.36191478495519463,
      "grad_norm": 6.772287845611572,
      "learning_rate": 4.205245713696304e-06,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8451321288943291,
      "num_tokens": 140375101.0,
      "step": 116750
    },
    {
      "entropy": 1.9226402550935746,
      "epoch": 0.36194578408024436,
      "grad_norm": 8.788530349731445,
      "learning_rate": 4.205065627546256e-06,
      "loss": 0.5342,
      "mean_token_accuracy": 0.8383546307682991,
      "num_tokens": 140386168.0,
      "step": 116760
    },
    {
      "entropy": 1.7971034452319146,
      "epoch": 0.36197678320529403,
      "grad_norm": 9.055315017700195,
      "learning_rate": 4.204885564530345e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8534369856119156,
      "num_tokens": 140399104.0,
      "step": 116770
    },
    {
      "entropy": 1.903681069612503,
      "epoch": 0.36200778233034375,
      "grad_norm": 8.377460479736328,
      "learning_rate": 4.204705524643619e-06,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8296264916658401,
      "num_tokens": 140409966.0,
      "step": 116780
    },
    {
      "entropy": 1.8770312041044235,
      "epoch": 0.3620387814553934,
      "grad_norm": 7.98540735244751,
      "learning_rate": 4.204525507881126e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8401692688465119,
      "num_tokens": 140423005.0,
      "step": 116790
    },
    {
      "entropy": 1.8793371990323067,
      "epoch": 0.3620697805804431,
      "grad_norm": 8.905379295349121,
      "learning_rate": 4.204345514237917e-06,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8462567001581192,
      "num_tokens": 140434436.0,
      "step": 116800
    },
    {
      "entropy": 1.8640392243862152,
      "epoch": 0.3621007797054928,
      "grad_norm": 4.022253036499023,
      "learning_rate": 4.204165543709043e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8513988614082336,
      "num_tokens": 140446710.0,
      "step": 116810
    },
    {
      "entropy": 1.8475000470876695,
      "epoch": 0.3621317788305425,
      "grad_norm": 8.355897903442383,
      "learning_rate": 4.2039855962895605e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8459990754723549,
      "num_tokens": 140458545.0,
      "step": 116820
    },
    {
      "entropy": 1.8094294220209122,
      "epoch": 0.3621627779555922,
      "grad_norm": 2.8843464851379395,
      "learning_rate": 4.203805671974519e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8503652885556221,
      "num_tokens": 140471070.0,
      "step": 116830
    },
    {
      "entropy": 1.8952600628137588,
      "epoch": 0.3621937770806419,
      "grad_norm": 7.821008682250977,
      "learning_rate": 4.203625770758979e-06,
      "loss": 0.5673,
      "mean_token_accuracy": 0.8331520974636077,
      "num_tokens": 140482920.0,
      "step": 116840
    },
    {
      "entropy": 1.827965249121189,
      "epoch": 0.3622247762056916,
      "grad_norm": 9.462203025817871,
      "learning_rate": 4.2034458926379965e-06,
      "loss": 0.4623,
      "mean_token_accuracy": 0.854870118200779,
      "num_tokens": 140495151.0,
      "step": 116850
    },
    {
      "entropy": 1.8622143000364304,
      "epoch": 0.36225577533074127,
      "grad_norm": 8.208244323730469,
      "learning_rate": 4.20326603760663e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8542701035737992,
      "num_tokens": 140507333.0,
      "step": 116860
    },
    {
      "entropy": 1.7744985669851303,
      "epoch": 0.362286774455791,
      "grad_norm": 8.417732238769531,
      "learning_rate": 4.2030862056599415e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8632864966988564,
      "num_tokens": 140520558.0,
      "step": 116870
    },
    {
      "entropy": 1.8826520085334777,
      "epoch": 0.36231777358084066,
      "grad_norm": 8.546829223632812,
      "learning_rate": 4.202906396792993e-06,
      "loss": 0.4866,
      "mean_token_accuracy": 0.8448991179466248,
      "num_tokens": 140532773.0,
      "step": 116880
    },
    {
      "entropy": 1.8119245707988738,
      "epoch": 0.3623487727058904,
      "grad_norm": 7.6031599044799805,
      "learning_rate": 4.2027266110008475e-06,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8597419142723084,
      "num_tokens": 140544815.0,
      "step": 116890
    },
    {
      "entropy": 1.8004948943853378,
      "epoch": 0.36237977183094006,
      "grad_norm": 9.985618591308594,
      "learning_rate": 4.202546848278572e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8449292674660682,
      "num_tokens": 140557252.0,
      "step": 116900
    },
    {
      "entropy": 1.8618132412433623,
      "epoch": 0.3624107709559898,
      "grad_norm": 6.6439290046691895,
      "learning_rate": 4.2023671086212295e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8493566080927849,
      "num_tokens": 140568584.0,
      "step": 116910
    },
    {
      "entropy": 1.7272257059812546,
      "epoch": 0.36244177008103945,
      "grad_norm": 5.679489612579346,
      "learning_rate": 4.202187392023891e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8584482952952385,
      "num_tokens": 140582578.0,
      "step": 116920
    },
    {
      "entropy": 1.9361666440963745,
      "epoch": 0.3624727692060892,
      "grad_norm": 7.492152690887451,
      "learning_rate": 4.202007698481626e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8505539804697037,
      "num_tokens": 140593156.0,
      "step": 116930
    },
    {
      "entropy": 1.9066250026226044,
      "epoch": 0.36250376833113884,
      "grad_norm": 9.594344139099121,
      "learning_rate": 4.201828027989504e-06,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8503702595829964,
      "num_tokens": 140604659.0,
      "step": 116940
    },
    {
      "entropy": 1.8485996812582015,
      "epoch": 0.36253476745618857,
      "grad_norm": 9.402889251708984,
      "learning_rate": 4.201648380542599e-06,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8464497730135918,
      "num_tokens": 140616897.0,
      "step": 116950
    },
    {
      "entropy": 1.8569966793060302,
      "epoch": 0.36256576658123824,
      "grad_norm": 8.817802429199219,
      "learning_rate": 4.201468756135983e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.839309224486351,
      "num_tokens": 140628644.0,
      "step": 116960
    },
    {
      "entropy": 1.827142098546028,
      "epoch": 0.36259676570628796,
      "grad_norm": 7.352452278137207,
      "learning_rate": 4.201289154764733e-06,
      "loss": 0.477,
      "mean_token_accuracy": 0.8442728266119957,
      "num_tokens": 140640707.0,
      "step": 116970
    },
    {
      "entropy": 1.807940025627613,
      "epoch": 0.36262776483133763,
      "grad_norm": 4.695886135101318,
      "learning_rate": 4.201109576423926e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8374917089939118,
      "num_tokens": 140653668.0,
      "step": 116980
    },
    {
      "entropy": 1.8596377670764923,
      "epoch": 0.36265876395638735,
      "grad_norm": 9.165081024169922,
      "learning_rate": 4.20093002110864e-06,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8492574363946914,
      "num_tokens": 140665982.0,
      "step": 116990
    },
    {
      "entropy": 1.8663609504699707,
      "epoch": 0.362689763081437,
      "grad_norm": 7.703036785125732,
      "learning_rate": 4.200750488813955e-06,
      "loss": 0.4468,
      "mean_token_accuracy": 0.8582181662321091,
      "num_tokens": 140677607.0,
      "step": 117000
    },
    {
      "entropy": 1.808670112490654,
      "epoch": 0.36272076220648675,
      "grad_norm": 3.881373405456543,
      "learning_rate": 4.200570979534951e-06,
      "loss": 0.3998,
      "mean_token_accuracy": 0.856096900999546,
      "num_tokens": 140690750.0,
      "step": 117010
    },
    {
      "entropy": 1.8351006895303725,
      "epoch": 0.3627517613315364,
      "grad_norm": 3.9870259761810303,
      "learning_rate": 4.200391493266714e-06,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8495851859450341,
      "num_tokens": 140702826.0,
      "step": 117020
    },
    {
      "entropy": 1.8416362181305885,
      "epoch": 0.36278276045658614,
      "grad_norm": 7.493378639221191,
      "learning_rate": 4.200212030004326e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8568256065249443,
      "num_tokens": 140714882.0,
      "step": 117030
    },
    {
      "entropy": 1.8965220645070076,
      "epoch": 0.3628137595816358,
      "grad_norm": 7.964942455291748,
      "learning_rate": 4.200032589742872e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8496366932988166,
      "num_tokens": 140726128.0,
      "step": 117040
    },
    {
      "entropy": 1.835560789704323,
      "epoch": 0.3628447587066855,
      "grad_norm": 8.899863243103027,
      "learning_rate": 4.1998531724774405e-06,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8567850336432457,
      "num_tokens": 140738179.0,
      "step": 117050
    },
    {
      "entropy": 1.8459725484251976,
      "epoch": 0.3628757578317352,
      "grad_norm": 6.545345306396484,
      "learning_rate": 4.19967377820312e-06,
      "loss": 0.4149,
      "mean_token_accuracy": 0.8515329957008362,
      "num_tokens": 140750630.0,
      "step": 117060
    },
    {
      "entropy": 1.9014078021049499,
      "epoch": 0.36290675695678487,
      "grad_norm": 4.256891250610352,
      "learning_rate": 4.199494406915001e-06,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8485517993569374,
      "num_tokens": 140762145.0,
      "step": 117070
    },
    {
      "entropy": 1.8998582303524016,
      "epoch": 0.3629377560818346,
      "grad_norm": 7.5159430503845215,
      "learning_rate": 4.1993150586081755e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.845410218834877,
      "num_tokens": 140773817.0,
      "step": 117080
    },
    {
      "entropy": 1.8339273512363434,
      "epoch": 0.36296875520688426,
      "grad_norm": 3.6679487228393555,
      "learning_rate": 4.1991357332777346e-06,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8533756047487259,
      "num_tokens": 140786301.0,
      "step": 117090
    },
    {
      "entropy": 1.8324194088578225,
      "epoch": 0.362999754331934,
      "grad_norm": 8.492642402648926,
      "learning_rate": 4.198956430918775e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8549712345004081,
      "num_tokens": 140798870.0,
      "step": 117100
    },
    {
      "entropy": 1.8803598403930664,
      "epoch": 0.36303075345698366,
      "grad_norm": 7.195762634277344,
      "learning_rate": 4.1987771515263905e-06,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8416722387075424,
      "num_tokens": 140811461.0,
      "step": 117110
    },
    {
      "entropy": 1.9139654606580734,
      "epoch": 0.3630617525820334,
      "grad_norm": 4.575145244598389,
      "learning_rate": 4.198597895095681e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8445180609822274,
      "num_tokens": 140822762.0,
      "step": 117120
    },
    {
      "entropy": 1.9604573130607605,
      "epoch": 0.36309275170708305,
      "grad_norm": 8.761122703552246,
      "learning_rate": 4.198418661621745e-06,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8457563772797585,
      "num_tokens": 140833186.0,
      "step": 117130
    },
    {
      "entropy": 1.8488843753933906,
      "epoch": 0.3631237508321328,
      "grad_norm": 8.668008804321289,
      "learning_rate": 4.198239451099681e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.848555526137352,
      "num_tokens": 140845017.0,
      "step": 117140
    },
    {
      "entropy": 1.8061452731490135,
      "epoch": 0.36315474995718244,
      "grad_norm": 8.766512870788574,
      "learning_rate": 4.198060263524593e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8509079724550247,
      "num_tokens": 140857708.0,
      "step": 117150
    },
    {
      "entropy": 1.8900872439146041,
      "epoch": 0.36318574908223217,
      "grad_norm": 8.589384078979492,
      "learning_rate": 4.197881098891584e-06,
      "loss": 0.454,
      "mean_token_accuracy": 0.84776521474123,
      "num_tokens": 140869482.0,
      "step": 117160
    },
    {
      "entropy": 1.887955754995346,
      "epoch": 0.36321674820728184,
      "grad_norm": 8.895465850830078,
      "learning_rate": 4.197701957195758e-06,
      "loss": 0.4763,
      "mean_token_accuracy": 0.8418793961405754,
      "num_tokens": 140881457.0,
      "step": 117170
    },
    {
      "entropy": 1.9077037498354912,
      "epoch": 0.36324774733233156,
      "grad_norm": 8.192936897277832,
      "learning_rate": 4.197522838432221e-06,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8485002934932708,
      "num_tokens": 140893109.0,
      "step": 117180
    },
    {
      "entropy": 1.9021349415183066,
      "epoch": 0.36327874645738123,
      "grad_norm": 8.755243301391602,
      "learning_rate": 4.197343742596083e-06,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8397643223404885,
      "num_tokens": 140904533.0,
      "step": 117190
    },
    {
      "entropy": 1.8732059866189956,
      "epoch": 0.36330974558243095,
      "grad_norm": 8.205700874328613,
      "learning_rate": 4.19716466968245e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8494512528181076,
      "num_tokens": 140917332.0,
      "step": 117200
    },
    {
      "entropy": 1.8849352940917015,
      "epoch": 0.3633407447074806,
      "grad_norm": 3.967863082885742,
      "learning_rate": 4.196985619686435e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8578123107552529,
      "num_tokens": 140929314.0,
      "step": 117210
    },
    {
      "entropy": 1.8729075148701668,
      "epoch": 0.36337174383253035,
      "grad_norm": 8.107409477233887,
      "learning_rate": 4.196806592603149e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8556957468390465,
      "num_tokens": 140941339.0,
      "step": 117220
    },
    {
      "entropy": 1.9158906683325767,
      "epoch": 0.36340274295758,
      "grad_norm": 9.10888385772705,
      "learning_rate": 4.196627588427705e-06,
      "loss": 0.4664,
      "mean_token_accuracy": 0.849520568549633,
      "num_tokens": 140952013.0,
      "step": 117230
    },
    {
      "entropy": 1.8629322737455367,
      "epoch": 0.36343374208262974,
      "grad_norm": 7.760976314544678,
      "learning_rate": 4.196448607155221e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8583438560366631,
      "num_tokens": 140963744.0,
      "step": 117240
    },
    {
      "entropy": 1.7615326926112176,
      "epoch": 0.3634647412076794,
      "grad_norm": 8.887621879577637,
      "learning_rate": 4.19626964878081e-06,
      "loss": 0.3829,
      "mean_token_accuracy": 0.8631403967738152,
      "num_tokens": 140976932.0,
      "step": 117250
    },
    {
      "entropy": 1.8309187158942222,
      "epoch": 0.36349574033272913,
      "grad_norm": 7.472461223602295,
      "learning_rate": 4.196090713299592e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8512938529253006,
      "num_tokens": 140989695.0,
      "step": 117260
    },
    {
      "entropy": 1.8179723098874092,
      "epoch": 0.3635267394577788,
      "grad_norm": 8.319780349731445,
      "learning_rate": 4.195911800706686e-06,
      "loss": 0.4561,
      "mean_token_accuracy": 0.855165833234787,
      "num_tokens": 141001678.0,
      "step": 117270
    },
    {
      "entropy": 1.873963290452957,
      "epoch": 0.3635577385828285,
      "grad_norm": 8.117876052856445,
      "learning_rate": 4.195732910997212e-06,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8410425320267677,
      "num_tokens": 141012566.0,
      "step": 117280
    },
    {
      "entropy": 1.8815976113080979,
      "epoch": 0.3635887377078782,
      "grad_norm": 9.272675514221191,
      "learning_rate": 4.195554044166294e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8484034299850464,
      "num_tokens": 141023991.0,
      "step": 117290
    },
    {
      "entropy": 1.8928146451711654,
      "epoch": 0.36361973683292786,
      "grad_norm": 10.41171646118164,
      "learning_rate": 4.195375200209055e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8498426347970962,
      "num_tokens": 141034844.0,
      "step": 117300
    },
    {
      "entropy": 1.9140311002731323,
      "epoch": 0.3636507359579776,
      "grad_norm": 9.667838096618652,
      "learning_rate": 4.195196379120619e-06,
      "loss": 0.5428,
      "mean_token_accuracy": 0.8315603911876679,
      "num_tokens": 141045930.0,
      "step": 117310
    },
    {
      "entropy": 1.817529346048832,
      "epoch": 0.36368173508302726,
      "grad_norm": 7.76594877243042,
      "learning_rate": 4.195017580896114e-06,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8598337545990944,
      "num_tokens": 141058763.0,
      "step": 117320
    },
    {
      "entropy": 1.971425500512123,
      "epoch": 0.363712734208077,
      "grad_norm": 8.319217681884766,
      "learning_rate": 4.194838805530668e-06,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8371591001749039,
      "num_tokens": 141068907.0,
      "step": 117330
    },
    {
      "entropy": 1.9198841854929924,
      "epoch": 0.36374373333312665,
      "grad_norm": 8.907166481018066,
      "learning_rate": 4.19466005301941e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8449474439024925,
      "num_tokens": 141080433.0,
      "step": 117340
    },
    {
      "entropy": 1.8737755030393601,
      "epoch": 0.3637747324581764,
      "grad_norm": 8.013126373291016,
      "learning_rate": 4.194481323357473e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8571465671062469,
      "num_tokens": 141092637.0,
      "step": 117350
    },
    {
      "entropy": 1.9522608071565628,
      "epoch": 0.36380573158322604,
      "grad_norm": 6.9084296226501465,
      "learning_rate": 4.194302616539986e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8557817369699479,
      "num_tokens": 141103184.0,
      "step": 117360
    },
    {
      "entropy": 1.849492047727108,
      "epoch": 0.36383673070827577,
      "grad_norm": 9.977517127990723,
      "learning_rate": 4.194123932562086e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8567962691187858,
      "num_tokens": 141115374.0,
      "step": 117370
    },
    {
      "entropy": 1.8683731004595756,
      "epoch": 0.36386772983332544,
      "grad_norm": 3.932969093322754,
      "learning_rate": 4.193945271418908e-06,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8401071175932884,
      "num_tokens": 141127677.0,
      "step": 117380
    },
    {
      "entropy": 1.8759948313236237,
      "epoch": 0.36389872895837516,
      "grad_norm": 6.948093414306641,
      "learning_rate": 4.193766633105587e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8567280665040016,
      "num_tokens": 141139281.0,
      "step": 117390
    },
    {
      "entropy": 1.9030549958348275,
      "epoch": 0.36392972808342483,
      "grad_norm": 4.025951862335205,
      "learning_rate": 4.1935880176172626e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8426888778805732,
      "num_tokens": 141150437.0,
      "step": 117400
    },
    {
      "entropy": 1.790473000705242,
      "epoch": 0.36396072720847455,
      "grad_norm": 7.977076053619385,
      "learning_rate": 4.193409424949075e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8608955994248391,
      "num_tokens": 141163304.0,
      "step": 117410
    },
    {
      "entropy": 1.9095865696668626,
      "epoch": 0.3639917263335242,
      "grad_norm": 6.860559940338135,
      "learning_rate": 4.193230855096164e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8461943343281746,
      "num_tokens": 141174839.0,
      "step": 117420
    },
    {
      "entropy": 1.8187424436211586,
      "epoch": 0.36402272545857395,
      "grad_norm": 10.87106704711914,
      "learning_rate": 4.193052308053674e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8446593895554543,
      "num_tokens": 141187933.0,
      "step": 117430
    },
    {
      "entropy": 1.8248173132538796,
      "epoch": 0.3640537245836236,
      "grad_norm": 8.197173118591309,
      "learning_rate": 4.192873783816748e-06,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8538053125143051,
      "num_tokens": 141200760.0,
      "step": 117440
    },
    {
      "entropy": 1.7480739682912827,
      "epoch": 0.36408472370867334,
      "grad_norm": 5.014430046081543,
      "learning_rate": 4.192695282380531e-06,
      "loss": 0.3556,
      "mean_token_accuracy": 0.8659795328974724,
      "num_tokens": 141215184.0,
      "step": 117450
    },
    {
      "entropy": 1.9292996317148208,
      "epoch": 0.364115722833723,
      "grad_norm": 8.716943740844727,
      "learning_rate": 4.192516803740172e-06,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8481010347604752,
      "num_tokens": 141226311.0,
      "step": 117460
    },
    {
      "entropy": 1.8196698293089866,
      "epoch": 0.36414672195877273,
      "grad_norm": 6.581869602203369,
      "learning_rate": 4.192338347890818e-06,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8496568530797959,
      "num_tokens": 141238353.0,
      "step": 117470
    },
    {
      "entropy": 1.8798624217510223,
      "epoch": 0.3641777210838224,
      "grad_norm": 8.914942741394043,
      "learning_rate": 4.192159914827618e-06,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8403584629297256,
      "num_tokens": 141250412.0,
      "step": 117480
    },
    {
      "entropy": 1.9054614737629891,
      "epoch": 0.3642087202088721,
      "grad_norm": 8.384186744689941,
      "learning_rate": 4.191981504545725e-06,
      "loss": 0.497,
      "mean_token_accuracy": 0.8418497174978257,
      "num_tokens": 141261399.0,
      "step": 117490
    },
    {
      "entropy": 1.8841155782341956,
      "epoch": 0.3642397193339218,
      "grad_norm": 9.407716751098633,
      "learning_rate": 4.191803117040292e-06,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8366957977414131,
      "num_tokens": 141273160.0,
      "step": 117500
    },
    {
      "entropy": 1.82247234582901,
      "epoch": 0.3642707184589715,
      "grad_norm": 10.961734771728516,
      "learning_rate": 4.191624752306471e-06,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8399570837616921,
      "num_tokens": 141285805.0,
      "step": 117510
    },
    {
      "entropy": 1.8197978034615516,
      "epoch": 0.3643017175840212,
      "grad_norm": 9.942026138305664,
      "learning_rate": 4.191446410339419e-06,
      "loss": 0.4457,
      "mean_token_accuracy": 0.8496342837810517,
      "num_tokens": 141297836.0,
      "step": 117520
    },
    {
      "entropy": 1.8120769903063774,
      "epoch": 0.3643327167090709,
      "grad_norm": 4.632111549377441,
      "learning_rate": 4.191268091134293e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.855706425011158,
      "num_tokens": 141309962.0,
      "step": 117530
    },
    {
      "entropy": 1.9124917179346084,
      "epoch": 0.3643637158341206,
      "grad_norm": 8.424860000610352,
      "learning_rate": 4.191089794686252e-06,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8378676295280456,
      "num_tokens": 141320584.0,
      "step": 117540
    },
    {
      "entropy": 1.871908935904503,
      "epoch": 0.36439471495917025,
      "grad_norm": 3.5765280723571777,
      "learning_rate": 4.190911520990456e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8616762563586235,
      "num_tokens": 141331690.0,
      "step": 117550
    },
    {
      "entropy": 1.827473521232605,
      "epoch": 0.36442571408422,
      "grad_norm": 8.839676856994629,
      "learning_rate": 4.190733270042066e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.856446447968483,
      "num_tokens": 141344147.0,
      "step": 117560
    },
    {
      "entropy": 1.9029761299490928,
      "epoch": 0.36445671320926964,
      "grad_norm": 10.330524444580078,
      "learning_rate": 4.190555041836245e-06,
      "loss": 0.4899,
      "mean_token_accuracy": 0.8440795317292213,
      "num_tokens": 141355707.0,
      "step": 117570
    },
    {
      "entropy": 1.8510171189904212,
      "epoch": 0.36448771233431937,
      "grad_norm": 10.09470272064209,
      "learning_rate": 4.190376836368157e-06,
      "loss": 0.5337,
      "mean_token_accuracy": 0.8413840815424919,
      "num_tokens": 141367817.0,
      "step": 117580
    },
    {
      "entropy": 1.9088376432657241,
      "epoch": 0.36451871145936904,
      "grad_norm": 7.446246147155762,
      "learning_rate": 4.190198653632968e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.850058002769947,
      "num_tokens": 141378606.0,
      "step": 117590
    },
    {
      "entropy": 1.7923038378357887,
      "epoch": 0.36454971058441876,
      "grad_norm": 4.018001556396484,
      "learning_rate": 4.190020493625845e-06,
      "loss": 0.4045,
      "mean_token_accuracy": 0.8695195659995079,
      "num_tokens": 141390643.0,
      "step": 117600
    },
    {
      "entropy": 1.9072920486330986,
      "epoch": 0.36458070970946843,
      "grad_norm": 7.586421489715576,
      "learning_rate": 4.1898423563419565e-06,
      "loss": 0.496,
      "mean_token_accuracy": 0.8380471915006638,
      "num_tokens": 141401942.0,
      "step": 117610
    },
    {
      "entropy": 1.8963149681687355,
      "epoch": 0.36461170883451816,
      "grad_norm": Infinity,
      "learning_rate": 4.1896642417764735e-06,
      "loss": 0.5418,
      "mean_token_accuracy": 0.8512624606490136,
      "num_tokens": 141414153.0,
      "step": 117620
    },
    {
      "entropy": 1.8850282415747643,
      "epoch": 0.3646427079595678,
      "grad_norm": 10.361580848693848,
      "learning_rate": 4.189486149924567e-06,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8521965146064758,
      "num_tokens": 141425510.0,
      "step": 117630
    },
    {
      "entropy": 1.8708653166890143,
      "epoch": 0.36467370708461755,
      "grad_norm": 9.141669273376465,
      "learning_rate": 4.189308080781409e-06,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8514803513884545,
      "num_tokens": 141436975.0,
      "step": 117640
    },
    {
      "entropy": 1.86157566010952,
      "epoch": 0.3647047062096672,
      "grad_norm": 7.51423454284668,
      "learning_rate": 4.189130034342174e-06,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8555971041321755,
      "num_tokens": 141449229.0,
      "step": 117650
    },
    {
      "entropy": 1.8771350249648093,
      "epoch": 0.36473570533471694,
      "grad_norm": 4.558736801147461,
      "learning_rate": 4.18895201060204e-06,
      "loss": 0.5087,
      "mean_token_accuracy": 0.8356175437569618,
      "num_tokens": 141460708.0,
      "step": 117660
    },
    {
      "entropy": 1.884423953294754,
      "epoch": 0.3647667044597666,
      "grad_norm": 7.785421848297119,
      "learning_rate": 4.188774009556181e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.845032088458538,
      "num_tokens": 141472534.0,
      "step": 117670
    },
    {
      "entropy": 1.8946044102311135,
      "epoch": 0.36479770358481634,
      "grad_norm": 8.261967658996582,
      "learning_rate": 4.188596031199778e-06,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8496976479887962,
      "num_tokens": 141483714.0,
      "step": 117680
    },
    {
      "entropy": 1.7952809534966945,
      "epoch": 0.364828702709866,
      "grad_norm": 7.456904888153076,
      "learning_rate": 4.188418075528011e-06,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8519541397690773,
      "num_tokens": 141497189.0,
      "step": 117690
    },
    {
      "entropy": 1.9262281000614165,
      "epoch": 0.36485970183491573,
      "grad_norm": 8.199373245239258,
      "learning_rate": 4.188240142536061e-06,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8272052019834518,
      "num_tokens": 141508512.0,
      "step": 117700
    },
    {
      "entropy": 1.8607816338539123,
      "epoch": 0.3648907009599654,
      "grad_norm": 8.653282165527344,
      "learning_rate": 4.18806223221911e-06,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8485778480768204,
      "num_tokens": 141521091.0,
      "step": 117710
    },
    {
      "entropy": 1.9451293349266052,
      "epoch": 0.3649217000850151,
      "grad_norm": 3.9272289276123047,
      "learning_rate": 4.1878843445723445e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8386497780680656,
      "num_tokens": 141532558.0,
      "step": 117720
    },
    {
      "entropy": 1.843569315969944,
      "epoch": 0.3649526992100648,
      "grad_norm": 9.565539360046387,
      "learning_rate": 4.187706479590949e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8429495915770531,
      "num_tokens": 141545391.0,
      "step": 117730
    },
    {
      "entropy": 1.8172935619950294,
      "epoch": 0.3649836983351145,
      "grad_norm": 3.4644248485565186,
      "learning_rate": 4.18752863727011e-06,
      "loss": 0.4049,
      "mean_token_accuracy": 0.8586100369691849,
      "num_tokens": 141557495.0,
      "step": 117740
    },
    {
      "entropy": 1.8373050913214684,
      "epoch": 0.3650146974601642,
      "grad_norm": 9.560324668884277,
      "learning_rate": 4.187350817605018e-06,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8568969145417213,
      "num_tokens": 141569946.0,
      "step": 117750
    },
    {
      "entropy": 1.8132469549775123,
      "epoch": 0.3650456965852139,
      "grad_norm": 10.291223526000977,
      "learning_rate": 4.187173020590862e-06,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8503494530916214,
      "num_tokens": 141583063.0,
      "step": 117760
    },
    {
      "entropy": 1.8738684132695198,
      "epoch": 0.3650766957102636,
      "grad_norm": 7.412810802459717,
      "learning_rate": 4.186995246222834e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8561365365982055,
      "num_tokens": 141595030.0,
      "step": 117770
    },
    {
      "entropy": 1.9303662449121475,
      "epoch": 0.36510769483531325,
      "grad_norm": 7.82151985168457,
      "learning_rate": 4.186817494496126e-06,
      "loss": 0.4914,
      "mean_token_accuracy": 0.84500802308321,
      "num_tokens": 141606726.0,
      "step": 117780
    },
    {
      "entropy": 1.8065059944987296,
      "epoch": 0.36513869396036297,
      "grad_norm": 7.31696081161499,
      "learning_rate": 4.186639765405935e-06,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8545883074402809,
      "num_tokens": 141620009.0,
      "step": 117790
    },
    {
      "entropy": 1.8446330174803733,
      "epoch": 0.36516969308541264,
      "grad_norm": 8.894010543823242,
      "learning_rate": 4.186462058947453e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8492456763982773,
      "num_tokens": 141631772.0,
      "step": 117800
    },
    {
      "entropy": 1.773499608039856,
      "epoch": 0.36520069221046236,
      "grad_norm": 8.691332817077637,
      "learning_rate": 4.186284375115881e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8507530987262726,
      "num_tokens": 141646041.0,
      "step": 117810
    },
    {
      "entropy": 1.8811022609472274,
      "epoch": 0.36523169133551203,
      "grad_norm": 10.04056453704834,
      "learning_rate": 4.186106713906415e-06,
      "loss": 0.4177,
      "mean_token_accuracy": 0.8536274120211601,
      "num_tokens": 141657936.0,
      "step": 117820
    },
    {
      "entropy": 1.8523863427340985,
      "epoch": 0.36526269046056176,
      "grad_norm": 9.139519691467285,
      "learning_rate": 4.1859290753142566e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8525974780321122,
      "num_tokens": 141670704.0,
      "step": 117830
    },
    {
      "entropy": 1.8781090155243874,
      "epoch": 0.3652936895856114,
      "grad_norm": 11.72330379486084,
      "learning_rate": 4.185751459334607e-06,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8432107910513877,
      "num_tokens": 141682045.0,
      "step": 117840
    },
    {
      "entropy": 1.7702817119657994,
      "epoch": 0.36532468871066115,
      "grad_norm": 7.2551188468933105,
      "learning_rate": 4.185573865962669e-06,
      "loss": 0.4034,
      "mean_token_accuracy": 0.8568816781044006,
      "num_tokens": 141696085.0,
      "step": 117850
    },
    {
      "entropy": 1.9511079430580138,
      "epoch": 0.3653556878357108,
      "grad_norm": 9.1026029586792,
      "learning_rate": 4.185396295193647e-06,
      "loss": 0.5668,
      "mean_token_accuracy": 0.8349045425653457,
      "num_tokens": 141707166.0,
      "step": 117860
    },
    {
      "entropy": 1.8783316642045975,
      "epoch": 0.36538668696076054,
      "grad_norm": 8.064900398254395,
      "learning_rate": 4.185218747022747e-06,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8456951096653939,
      "num_tokens": 141719014.0,
      "step": 117870
    },
    {
      "entropy": 1.8309227600693703,
      "epoch": 0.3654176860858102,
      "grad_norm": 8.56662654876709,
      "learning_rate": 4.185041221445176e-06,
      "loss": 0.5239,
      "mean_token_accuracy": 0.8421471819281579,
      "num_tokens": 141732151.0,
      "step": 117880
    },
    {
      "entropy": 1.8460636466741562,
      "epoch": 0.36544868521085994,
      "grad_norm": 10.314547538757324,
      "learning_rate": 4.184863718456143e-06,
      "loss": 0.4106,
      "mean_token_accuracy": 0.8483897626399994,
      "num_tokens": 141744646.0,
      "step": 117890
    },
    {
      "entropy": 1.8618800938129425,
      "epoch": 0.3654796843359096,
      "grad_norm": 8.248429298400879,
      "learning_rate": 4.184686238050858e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8518645599484443,
      "num_tokens": 141756281.0,
      "step": 117900
    },
    {
      "entropy": 1.8451855972409248,
      "epoch": 0.36551068346095933,
      "grad_norm": 7.507343769073486,
      "learning_rate": 4.184508780224532e-06,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8565283268690109,
      "num_tokens": 141768488.0,
      "step": 117910
    },
    {
      "entropy": 1.8886141359806061,
      "epoch": 0.365541682586009,
      "grad_norm": 8.9036865234375,
      "learning_rate": 4.1843313449723795e-06,
      "loss": 0.498,
      "mean_token_accuracy": 0.8407893583178521,
      "num_tokens": 141780324.0,
      "step": 117920
    },
    {
      "entropy": 1.8497443065047263,
      "epoch": 0.3655726817110587,
      "grad_norm": 4.070679187774658,
      "learning_rate": 4.184153932289612e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8470659956336022,
      "num_tokens": 141791515.0,
      "step": 117930
    },
    {
      "entropy": 1.8582046449184417,
      "epoch": 0.3656036808361084,
      "grad_norm": 8.340042114257812,
      "learning_rate": 4.183976542171449e-06,
      "loss": 0.445,
      "mean_token_accuracy": 0.8539603292942047,
      "num_tokens": 141803922.0,
      "step": 117940
    },
    {
      "entropy": 1.8717338189482688,
      "epoch": 0.3656346799611581,
      "grad_norm": 3.3975770473480225,
      "learning_rate": 4.183799174613104e-06,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8473453000187874,
      "num_tokens": 141815267.0,
      "step": 117950
    },
    {
      "entropy": 1.821525427699089,
      "epoch": 0.3656656790862078,
      "grad_norm": 6.9430251121521,
      "learning_rate": 4.183621829609798e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8513611420989037,
      "num_tokens": 141827085.0,
      "step": 117960
    },
    {
      "entropy": 1.8612057372927666,
      "epoch": 0.3656966782112575,
      "grad_norm": 8.761266708374023,
      "learning_rate": 4.1834445071567505e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8551434949040413,
      "num_tokens": 141839369.0,
      "step": 117970
    },
    {
      "entropy": 1.7933984741568565,
      "epoch": 0.3657276773363072,
      "grad_norm": 8.286885261535645,
      "learning_rate": 4.183267207249182e-06,
      "loss": 0.4084,
      "mean_token_accuracy": 0.8545448705554008,
      "num_tokens": 141852553.0,
      "step": 117980
    },
    {
      "entropy": 1.702750214934349,
      "epoch": 0.3657586764613569,
      "grad_norm": 6.650256633758545,
      "learning_rate": 4.183089929882318e-06,
      "loss": 0.3437,
      "mean_token_accuracy": 0.8643985167145729,
      "num_tokens": 141867491.0,
      "step": 117990
    },
    {
      "entropy": 1.8753923997282982,
      "epoch": 0.36578967558640657,
      "grad_norm": 6.874874114990234,
      "learning_rate": 4.182912675051379e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8641251236200332,
      "num_tokens": 141879390.0,
      "step": 118000
    },
    {
      "entropy": 1.926875615119934,
      "epoch": 0.3658206747114563,
      "grad_norm": 6.646308422088623,
      "learning_rate": 4.182735442751594e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8443400964140892,
      "num_tokens": 141891040.0,
      "step": 118010
    },
    {
      "entropy": 1.8918773487210274,
      "epoch": 0.36585167383650596,
      "grad_norm": 6.332447528839111,
      "learning_rate": 4.182558232978188e-06,
      "loss": 0.492,
      "mean_token_accuracy": 0.8434045284986496,
      "num_tokens": 141902986.0,
      "step": 118020
    },
    {
      "entropy": 1.9184204563498497,
      "epoch": 0.36588267296155563,
      "grad_norm": 7.81022834777832,
      "learning_rate": 4.182381045726391e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.8520988777279854,
      "num_tokens": 141914126.0,
      "step": 118030
    },
    {
      "entropy": 1.9360093891620636,
      "epoch": 0.36591367208660536,
      "grad_norm": 8.875545501708984,
      "learning_rate": 4.182203880991431e-06,
      "loss": 0.5721,
      "mean_token_accuracy": 0.8364198818802834,
      "num_tokens": 141925360.0,
      "step": 118040
    },
    {
      "entropy": 1.7633348166942597,
      "epoch": 0.365944671211655,
      "grad_norm": 3.641873836517334,
      "learning_rate": 4.182026738768541e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8556516453623771,
      "num_tokens": 141939081.0,
      "step": 118050
    },
    {
      "entropy": 1.9223826959729196,
      "epoch": 0.36597567033670475,
      "grad_norm": 7.793611526489258,
      "learning_rate": 4.181849619052955e-06,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8365512430667877,
      "num_tokens": 141950887.0,
      "step": 118060
    },
    {
      "entropy": 1.910523234307766,
      "epoch": 0.3660066694617544,
      "grad_norm": 3.47296404838562,
      "learning_rate": 4.181672521839904e-06,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8425001531839371,
      "num_tokens": 141962913.0,
      "step": 118070
    },
    {
      "entropy": 1.8669334262609483,
      "epoch": 0.36603766858680414,
      "grad_norm": 7.858112812042236,
      "learning_rate": 4.1814954471246254e-06,
      "loss": 0.4452,
      "mean_token_accuracy": 0.8510735929012299,
      "num_tokens": 141974965.0,
      "step": 118080
    },
    {
      "entropy": 1.8188461996614933,
      "epoch": 0.3660686677118538,
      "grad_norm": 9.689606666564941,
      "learning_rate": 4.181318394902356e-06,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8512760177254677,
      "num_tokens": 141987859.0,
      "step": 118090
    },
    {
      "entropy": 1.8165872499346734,
      "epoch": 0.36609966683690354,
      "grad_norm": 8.783658981323242,
      "learning_rate": 4.181141365168336e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8553719744086266,
      "num_tokens": 142000311.0,
      "step": 118100
    },
    {
      "entropy": 1.747716872394085,
      "epoch": 0.3661306659619532,
      "grad_norm": 3.489741802215576,
      "learning_rate": 4.1809643579178005e-06,
      "loss": 0.3822,
      "mean_token_accuracy": 0.8664152339100838,
      "num_tokens": 142013950.0,
      "step": 118110
    },
    {
      "entropy": 1.8506772994995118,
      "epoch": 0.36616166508700293,
      "grad_norm": 7.3857316970825195,
      "learning_rate": 4.180787373145996e-06,
      "loss": 0.4218,
      "mean_token_accuracy": 0.863612025976181,
      "num_tokens": 142025425.0,
      "step": 118120
    },
    {
      "entropy": 1.8373650014400482,
      "epoch": 0.3661926642120526,
      "grad_norm": 8.373377799987793,
      "learning_rate": 4.180610410848162e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8401133880019188,
      "num_tokens": 142038565.0,
      "step": 118130
    },
    {
      "entropy": 1.893815578520298,
      "epoch": 0.3662236633371023,
      "grad_norm": 3.8234965801239014,
      "learning_rate": 4.1804334710195425e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8400477096438408,
      "num_tokens": 142050090.0,
      "step": 118140
    },
    {
      "entropy": 1.825938382744789,
      "epoch": 0.366254662462152,
      "grad_norm": 4.16547966003418,
      "learning_rate": 4.180256553655385e-06,
      "loss": 0.43,
      "mean_token_accuracy": 0.8617624938488007,
      "num_tokens": 142062662.0,
      "step": 118150
    },
    {
      "entropy": 1.801938709616661,
      "epoch": 0.3662856615872017,
      "grad_norm": 3.787094831466675,
      "learning_rate": 4.180079658750934e-06,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8575030252337456,
      "num_tokens": 142075475.0,
      "step": 118160
    },
    {
      "entropy": 1.830231237411499,
      "epoch": 0.3663166607122514,
      "grad_norm": 3.527515172958374,
      "learning_rate": 4.179902786301441e-06,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8546881586313247,
      "num_tokens": 142087897.0,
      "step": 118170
    },
    {
      "entropy": 1.8849106505513191,
      "epoch": 0.3663476598373011,
      "grad_norm": 7.191827774047852,
      "learning_rate": 4.179725936302153e-06,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8488413885235786,
      "num_tokens": 142099236.0,
      "step": 118180
    },
    {
      "entropy": 1.9353514075279237,
      "epoch": 0.3663786589623508,
      "grad_norm": 9.70740032196045,
      "learning_rate": 4.1795491087483225e-06,
      "loss": 0.5316,
      "mean_token_accuracy": 0.843434551358223,
      "num_tokens": 142110114.0,
      "step": 118190
    },
    {
      "entropy": 1.7677051618695259,
      "epoch": 0.3664096580874005,
      "grad_norm": 6.214950084686279,
      "learning_rate": 4.1793723036352e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8696540012955666,
      "num_tokens": 142123202.0,
      "step": 118200
    },
    {
      "entropy": 1.8973610028624535,
      "epoch": 0.36644065721245017,
      "grad_norm": 7.601381301879883,
      "learning_rate": 4.179195520958042e-06,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8459401935338974,
      "num_tokens": 142134423.0,
      "step": 118210
    },
    {
      "entropy": 1.877578319609165,
      "epoch": 0.3664716563374999,
      "grad_norm": 6.392949104309082,
      "learning_rate": 4.179018760712103e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8500817194581032,
      "num_tokens": 142145725.0,
      "step": 118220
    },
    {
      "entropy": 1.8080515787005424,
      "epoch": 0.36650265546254956,
      "grad_norm": 2.9493696689605713,
      "learning_rate": 4.178842022892638e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8623009279370308,
      "num_tokens": 142157978.0,
      "step": 118230
    },
    {
      "entropy": 1.8072045862674713,
      "epoch": 0.3665336545875993,
      "grad_norm": 8.654043197631836,
      "learning_rate": 4.178665307494907e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8556650042533874,
      "num_tokens": 142170679.0,
      "step": 118240
    },
    {
      "entropy": 1.8863827347755433,
      "epoch": 0.36656465371264896,
      "grad_norm": 8.228194236755371,
      "learning_rate": 4.178488614514169e-06,
      "loss": 0.4769,
      "mean_token_accuracy": 0.843590895831585,
      "num_tokens": 142182700.0,
      "step": 118250
    },
    {
      "entropy": 1.8511935338377952,
      "epoch": 0.3665956528376987,
      "grad_norm": 8.606576919555664,
      "learning_rate": 4.1783119439456844e-06,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8457527369260788,
      "num_tokens": 142194831.0,
      "step": 118260
    },
    {
      "entropy": 1.8353765562176705,
      "epoch": 0.36662665196274835,
      "grad_norm": 9.248912811279297,
      "learning_rate": 4.178135295784717e-06,
      "loss": 0.452,
      "mean_token_accuracy": 0.8504055172204972,
      "num_tokens": 142207211.0,
      "step": 118270
    },
    {
      "entropy": 1.883970457315445,
      "epoch": 0.366657651087798,
      "grad_norm": 7.958624362945557,
      "learning_rate": 4.17795867002653e-06,
      "loss": 0.5167,
      "mean_token_accuracy": 0.8472073882818222,
      "num_tokens": 142218770.0,
      "step": 118280
    },
    {
      "entropy": 1.866176003217697,
      "epoch": 0.36668865021284774,
      "grad_norm": 9.430354118347168,
      "learning_rate": 4.177782066666388e-06,
      "loss": 0.5162,
      "mean_token_accuracy": 0.8443237110972405,
      "num_tokens": 142230034.0,
      "step": 118290
    },
    {
      "entropy": 1.8722920432686805,
      "epoch": 0.3667196493378974,
      "grad_norm": 8.045259475708008,
      "learning_rate": 4.177605485699558e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8541638612747192,
      "num_tokens": 142242294.0,
      "step": 118300
    },
    {
      "entropy": 1.8110121667385102,
      "epoch": 0.36675064846294714,
      "grad_norm": 3.3392722606658936,
      "learning_rate": 4.177428927121307e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8573125541210175,
      "num_tokens": 142255463.0,
      "step": 118310
    },
    {
      "entropy": 1.826652705669403,
      "epoch": 0.3667816475879968,
      "grad_norm": 7.730444431304932,
      "learning_rate": 4.177252390926905e-06,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8576511397957802,
      "num_tokens": 142267885.0,
      "step": 118320
    },
    {
      "entropy": 1.8853152304887772,
      "epoch": 0.36681264671304653,
      "grad_norm": 7.873587131500244,
      "learning_rate": 4.1770758771116235e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8457479134202004,
      "num_tokens": 142279026.0,
      "step": 118330
    },
    {
      "entropy": 1.7946649104356767,
      "epoch": 0.3668436458380962,
      "grad_norm": 3.6855735778808594,
      "learning_rate": 4.176899385670734e-06,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8608448430895805,
      "num_tokens": 142292139.0,
      "step": 118340
    },
    {
      "entropy": 1.8492568552494049,
      "epoch": 0.3668746449631459,
      "grad_norm": 9.059473991394043,
      "learning_rate": 4.1767229165995095e-06,
      "loss": 0.4399,
      "mean_token_accuracy": 0.8556591719388962,
      "num_tokens": 142304080.0,
      "step": 118350
    },
    {
      "entropy": 1.8320197984576225,
      "epoch": 0.3669056440881956,
      "grad_norm": 7.440866470336914,
      "learning_rate": 4.176546469893225e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8550046220421791,
      "num_tokens": 142315495.0,
      "step": 118360
    },
    {
      "entropy": 1.9020672082901,
      "epoch": 0.3669366432132453,
      "grad_norm": 7.690849781036377,
      "learning_rate": 4.176370045547157e-06,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8459215462207794,
      "num_tokens": 142326196.0,
      "step": 118370
    },
    {
      "entropy": 1.8766709178686143,
      "epoch": 0.366967642338295,
      "grad_norm": 7.724151134490967,
      "learning_rate": 4.176193643556584e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8508606314659118,
      "num_tokens": 142337696.0,
      "step": 118380
    },
    {
      "entropy": 1.8881130993366242,
      "epoch": 0.3669986414633447,
      "grad_norm": 8.17661190032959,
      "learning_rate": 4.176017263916784e-06,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8483958140015602,
      "num_tokens": 142349838.0,
      "step": 118390
    },
    {
      "entropy": 1.9459423005580903,
      "epoch": 0.3670296405883944,
      "grad_norm": 8.967854499816895,
      "learning_rate": 4.17584090662304e-06,
      "loss": 0.5726,
      "mean_token_accuracy": 0.8301825195550918,
      "num_tokens": 142360518.0,
      "step": 118400
    },
    {
      "entropy": 1.889550267159939,
      "epoch": 0.3670606397134441,
      "grad_norm": 7.768711566925049,
      "learning_rate": 4.175664571670631e-06,
      "loss": 0.5122,
      "mean_token_accuracy": 0.8391052410006523,
      "num_tokens": 142372416.0,
      "step": 118410
    },
    {
      "entropy": 1.8851889297366142,
      "epoch": 0.3670916388384938,
      "grad_norm": 7.86566686630249,
      "learning_rate": 4.175488259054841e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.857041896879673,
      "num_tokens": 142384483.0,
      "step": 118420
    },
    {
      "entropy": 1.8617109119892121,
      "epoch": 0.3671226379635435,
      "grad_norm": 3.7308735847473145,
      "learning_rate": 4.175311968770956e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8486345887184144,
      "num_tokens": 142396492.0,
      "step": 118430
    },
    {
      "entropy": 1.8434831380844117,
      "epoch": 0.36715363708859317,
      "grad_norm": 3.5595266819000244,
      "learning_rate": 4.175135700814261e-06,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8539713755249977,
      "num_tokens": 142408880.0,
      "step": 118440
    },
    {
      "entropy": 1.8063078075647354,
      "epoch": 0.3671846362136429,
      "grad_norm": 7.024228096008301,
      "learning_rate": 4.174959455180043e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8617870509624481,
      "num_tokens": 142422170.0,
      "step": 118450
    },
    {
      "entropy": 1.9420251905918122,
      "epoch": 0.36721563533869256,
      "grad_norm": 7.5710039138793945,
      "learning_rate": 4.174783231863592e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8481424629688263,
      "num_tokens": 142432968.0,
      "step": 118460
    },
    {
      "entropy": 1.8173451155424118,
      "epoch": 0.3672466344637423,
      "grad_norm": 7.755475997924805,
      "learning_rate": 4.174607030860197e-06,
      "loss": 0.4012,
      "mean_token_accuracy": 0.8645861312747002,
      "num_tokens": 142446003.0,
      "step": 118470
    },
    {
      "entropy": 1.8867697641253471,
      "epoch": 0.36727763358879195,
      "grad_norm": 8.526572227478027,
      "learning_rate": 4.174430852165151e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8466065526008606,
      "num_tokens": 142457251.0,
      "step": 118480
    },
    {
      "entropy": 1.9442676544189452,
      "epoch": 0.3673086327138417,
      "grad_norm": 8.537225723266602,
      "learning_rate": 4.1742546957737465e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8544994726777076,
      "num_tokens": 142468748.0,
      "step": 118490
    },
    {
      "entropy": 1.846812443435192,
      "epoch": 0.36733963183889135,
      "grad_norm": 7.822627544403076,
      "learning_rate": 4.174078561681279e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8543079942464828,
      "num_tokens": 142480975.0,
      "step": 118500
    },
    {
      "entropy": 1.87229622900486,
      "epoch": 0.36737063096394107,
      "grad_norm": 4.54270076751709,
      "learning_rate": 4.173902449883043e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8584425121545791,
      "num_tokens": 142492336.0,
      "step": 118510
    },
    {
      "entropy": 1.889968466758728,
      "epoch": 0.36740163008899074,
      "grad_norm": 3.9654746055603027,
      "learning_rate": 4.173726360374335e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.850946743786335,
      "num_tokens": 142504212.0,
      "step": 118520
    },
    {
      "entropy": 1.8861878275871278,
      "epoch": 0.3674326292140404,
      "grad_norm": 11.20728588104248,
      "learning_rate": 4.173550293150456e-06,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8468590840697289,
      "num_tokens": 142516169.0,
      "step": 118530
    },
    {
      "entropy": 1.842411059141159,
      "epoch": 0.36746362833909013,
      "grad_norm": 8.36995792388916,
      "learning_rate": 4.1733742482067035e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8587794110178948,
      "num_tokens": 142528006.0,
      "step": 118540
    },
    {
      "entropy": 1.873109185695648,
      "epoch": 0.3674946274641398,
      "grad_norm": 8.865880012512207,
      "learning_rate": 4.173198225538381e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8478357747197152,
      "num_tokens": 142540361.0,
      "step": 118550
    },
    {
      "entropy": 1.861200602352619,
      "epoch": 0.3675256265891895,
      "grad_norm": 7.625993728637695,
      "learning_rate": 4.173022225140791e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8592444315552712,
      "num_tokens": 142553408.0,
      "step": 118560
    },
    {
      "entropy": 1.8547646909952165,
      "epoch": 0.3675566257142392,
      "grad_norm": 8.53388500213623,
      "learning_rate": 4.172846247009236e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8481455728411674,
      "num_tokens": 142566455.0,
      "step": 118570
    },
    {
      "entropy": 1.8456396833062172,
      "epoch": 0.3675876248392889,
      "grad_norm": 3.9087069034576416,
      "learning_rate": 4.1726702911390225e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8550665810704231,
      "num_tokens": 142578809.0,
      "step": 118580
    },
    {
      "entropy": 1.8159615725278855,
      "epoch": 0.3676186239643386,
      "grad_norm": 4.247369289398193,
      "learning_rate": 4.172494357525458e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8549291148781777,
      "num_tokens": 142591777.0,
      "step": 118590
    },
    {
      "entropy": 1.84511306732893,
      "epoch": 0.3676496230893883,
      "grad_norm": 4.138094425201416,
      "learning_rate": 4.17231844616385e-06,
      "loss": 0.3894,
      "mean_token_accuracy": 0.8563702628016472,
      "num_tokens": 142605209.0,
      "step": 118600
    },
    {
      "entropy": 1.9339896500110627,
      "epoch": 0.367680622214438,
      "grad_norm": 7.239840030670166,
      "learning_rate": 4.17214255704951e-06,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8529000446200371,
      "num_tokens": 142616618.0,
      "step": 118610
    },
    {
      "entropy": 1.9016248881816864,
      "epoch": 0.3677116213394877,
      "grad_norm": 10.453241348266602,
      "learning_rate": 4.171966690177746e-06,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8428125187754631,
      "num_tokens": 142628132.0,
      "step": 118620
    },
    {
      "entropy": 1.9518603295087815,
      "epoch": 0.3677426204645374,
      "grad_norm": 8.346458435058594,
      "learning_rate": 4.171790845543873e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8541428744792938,
      "num_tokens": 142639425.0,
      "step": 118630
    },
    {
      "entropy": 1.8354076534509658,
      "epoch": 0.3677736195895871,
      "grad_norm": 10.148712158203125,
      "learning_rate": 4.171615023143204e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.855410099029541,
      "num_tokens": 142651425.0,
      "step": 118640
    },
    {
      "entropy": 1.9070971041917801,
      "epoch": 0.36780461871463677,
      "grad_norm": 8.946581840515137,
      "learning_rate": 4.1714392229710536e-06,
      "loss": 0.4988,
      "mean_token_accuracy": 0.8442860797047615,
      "num_tokens": 142662410.0,
      "step": 118650
    },
    {
      "entropy": 1.8595642820000648,
      "epoch": 0.3678356178396865,
      "grad_norm": 9.433381080627441,
      "learning_rate": 4.17126344502274e-06,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8475638180971146,
      "num_tokens": 142674406.0,
      "step": 118660
    },
    {
      "entropy": 1.951467391848564,
      "epoch": 0.36786661696473616,
      "grad_norm": 7.69748592376709,
      "learning_rate": 4.17108768929358e-06,
      "loss": 0.5543,
      "mean_token_accuracy": 0.8346621260046959,
      "num_tokens": 142685520.0,
      "step": 118670
    },
    {
      "entropy": 1.8777378484606744,
      "epoch": 0.3678976160897859,
      "grad_norm": 3.798490285873413,
      "learning_rate": 4.170911955778893e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.851957768201828,
      "num_tokens": 142696814.0,
      "step": 118680
    },
    {
      "entropy": 1.8550070881843568,
      "epoch": 0.36792861521483555,
      "grad_norm": 6.965719699859619,
      "learning_rate": 4.170736244474e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8539418116211891,
      "num_tokens": 142709691.0,
      "step": 118690
    },
    {
      "entropy": 1.8812433794140815,
      "epoch": 0.3679596143398853,
      "grad_norm": 8.439732551574707,
      "learning_rate": 4.170560555374224e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8429150357842445,
      "num_tokens": 142721133.0,
      "step": 118700
    },
    {
      "entropy": 1.8941943876445293,
      "epoch": 0.36799061346493495,
      "grad_norm": 3.9443578720092773,
      "learning_rate": 4.1703848884748875e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.84272540807724,
      "num_tokens": 142734302.0,
      "step": 118710
    },
    {
      "entropy": 1.8599026456475258,
      "epoch": 0.36802161258998467,
      "grad_norm": 9.261316299438477,
      "learning_rate": 4.170209243771315e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8592417910695076,
      "num_tokens": 142745854.0,
      "step": 118720
    },
    {
      "entropy": 1.8526142731308937,
      "epoch": 0.36805261171503434,
      "grad_norm": 8.83598804473877,
      "learning_rate": 4.1700336212588336e-06,
      "loss": 0.4263,
      "mean_token_accuracy": 0.850617028772831,
      "num_tokens": 142758084.0,
      "step": 118730
    },
    {
      "entropy": 1.9292904302477836,
      "epoch": 0.36808361084008406,
      "grad_norm": 8.976339340209961,
      "learning_rate": 4.169858020932772e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8495315477252007,
      "num_tokens": 142769212.0,
      "step": 118740
    },
    {
      "entropy": 1.9031863331794738,
      "epoch": 0.36811460996513373,
      "grad_norm": 9.556485176086426,
      "learning_rate": 4.169682442788456e-06,
      "loss": 0.5131,
      "mean_token_accuracy": 0.8372549533843994,
      "num_tokens": 142780283.0,
      "step": 118750
    },
    {
      "entropy": 1.903608924150467,
      "epoch": 0.36814560909018346,
      "grad_norm": 8.73618221282959,
      "learning_rate": 4.16950688682122e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.841778826713562,
      "num_tokens": 142791778.0,
      "step": 118760
    },
    {
      "entropy": 1.8728568017482758,
      "epoch": 0.3681766082152331,
      "grad_norm": 7.941903591156006,
      "learning_rate": 4.1693313530263924e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8445895060896873,
      "num_tokens": 142804320.0,
      "step": 118770
    },
    {
      "entropy": 1.822492091357708,
      "epoch": 0.3682076073402828,
      "grad_norm": 4.276817798614502,
      "learning_rate": 4.1691558413993075e-06,
      "loss": 0.3873,
      "mean_token_accuracy": 0.8638914600014687,
      "num_tokens": 142816773.0,
      "step": 118780
    },
    {
      "entropy": 1.846034236252308,
      "epoch": 0.3682386064653325,
      "grad_norm": 9.146262168884277,
      "learning_rate": 4.168980351935301e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8438366368412972,
      "num_tokens": 142828530.0,
      "step": 118790
    },
    {
      "entropy": 1.9032179087400436,
      "epoch": 0.3682696055903822,
      "grad_norm": 8.37747573852539,
      "learning_rate": 4.168804884629708e-06,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8452135339379311,
      "num_tokens": 142839769.0,
      "step": 118800
    },
    {
      "entropy": 1.9008402451872826,
      "epoch": 0.3683006047154319,
      "grad_norm": 7.905856609344482,
      "learning_rate": 4.1686294394778654e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8504106491804123,
      "num_tokens": 142851380.0,
      "step": 118810
    },
    {
      "entropy": 1.8856339365243913,
      "epoch": 0.3683316038404816,
      "grad_norm": 6.864739894866943,
      "learning_rate": 4.168454016475113e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8480422243475914,
      "num_tokens": 142863896.0,
      "step": 118820
    },
    {
      "entropy": 1.872384211421013,
      "epoch": 0.3683626029655313,
      "grad_norm": 8.543014526367188,
      "learning_rate": 4.168278615616788e-06,
      "loss": 0.4822,
      "mean_token_accuracy": 0.8365708619356156,
      "num_tokens": 142876025.0,
      "step": 118830
    },
    {
      "entropy": 1.8250932022929192,
      "epoch": 0.368393602090581,
      "grad_norm": 3.684549570083618,
      "learning_rate": 4.168103236898236e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8542123079299927,
      "num_tokens": 142888811.0,
      "step": 118840
    },
    {
      "entropy": 1.9026092737913132,
      "epoch": 0.3684246012156307,
      "grad_norm": 7.081031322479248,
      "learning_rate": 4.167927880314796e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8558070927858352,
      "num_tokens": 142900186.0,
      "step": 118850
    },
    {
      "entropy": 1.9256568044424056,
      "epoch": 0.36845560034068037,
      "grad_norm": 7.484577178955078,
      "learning_rate": 4.167752545861815e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8413769558072091,
      "num_tokens": 142912507.0,
      "step": 118860
    },
    {
      "entropy": 1.7556425005197525,
      "epoch": 0.3684865994657301,
      "grad_norm": 8.272164344787598,
      "learning_rate": 4.167577233534637e-06,
      "loss": 0.3423,
      "mean_token_accuracy": 0.8660473376512527,
      "num_tokens": 142927368.0,
      "step": 118870
    },
    {
      "entropy": 1.9561314284801483,
      "epoch": 0.36851759859077976,
      "grad_norm": 8.683992385864258,
      "learning_rate": 4.167401943328609e-06,
      "loss": 0.5259,
      "mean_token_accuracy": 0.8412436872720719,
      "num_tokens": 142938818.0,
      "step": 118880
    },
    {
      "entropy": 1.854485747218132,
      "epoch": 0.3685485977158295,
      "grad_norm": 4.319868087768555,
      "learning_rate": 4.167226675239079e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8507488906383515,
      "num_tokens": 142951007.0,
      "step": 118890
    },
    {
      "entropy": 1.7686067208647729,
      "epoch": 0.36857959684087915,
      "grad_norm": 7.7503581047058105,
      "learning_rate": 4.167051429261398e-06,
      "loss": 0.4149,
      "mean_token_accuracy": 0.8598582178354264,
      "num_tokens": 142964704.0,
      "step": 118900
    },
    {
      "entropy": 1.898981074988842,
      "epoch": 0.3686105959659289,
      "grad_norm": 6.987488269805908,
      "learning_rate": 4.166876205390915e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8474484965205192,
      "num_tokens": 142976441.0,
      "step": 118910
    },
    {
      "entropy": 1.8347061678767205,
      "epoch": 0.36864159509097855,
      "grad_norm": 7.668900489807129,
      "learning_rate": 4.166701003622984e-06,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8599162891507148,
      "num_tokens": 142989393.0,
      "step": 118920
    },
    {
      "entropy": 1.8358816392719746,
      "epoch": 0.36867259421602827,
      "grad_norm": 8.171165466308594,
      "learning_rate": 4.166525823952959e-06,
      "loss": 0.4274,
      "mean_token_accuracy": 0.8551807060837746,
      "num_tokens": 143001626.0,
      "step": 118930
    },
    {
      "entropy": 1.7875201195478438,
      "epoch": 0.36870359334107794,
      "grad_norm": 4.123532295227051,
      "learning_rate": 4.166350666376194e-06,
      "loss": 0.4216,
      "mean_token_accuracy": 0.8510026276111603,
      "num_tokens": 143014393.0,
      "step": 118940
    },
    {
      "entropy": 1.9123052790760995,
      "epoch": 0.36873459246612766,
      "grad_norm": 9.459202766418457,
      "learning_rate": 4.1661755308880446e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8408966720104217,
      "num_tokens": 143025392.0,
      "step": 118950
    },
    {
      "entropy": 1.8176440641283989,
      "epoch": 0.36876559159117733,
      "grad_norm": 7.813144683837891,
      "learning_rate": 4.166000417483871e-06,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8539251640439034,
      "num_tokens": 143038069.0,
      "step": 118960
    },
    {
      "entropy": 1.8900042787194251,
      "epoch": 0.36879659071622706,
      "grad_norm": 3.9542362689971924,
      "learning_rate": 4.16582532615903e-06,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8539608538150787,
      "num_tokens": 143049603.0,
      "step": 118970
    },
    {
      "entropy": 1.83581335991621,
      "epoch": 0.3688275898412767,
      "grad_norm": 3.050680160522461,
      "learning_rate": 4.165650256908884e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8514253705739975,
      "num_tokens": 143062623.0,
      "step": 118980
    },
    {
      "entropy": 1.7662447020411491,
      "epoch": 0.36885858896632645,
      "grad_norm": 5.801089763641357,
      "learning_rate": 4.165475209728794e-06,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8633951663970947,
      "num_tokens": 143075670.0,
      "step": 118990
    },
    {
      "entropy": 1.8178039237856864,
      "epoch": 0.3688895880913761,
      "grad_norm": 3.8043034076690674,
      "learning_rate": 4.165300184614124e-06,
      "loss": 0.4101,
      "mean_token_accuracy": 0.8604209214448929,
      "num_tokens": 143088556.0,
      "step": 119000
    },
    {
      "entropy": 1.841179284453392,
      "epoch": 0.36892058721642584,
      "grad_norm": 7.6272292137146,
      "learning_rate": 4.1651251815602385e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8553322166204452,
      "num_tokens": 143100756.0,
      "step": 119010
    },
    {
      "entropy": 1.8461539566516876,
      "epoch": 0.3689515863414755,
      "grad_norm": 9.249650955200195,
      "learning_rate": 4.1649502005625024e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.85612553358078,
      "num_tokens": 143112605.0,
      "step": 119020
    },
    {
      "entropy": 1.902879323065281,
      "epoch": 0.3689825854665252,
      "grad_norm": 7.911184310913086,
      "learning_rate": 4.164775241616285e-06,
      "loss": 0.5353,
      "mean_token_accuracy": 0.835160045325756,
      "num_tokens": 143124403.0,
      "step": 119030
    },
    {
      "entropy": 1.9350175976753234,
      "epoch": 0.3690135845915749,
      "grad_norm": 9.767508506774902,
      "learning_rate": 4.164600304716953e-06,
      "loss": 0.5188,
      "mean_token_accuracy": 0.8444019317626953,
      "num_tokens": 143135579.0,
      "step": 119040
    },
    {
      "entropy": 1.797468328475952,
      "epoch": 0.3690445837166246,
      "grad_norm": 8.605935096740723,
      "learning_rate": 4.164425389859878e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8529829382896423,
      "num_tokens": 143148800.0,
      "step": 119050
    },
    {
      "entropy": 1.8765917882323264,
      "epoch": 0.3690755828416743,
      "grad_norm": 6.596861839294434,
      "learning_rate": 4.164250497040431e-06,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8517785176634789,
      "num_tokens": 143160367.0,
      "step": 119060
    },
    {
      "entropy": 1.8928829789161683,
      "epoch": 0.36910658196672397,
      "grad_norm": 7.310998916625977,
      "learning_rate": 4.164075626253985e-06,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8375116720795631,
      "num_tokens": 143172139.0,
      "step": 119070
    },
    {
      "entropy": 1.8701190561056138,
      "epoch": 0.3691375810917737,
      "grad_norm": 9.065203666687012,
      "learning_rate": 4.163900777495915e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8435375481843949,
      "num_tokens": 143184281.0,
      "step": 119080
    },
    {
      "entropy": 1.8811526045203208,
      "epoch": 0.36916858021682336,
      "grad_norm": 3.9514009952545166,
      "learning_rate": 4.1637259507615935e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8543221414089203,
      "num_tokens": 143196848.0,
      "step": 119090
    },
    {
      "entropy": 1.8915190115571021,
      "epoch": 0.3691995793418731,
      "grad_norm": 9.159911155700684,
      "learning_rate": 4.163551146046401e-06,
      "loss": 0.4463,
      "mean_token_accuracy": 0.8477994874119759,
      "num_tokens": 143207993.0,
      "step": 119100
    },
    {
      "entropy": 1.9243160128593444,
      "epoch": 0.36923057846692275,
      "grad_norm": 8.874519348144531,
      "learning_rate": 4.163376363345714e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8604633152484894,
      "num_tokens": 143218596.0,
      "step": 119110
    },
    {
      "entropy": 1.9454612672328948,
      "epoch": 0.3692615775919725,
      "grad_norm": 9.145526885986328,
      "learning_rate": 4.163201602654912e-06,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8401903316378594,
      "num_tokens": 143230338.0,
      "step": 119120
    },
    {
      "entropy": 1.896520721912384,
      "epoch": 0.36929257671702215,
      "grad_norm": 8.464309692382812,
      "learning_rate": 4.163026863969376e-06,
      "loss": 0.4793,
      "mean_token_accuracy": 0.8491705074906349,
      "num_tokens": 143242088.0,
      "step": 119130
    },
    {
      "entropy": 1.943563875555992,
      "epoch": 0.36932357584207187,
      "grad_norm": 7.9828033447265625,
      "learning_rate": 4.162852147284489e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8447094395756721,
      "num_tokens": 143253099.0,
      "step": 119140
    },
    {
      "entropy": 1.9148243635892868,
      "epoch": 0.36935457496712154,
      "grad_norm": 8.171924591064453,
      "learning_rate": 4.162677452595636e-06,
      "loss": 0.5437,
      "mean_token_accuracy": 0.846064169704914,
      "num_tokens": 143264931.0,
      "step": 119150
    },
    {
      "entropy": 1.9138666808605194,
      "epoch": 0.36938557409217126,
      "grad_norm": 9.261260986328125,
      "learning_rate": 4.162502779898198e-06,
      "loss": 0.512,
      "mean_token_accuracy": 0.8433482199907303,
      "num_tokens": 143275598.0,
      "step": 119160
    },
    {
      "entropy": 1.8780044361948967,
      "epoch": 0.36941657321722093,
      "grad_norm": 8.321491241455078,
      "learning_rate": 4.162328129187566e-06,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8502790480852127,
      "num_tokens": 143288132.0,
      "step": 119170
    },
    {
      "entropy": 1.9145966663956642,
      "epoch": 0.36944757234227066,
      "grad_norm": 7.993964672088623,
      "learning_rate": 4.162153500459123e-06,
      "loss": 0.4806,
      "mean_token_accuracy": 0.8421851694583893,
      "num_tokens": 143300040.0,
      "step": 119180
    },
    {
      "entropy": 1.8805766895413398,
      "epoch": 0.3694785714673203,
      "grad_norm": 8.223487854003906,
      "learning_rate": 4.161978893708263e-06,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8541955664753914,
      "num_tokens": 143311847.0,
      "step": 119190
    },
    {
      "entropy": 1.93892260491848,
      "epoch": 0.36950957059237005,
      "grad_norm": 9.87500286102295,
      "learning_rate": 4.161804308930374e-06,
      "loss": 0.548,
      "mean_token_accuracy": 0.8372417896986007,
      "num_tokens": 143324118.0,
      "step": 119200
    },
    {
      "entropy": 1.892756275832653,
      "epoch": 0.3695405697174197,
      "grad_norm": 7.675792694091797,
      "learning_rate": 4.161629746120847e-06,
      "loss": 0.4608,
      "mean_token_accuracy": 0.849764832854271,
      "num_tokens": 143335130.0,
      "step": 119210
    },
    {
      "entropy": 1.9276916295289994,
      "epoch": 0.36957156884246944,
      "grad_norm": 8.37989330291748,
      "learning_rate": 4.161455205275077e-06,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8419989302754403,
      "num_tokens": 143346444.0,
      "step": 119220
    },
    {
      "entropy": 1.846215435862541,
      "epoch": 0.3696025679675191,
      "grad_norm": 8.054888725280762,
      "learning_rate": 4.161280686388458e-06,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8463585451245308,
      "num_tokens": 143358709.0,
      "step": 119230
    },
    {
      "entropy": 1.8919729992747307,
      "epoch": 0.36963356709256884,
      "grad_norm": 8.983673095703125,
      "learning_rate": 4.161106189456385e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8458945691585541,
      "num_tokens": 143371208.0,
      "step": 119240
    },
    {
      "entropy": 1.9590601921081543,
      "epoch": 0.3696645662176185,
      "grad_norm": 9.514650344848633,
      "learning_rate": 4.160931714474256e-06,
      "loss": 0.5253,
      "mean_token_accuracy": 0.8396427750587463,
      "num_tokens": 143381837.0,
      "step": 119250
    },
    {
      "entropy": 1.8351401954889297,
      "epoch": 0.36969556534266823,
      "grad_norm": 7.223183631896973,
      "learning_rate": 4.1607572614374696e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.863115793466568,
      "num_tokens": 143393900.0,
      "step": 119260
    },
    {
      "entropy": 1.7584014266729355,
      "epoch": 0.3697265644677179,
      "grad_norm": 3.2340762615203857,
      "learning_rate": 4.160582830341426e-06,
      "loss": 0.3708,
      "mean_token_accuracy": 0.8620016917586326,
      "num_tokens": 143408131.0,
      "step": 119270
    },
    {
      "entropy": 1.9106255337595939,
      "epoch": 0.36975756359276757,
      "grad_norm": 8.975741386413574,
      "learning_rate": 4.160408421181527e-06,
      "loss": 0.493,
      "mean_token_accuracy": 0.8459363922476768,
      "num_tokens": 143419497.0,
      "step": 119280
    },
    {
      "entropy": 1.9070470571517943,
      "epoch": 0.3697885627178173,
      "grad_norm": 9.103129386901855,
      "learning_rate": 4.160234033953172e-06,
      "loss": 0.4853,
      "mean_token_accuracy": 0.848831920325756,
      "num_tokens": 143430671.0,
      "step": 119290
    },
    {
      "entropy": 1.8115619271993637,
      "epoch": 0.36981956184286696,
      "grad_norm": 9.283622741699219,
      "learning_rate": 4.160059668651768e-06,
      "loss": 0.3954,
      "mean_token_accuracy": 0.8591952756047249,
      "num_tokens": 143442619.0,
      "step": 119300
    },
    {
      "entropy": 1.8160853952169418,
      "epoch": 0.3698505609679167,
      "grad_norm": 3.4729106426239014,
      "learning_rate": 4.15988532527272e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8528380259871483,
      "num_tokens": 143454996.0,
      "step": 119310
    },
    {
      "entropy": 1.9566965460777284,
      "epoch": 0.36988156009296635,
      "grad_norm": 8.186453819274902,
      "learning_rate": 4.159711003811434e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8404311820864677,
      "num_tokens": 143466235.0,
      "step": 119320
    },
    {
      "entropy": 1.770469543337822,
      "epoch": 0.3699125592180161,
      "grad_norm": 4.339020729064941,
      "learning_rate": 4.1595367042633196e-06,
      "loss": 0.3832,
      "mean_token_accuracy": 0.8623701587319375,
      "num_tokens": 143479903.0,
      "step": 119330
    },
    {
      "entropy": 1.8619060233235358,
      "epoch": 0.36994355834306575,
      "grad_norm": 7.124758243560791,
      "learning_rate": 4.159362426623783e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8501885548233986,
      "num_tokens": 143492204.0,
      "step": 119340
    },
    {
      "entropy": 1.781793449819088,
      "epoch": 0.3699745574681155,
      "grad_norm": 6.788565635681152,
      "learning_rate": 4.159188170888238e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8653283238410949,
      "num_tokens": 143504832.0,
      "step": 119350
    },
    {
      "entropy": 1.8910044834017754,
      "epoch": 0.37000555659316514,
      "grad_norm": 9.051020622253418,
      "learning_rate": 4.159013937052096e-06,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8442646443843842,
      "num_tokens": 143516104.0,
      "step": 119360
    },
    {
      "entropy": 1.9117089003324508,
      "epoch": 0.37003655571821487,
      "grad_norm": 6.4968390464782715,
      "learning_rate": 4.158839725110769e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.852028714120388,
      "num_tokens": 143527838.0,
      "step": 119370
    },
    {
      "entropy": 1.80558120906353,
      "epoch": 0.37006755484326453,
      "grad_norm": 8.500289916992188,
      "learning_rate": 4.158665535059673e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8530550315976143,
      "num_tokens": 143541211.0,
      "step": 119380
    },
    {
      "entropy": 1.8069082364439963,
      "epoch": 0.37009855396831426,
      "grad_norm": 6.210768699645996,
      "learning_rate": 4.158491366894224e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8599958032369613,
      "num_tokens": 143553620.0,
      "step": 119390
    },
    {
      "entropy": 1.7626231014728546,
      "epoch": 0.3701295530933639,
      "grad_norm": 8.250438690185547,
      "learning_rate": 4.158317220609839e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.860263791680336,
      "num_tokens": 143567049.0,
      "step": 119400
    },
    {
      "entropy": 1.888748326897621,
      "epoch": 0.37016055221841365,
      "grad_norm": 9.457777976989746,
      "learning_rate": 4.158143096201936e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8503834918141365,
      "num_tokens": 143578643.0,
      "step": 119410
    },
    {
      "entropy": 1.910116845369339,
      "epoch": 0.3701915513434633,
      "grad_norm": 6.411842346191406,
      "learning_rate": 4.157968993665937e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8387831076979637,
      "num_tokens": 143590634.0,
      "step": 119420
    },
    {
      "entropy": 1.8767773866653443,
      "epoch": 0.37022255046851305,
      "grad_norm": 9.648838996887207,
      "learning_rate": 4.157794912997263e-06,
      "loss": 0.4942,
      "mean_token_accuracy": 0.843813206255436,
      "num_tokens": 143602357.0,
      "step": 119430
    },
    {
      "entropy": 1.9029111877083777,
      "epoch": 0.3702535495935627,
      "grad_norm": 8.24268913269043,
      "learning_rate": 4.157620854191336e-06,
      "loss": 0.5262,
      "mean_token_accuracy": 0.8355750843882561,
      "num_tokens": 143613750.0,
      "step": 119440
    },
    {
      "entropy": 1.960300487279892,
      "epoch": 0.37028454871861244,
      "grad_norm": 7.390927791595459,
      "learning_rate": 4.1574468172435805e-06,
      "loss": 0.5408,
      "mean_token_accuracy": 0.8368476778268814,
      "num_tokens": 143624528.0,
      "step": 119450
    },
    {
      "entropy": 1.9035738706588745,
      "epoch": 0.3703155478436621,
      "grad_norm": 7.557567596435547,
      "learning_rate": 4.157272802149423e-06,
      "loss": 0.509,
      "mean_token_accuracy": 0.8415401116013527,
      "num_tokens": 143636080.0,
      "step": 119460
    },
    {
      "entropy": 1.9506605803966521,
      "epoch": 0.37034654696871183,
      "grad_norm": 7.982449054718018,
      "learning_rate": 4.157098808904288e-06,
      "loss": 0.5286,
      "mean_token_accuracy": 0.8373087286949158,
      "num_tokens": 143646846.0,
      "step": 119470
    },
    {
      "entropy": 1.8253500118851662,
      "epoch": 0.3703775460937615,
      "grad_norm": 4.673305034637451,
      "learning_rate": 4.156924837503605e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8508133858442306,
      "num_tokens": 143659349.0,
      "step": 119480
    },
    {
      "entropy": 1.8085382550954818,
      "epoch": 0.3704085452188112,
      "grad_norm": 3.9146220684051514,
      "learning_rate": 4.156750887942804e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8524488687515259,
      "num_tokens": 143672699.0,
      "step": 119490
    },
    {
      "entropy": 1.8396509125828744,
      "epoch": 0.3704395443438609,
      "grad_norm": 9.417864799499512,
      "learning_rate": 4.156576960217317e-06,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8548582971096039,
      "num_tokens": 143684766.0,
      "step": 119500
    },
    {
      "entropy": 1.8102483958005906,
      "epoch": 0.37047054346891056,
      "grad_norm": 8.302236557006836,
      "learning_rate": 4.156403054322573e-06,
      "loss": 0.3942,
      "mean_token_accuracy": 0.8558089345693588,
      "num_tokens": 143696483.0,
      "step": 119510
    },
    {
      "entropy": 1.8374242320656777,
      "epoch": 0.3705015425939603,
      "grad_norm": 8.9642972946167,
      "learning_rate": 4.156229170254007e-06,
      "loss": 0.434,
      "mean_token_accuracy": 0.8551341906189919,
      "num_tokens": 143709585.0,
      "step": 119520
    },
    {
      "entropy": 1.7632934302091599,
      "epoch": 0.37053254171900996,
      "grad_norm": 8.926246643066406,
      "learning_rate": 4.156055308007056e-06,
      "loss": 0.4146,
      "mean_token_accuracy": 0.858980093896389,
      "num_tokens": 143722796.0,
      "step": 119530
    },
    {
      "entropy": 1.7985145077109337,
      "epoch": 0.3705635408440597,
      "grad_norm": 3.6693923473358154,
      "learning_rate": 4.155881467577151e-06,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8529761865735054,
      "num_tokens": 143735885.0,
      "step": 119540
    },
    {
      "entropy": 1.8460396811366082,
      "epoch": 0.37059453996910935,
      "grad_norm": 8.484942436218262,
      "learning_rate": 4.1557076489597354e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8540690630674362,
      "num_tokens": 143747844.0,
      "step": 119550
    },
    {
      "entropy": 1.8268393889069556,
      "epoch": 0.3706255390941591,
      "grad_norm": 10.119608879089355,
      "learning_rate": 4.155533852150245e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8529582008719444,
      "num_tokens": 143760082.0,
      "step": 119560
    },
    {
      "entropy": 1.8313589990139008,
      "epoch": 0.37065653821920874,
      "grad_norm": 10.065152168273926,
      "learning_rate": 4.155360077144119e-06,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8478084012866021,
      "num_tokens": 143772556.0,
      "step": 119570
    },
    {
      "entropy": 1.7658352941274642,
      "epoch": 0.37068753734425847,
      "grad_norm": 4.34824800491333,
      "learning_rate": 4.155186323936802e-06,
      "loss": 0.3614,
      "mean_token_accuracy": 0.8594838708639145,
      "num_tokens": 143786868.0,
      "step": 119580
    },
    {
      "entropy": 1.9014542356133461,
      "epoch": 0.37071853646930814,
      "grad_norm": 8.814896583557129,
      "learning_rate": 4.155012592523735e-06,
      "loss": 0.4831,
      "mean_token_accuracy": 0.844634436070919,
      "num_tokens": 143799003.0,
      "step": 119590
    },
    {
      "entropy": 1.9206546247005463,
      "epoch": 0.37074953559435786,
      "grad_norm": 7.594090461730957,
      "learning_rate": 4.154838882900362e-06,
      "loss": 0.498,
      "mean_token_accuracy": 0.8516721650958061,
      "num_tokens": 143809724.0,
      "step": 119600
    },
    {
      "entropy": 1.920139655470848,
      "epoch": 0.37078053471940753,
      "grad_norm": 6.954849720001221,
      "learning_rate": 4.154665195062129e-06,
      "loss": 0.474,
      "mean_token_accuracy": 0.8434002518653869,
      "num_tokens": 143821147.0,
      "step": 119610
    },
    {
      "entropy": 1.8785637602210046,
      "epoch": 0.37081153384445725,
      "grad_norm": 7.791470527648926,
      "learning_rate": 4.154491529004484e-06,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8430327326059341,
      "num_tokens": 143832843.0,
      "step": 119620
    },
    {
      "entropy": 1.8350661814212799,
      "epoch": 0.3708425329695069,
      "grad_norm": 9.10515308380127,
      "learning_rate": 4.1543178847228734e-06,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8528729796409606,
      "num_tokens": 143844895.0,
      "step": 119630
    },
    {
      "entropy": 1.8383238837122917,
      "epoch": 0.37087353209455665,
      "grad_norm": 7.8374834060668945,
      "learning_rate": 4.154144262212747e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8570844009518623,
      "num_tokens": 143856936.0,
      "step": 119640
    },
    {
      "entropy": 1.8358725041151047,
      "epoch": 0.3709045312196063,
      "grad_norm": 10.492681503295898,
      "learning_rate": 4.153970661469557e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8490477934479713,
      "num_tokens": 143869530.0,
      "step": 119650
    },
    {
      "entropy": 1.889818374812603,
      "epoch": 0.37093553034465604,
      "grad_norm": 8.570408821105957,
      "learning_rate": 4.153797082488753e-06,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8489762708544731,
      "num_tokens": 143880824.0,
      "step": 119660
    },
    {
      "entropy": 1.827413833141327,
      "epoch": 0.3709665294697057,
      "grad_norm": 8.959956169128418,
      "learning_rate": 4.153623525265792e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8504405677318573,
      "num_tokens": 143893214.0,
      "step": 119670
    },
    {
      "entropy": 1.8852188110351562,
      "epoch": 0.37099752859475543,
      "grad_norm": 10.366182327270508,
      "learning_rate": 4.153449989796128e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8433555126190185,
      "num_tokens": 143904446.0,
      "step": 119680
    },
    {
      "entropy": 1.9464125022292138,
      "epoch": 0.3710285277198051,
      "grad_norm": 10.189809799194336,
      "learning_rate": 4.153276476075214e-06,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8357294231653214,
      "num_tokens": 143915556.0,
      "step": 119690
    },
    {
      "entropy": 1.825826181471348,
      "epoch": 0.3710595268448548,
      "grad_norm": 8.288381576538086,
      "learning_rate": 4.1531029840985115e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8584688425064086,
      "num_tokens": 143928024.0,
      "step": 119700
    },
    {
      "entropy": 1.8943134620785713,
      "epoch": 0.3710905259699045,
      "grad_norm": 7.004955768585205,
      "learning_rate": 4.152929513861477e-06,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8523566782474518,
      "num_tokens": 143940068.0,
      "step": 119710
    },
    {
      "entropy": 1.8529562443494796,
      "epoch": 0.3711215250949542,
      "grad_norm": 3.9424831867218018,
      "learning_rate": 4.152756065359572e-06,
      "loss": 0.4032,
      "mean_token_accuracy": 0.859981244802475,
      "num_tokens": 143952445.0,
      "step": 119720
    },
    {
      "entropy": 1.8910843506455421,
      "epoch": 0.3711525242200039,
      "grad_norm": 9.168560028076172,
      "learning_rate": 4.1525826385882565e-06,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8453802585601806,
      "num_tokens": 143964559.0,
      "step": 119730
    },
    {
      "entropy": 1.8304832607507706,
      "epoch": 0.3711835233450536,
      "grad_norm": 7.770691871643066,
      "learning_rate": 4.152409233542995e-06,
      "loss": 0.481,
      "mean_token_accuracy": 0.8464579433202744,
      "num_tokens": 143976473.0,
      "step": 119740
    },
    {
      "entropy": 1.8643721297383309,
      "epoch": 0.3712145224701033,
      "grad_norm": 9.255126953125,
      "learning_rate": 4.15223585021925e-06,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8462480574846267,
      "num_tokens": 143987917.0,
      "step": 119750
    },
    {
      "entropy": 1.9292099431157113,
      "epoch": 0.37124552159515295,
      "grad_norm": 10.51768684387207,
      "learning_rate": 4.1520624886124886e-06,
      "loss": 0.5321,
      "mean_token_accuracy": 0.8432277202606201,
      "num_tokens": 144000155.0,
      "step": 119760
    },
    {
      "entropy": 1.8904988139867782,
      "epoch": 0.3712765207202027,
      "grad_norm": 6.968792915344238,
      "learning_rate": 4.151889148718177e-06,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8615191966295243,
      "num_tokens": 144012142.0,
      "step": 119770
    },
    {
      "entropy": 1.9146782085299492,
      "epoch": 0.37130751984525234,
      "grad_norm": 8.367558479309082,
      "learning_rate": 4.151715830531783e-06,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8470437869429588,
      "num_tokens": 144023416.0,
      "step": 119780
    },
    {
      "entropy": 1.8538981348276138,
      "epoch": 0.37133851897030207,
      "grad_norm": 8.344622611999512,
      "learning_rate": 4.151542534048775e-06,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8540352538228035,
      "num_tokens": 144035624.0,
      "step": 119790
    },
    {
      "entropy": 1.9207333639264106,
      "epoch": 0.37136951809535174,
      "grad_norm": 7.7748942375183105,
      "learning_rate": 4.151369259264627e-06,
      "loss": 0.505,
      "mean_token_accuracy": 0.8414986193180084,
      "num_tokens": 144047105.0,
      "step": 119800
    },
    {
      "entropy": 1.9819932192564012,
      "epoch": 0.37140051722040146,
      "grad_norm": 9.306598663330078,
      "learning_rate": 4.151196006174808e-06,
      "loss": 0.5185,
      "mean_token_accuracy": 0.845866845548153,
      "num_tokens": 144058111.0,
      "step": 119810
    },
    {
      "entropy": 1.94317606985569,
      "epoch": 0.37143151634545113,
      "grad_norm": 8.76233196258545,
      "learning_rate": 4.151022774774793e-06,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8466348692774772,
      "num_tokens": 144068820.0,
      "step": 119820
    },
    {
      "entropy": 1.8876767694950103,
      "epoch": 0.37146251547050085,
      "grad_norm": 7.308814525604248,
      "learning_rate": 4.150849565060057e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8494908258318901,
      "num_tokens": 144080689.0,
      "step": 119830
    },
    {
      "entropy": 1.8951317608356475,
      "epoch": 0.3714935145955505,
      "grad_norm": 9.297974586486816,
      "learning_rate": 4.1506763770260735e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8519510939717293,
      "num_tokens": 144092194.0,
      "step": 119840
    },
    {
      "entropy": 1.9246593773365022,
      "epoch": 0.37152451372060025,
      "grad_norm": 7.682300090789795,
      "learning_rate": 4.150503210668323e-06,
      "loss": 0.4773,
      "mean_token_accuracy": 0.84807148873806,
      "num_tokens": 144103270.0,
      "step": 119850
    },
    {
      "entropy": 1.911957061290741,
      "epoch": 0.3715555128456499,
      "grad_norm": 6.910633087158203,
      "learning_rate": 4.150330065982283e-06,
      "loss": 0.5411,
      "mean_token_accuracy": 0.8236610382795334,
      "num_tokens": 144115512.0,
      "step": 119860
    },
    {
      "entropy": 1.7777829378843308,
      "epoch": 0.37158651197069964,
      "grad_norm": 8.125778198242188,
      "learning_rate": 4.1501569429634335e-06,
      "loss": 0.3631,
      "mean_token_accuracy": 0.8673719853162766,
      "num_tokens": 144128091.0,
      "step": 119870
    },
    {
      "entropy": 1.8605340436100959,
      "epoch": 0.3716175110957493,
      "grad_norm": 9.269027709960938,
      "learning_rate": 4.149983841607256e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8562753826379776,
      "num_tokens": 144139595.0,
      "step": 119880
    },
    {
      "entropy": 1.8629012137651444,
      "epoch": 0.37164851022079903,
      "grad_norm": 8.08617877960205,
      "learning_rate": 4.149810761909232e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8640567243099213,
      "num_tokens": 144151997.0,
      "step": 119890
    },
    {
      "entropy": 1.9436404347419738,
      "epoch": 0.3716795093458487,
      "grad_norm": 9.821636199951172,
      "learning_rate": 4.149637703864848e-06,
      "loss": 0.4963,
      "mean_token_accuracy": 0.848535805940628,
      "num_tokens": 144162535.0,
      "step": 119900
    },
    {
      "entropy": 1.890118558704853,
      "epoch": 0.3717105084708984,
      "grad_norm": 8.111401557922363,
      "learning_rate": 4.149464667469588e-06,
      "loss": 0.4725,
      "mean_token_accuracy": 0.846766683459282,
      "num_tokens": 144174680.0,
      "step": 119910
    },
    {
      "entropy": 1.8793727621436118,
      "epoch": 0.3717415075959481,
      "grad_norm": 9.598828315734863,
      "learning_rate": 4.149291652718937e-06,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8412741586565972,
      "num_tokens": 144186534.0,
      "step": 119920
    },
    {
      "entropy": 1.8049995869398117,
      "epoch": 0.3717725067209978,
      "grad_norm": 4.190101623535156,
      "learning_rate": 4.149118659608385e-06,
      "loss": 0.413,
      "mean_token_accuracy": 0.8464117422699928,
      "num_tokens": 144200420.0,
      "step": 119930
    },
    {
      "entropy": 1.9019189208745957,
      "epoch": 0.3718035058460475,
      "grad_norm": 9.553948402404785,
      "learning_rate": 4.14894568813342e-06,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8418327271938324,
      "num_tokens": 144212107.0,
      "step": 119940
    },
    {
      "entropy": 1.9129536800086497,
      "epoch": 0.3718345049710972,
      "grad_norm": 8.531197547912598,
      "learning_rate": 4.1487727382895345e-06,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8501315474510193,
      "num_tokens": 144223594.0,
      "step": 119950
    },
    {
      "entropy": 1.896179696917534,
      "epoch": 0.3718655040961469,
      "grad_norm": 8.905537605285645,
      "learning_rate": 4.148599810072218e-06,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8426758080720902,
      "num_tokens": 144235324.0,
      "step": 119960
    },
    {
      "entropy": 1.9160192415118218,
      "epoch": 0.3718965032211966,
      "grad_norm": 9.644220352172852,
      "learning_rate": 4.148426903476966e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8502436727285385,
      "num_tokens": 144246974.0,
      "step": 119970
    },
    {
      "entropy": 1.8606981292366982,
      "epoch": 0.3719275023462463,
      "grad_norm": 3.5319061279296875,
      "learning_rate": 4.148254018499271e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8643532186746598,
      "num_tokens": 144259328.0,
      "step": 119980
    },
    {
      "entropy": 1.8820769846439362,
      "epoch": 0.371958501471296,
      "grad_norm": 10.861024856567383,
      "learning_rate": 4.1480811551346305e-06,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8372126758098603,
      "num_tokens": 144270870.0,
      "step": 119990
    },
    {
      "entropy": 1.8569459572434426,
      "epoch": 0.37198950059634567,
      "grad_norm": 8.557576179504395,
      "learning_rate": 4.14790831337854e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8561231017112731,
      "num_tokens": 144282985.0,
      "step": 120000
    },
    {
      "entropy": 1.8007396288216113,
      "epoch": 0.37202049972139534,
      "grad_norm": 8.916121482849121,
      "learning_rate": 4.147735493226499e-06,
      "loss": 0.3946,
      "mean_token_accuracy": 0.8646645426750184,
      "num_tokens": 144296275.0,
      "step": 120010
    },
    {
      "entropy": 1.8762442633509635,
      "epoch": 0.37205149884644506,
      "grad_norm": 10.017187118530273,
      "learning_rate": 4.147562694674007e-06,
      "loss": 0.4473,
      "mean_token_accuracy": 0.8562163576483727,
      "num_tokens": 144307269.0,
      "step": 120020
    },
    {
      "entropy": 1.945700818300247,
      "epoch": 0.37208249797149473,
      "grad_norm": 9.23812198638916,
      "learning_rate": 4.147389917716566e-06,
      "loss": 0.5345,
      "mean_token_accuracy": 0.840525084733963,
      "num_tokens": 144318155.0,
      "step": 120030
    },
    {
      "entropy": 1.9071541115641595,
      "epoch": 0.37211349709654445,
      "grad_norm": 6.633033752441406,
      "learning_rate": 4.1472171623496765e-06,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8516213580965996,
      "num_tokens": 144330070.0,
      "step": 120040
    },
    {
      "entropy": 1.8605321899056435,
      "epoch": 0.3721444962215941,
      "grad_norm": 7.755608081817627,
      "learning_rate": 4.147044428568844e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.8484739094972611,
      "num_tokens": 144342082.0,
      "step": 120050
    },
    {
      "entropy": 1.8393002033233643,
      "epoch": 0.37217549534664385,
      "grad_norm": 3.898113489151001,
      "learning_rate": 4.146871716369573e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8561500370502472,
      "num_tokens": 144353537.0,
      "step": 120060
    },
    {
      "entropy": 1.9425443708896637,
      "epoch": 0.3722064944716935,
      "grad_norm": 7.680639743804932,
      "learning_rate": 4.146699025747368e-06,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8522611662745476,
      "num_tokens": 144364505.0,
      "step": 120070
    },
    {
      "entropy": 1.8757713958621025,
      "epoch": 0.37223749359674324,
      "grad_norm": 8.204666137695312,
      "learning_rate": 4.14652635669774e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.8506177857518196,
      "num_tokens": 144375941.0,
      "step": 120080
    },
    {
      "entropy": 1.9103585094213487,
      "epoch": 0.3722684927217929,
      "grad_norm": 9.457573890686035,
      "learning_rate": 4.146353709216196e-06,
      "loss": 0.5536,
      "mean_token_accuracy": 0.8371702343225479,
      "num_tokens": 144386996.0,
      "step": 120090
    },
    {
      "entropy": 1.7752160847187042,
      "epoch": 0.37229949184684263,
      "grad_norm": 9.947811126708984,
      "learning_rate": 4.146181083298246e-06,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8600396499037742,
      "num_tokens": 144400645.0,
      "step": 120100
    },
    {
      "entropy": 1.8245454356074333,
      "epoch": 0.3723304909718923,
      "grad_norm": 10.876919746398926,
      "learning_rate": 4.146008478939402e-06,
      "loss": 0.4198,
      "mean_token_accuracy": 0.8561910197138787,
      "num_tokens": 144412719.0,
      "step": 120110
    },
    {
      "entropy": 1.8593166559934615,
      "epoch": 0.372361490096942,
      "grad_norm": 7.8572587966918945,
      "learning_rate": 4.145835896135177e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8545159190893173,
      "num_tokens": 144424962.0,
      "step": 120120
    },
    {
      "entropy": 1.9504047334194183,
      "epoch": 0.3723924892219917,
      "grad_norm": 8.473873138427734,
      "learning_rate": 4.1456633348810855e-06,
      "loss": 0.5375,
      "mean_token_accuracy": 0.8311512500047684,
      "num_tokens": 144436324.0,
      "step": 120130
    },
    {
      "entropy": 1.8700911119580268,
      "epoch": 0.3724234883470414,
      "grad_norm": 4.052376747131348,
      "learning_rate": 4.145490795172642e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8466508388519287,
      "num_tokens": 144447723.0,
      "step": 120140
    },
    {
      "entropy": 1.8484340474009513,
      "epoch": 0.3724544874720911,
      "grad_norm": 3.679072141647339,
      "learning_rate": 4.145318277005364e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.8553288698196411,
      "num_tokens": 144459623.0,
      "step": 120150
    },
    {
      "entropy": 1.8352990061044694,
      "epoch": 0.3724854865971408,
      "grad_norm": 10.292163848876953,
      "learning_rate": 4.1451457803747705e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8461218729615212,
      "num_tokens": 144471544.0,
      "step": 120160
    },
    {
      "entropy": 1.9243695348501206,
      "epoch": 0.3725164857221905,
      "grad_norm": 8.172408103942871,
      "learning_rate": 4.1449733052763785e-06,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8539075449109077,
      "num_tokens": 144482528.0,
      "step": 120170
    },
    {
      "entropy": 1.8598498031497002,
      "epoch": 0.3725474848472402,
      "grad_norm": 7.883946895599365,
      "learning_rate": 4.144800851705711e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8527406945824623,
      "num_tokens": 144493717.0,
      "step": 120180
    },
    {
      "entropy": 1.8283675089478493,
      "epoch": 0.3725784839722899,
      "grad_norm": 8.100120544433594,
      "learning_rate": 4.1446284196582885e-06,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8576804473996162,
      "num_tokens": 144505962.0,
      "step": 120190
    },
    {
      "entropy": 1.910760723054409,
      "epoch": 0.3726094830973396,
      "grad_norm": 4.125510215759277,
      "learning_rate": 4.144456009129636e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.846273559331894,
      "num_tokens": 144517019.0,
      "step": 120200
    },
    {
      "entropy": 1.83195867985487,
      "epoch": 0.37264048222238927,
      "grad_norm": 7.632051944732666,
      "learning_rate": 4.144283620115277e-06,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8484606102108956,
      "num_tokens": 144529928.0,
      "step": 120210
    },
    {
      "entropy": 1.913992887735367,
      "epoch": 0.372671481347439,
      "grad_norm": 8.024348258972168,
      "learning_rate": 4.144111252610736e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.856801763176918,
      "num_tokens": 144541124.0,
      "step": 120220
    },
    {
      "entropy": 1.7974503114819527,
      "epoch": 0.37270248047248866,
      "grad_norm": 4.225975036621094,
      "learning_rate": 4.143938906611542e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.852984607219696,
      "num_tokens": 144554372.0,
      "step": 120230
    },
    {
      "entropy": 1.8398534119129182,
      "epoch": 0.3727334795975384,
      "grad_norm": 7.547396183013916,
      "learning_rate": 4.143766582113225e-06,
      "loss": 0.45,
      "mean_token_accuracy": 0.848320834338665,
      "num_tokens": 144567059.0,
      "step": 120240
    },
    {
      "entropy": 1.79450566470623,
      "epoch": 0.37276447872258806,
      "grad_norm": 7.509959697723389,
      "learning_rate": 4.143594279111312e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8508245050907135,
      "num_tokens": 144579668.0,
      "step": 120250
    },
    {
      "entropy": 1.9366710186004639,
      "epoch": 0.3727954778476377,
      "grad_norm": 7.8771209716796875,
      "learning_rate": 4.143421997601335e-06,
      "loss": 0.5205,
      "mean_token_accuracy": 0.8401233479380608,
      "num_tokens": 144590869.0,
      "step": 120260
    },
    {
      "entropy": 1.7975213006138802,
      "epoch": 0.37282647697268745,
      "grad_norm": 10.612153053283691,
      "learning_rate": 4.1432497375788275e-06,
      "loss": 0.417,
      "mean_token_accuracy": 0.8648232862353324,
      "num_tokens": 144603350.0,
      "step": 120270
    },
    {
      "entropy": 1.882333090901375,
      "epoch": 0.3728574760977371,
      "grad_norm": 9.56601619720459,
      "learning_rate": 4.143077499039321e-06,
      "loss": 0.488,
      "mean_token_accuracy": 0.8536757841706276,
      "num_tokens": 144614705.0,
      "step": 120280
    },
    {
      "entropy": 1.8980722784996034,
      "epoch": 0.37288847522278684,
      "grad_norm": 8.230175018310547,
      "learning_rate": 4.142905281978353e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8509363621473313,
      "num_tokens": 144625834.0,
      "step": 120290
    },
    {
      "entropy": 1.881382980942726,
      "epoch": 0.3729194743478365,
      "grad_norm": 8.221864700317383,
      "learning_rate": 4.14273308639146e-06,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8483862280845642,
      "num_tokens": 144636679.0,
      "step": 120300
    },
    {
      "entropy": 1.879370318353176,
      "epoch": 0.37295047347288623,
      "grad_norm": 7.582620620727539,
      "learning_rate": 4.142560912274176e-06,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8436816439032555,
      "num_tokens": 144648082.0,
      "step": 120310
    },
    {
      "entropy": 1.8741892874240875,
      "epoch": 0.3729814725979359,
      "grad_norm": 8.145028114318848,
      "learning_rate": 4.142388759622044e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8376132413744927,
      "num_tokens": 144659565.0,
      "step": 120320
    },
    {
      "entropy": 1.8751545161008836,
      "epoch": 0.37301247172298563,
      "grad_norm": 8.928181648254395,
      "learning_rate": 4.1422166284306016e-06,
      "loss": 0.5052,
      "mean_token_accuracy": 0.8437012255191803,
      "num_tokens": 144670975.0,
      "step": 120330
    },
    {
      "entropy": 1.8685587406158448,
      "epoch": 0.3730434708480353,
      "grad_norm": 9.061539649963379,
      "learning_rate": 4.142044518695391e-06,
      "loss": 0.4559,
      "mean_token_accuracy": 0.852770148217678,
      "num_tokens": 144682788.0,
      "step": 120340
    },
    {
      "entropy": 1.7655042618513108,
      "epoch": 0.373074469973085,
      "grad_norm": 6.830787181854248,
      "learning_rate": 4.141872430411956e-06,
      "loss": 0.4142,
      "mean_token_accuracy": 0.8638226523995399,
      "num_tokens": 144695970.0,
      "step": 120350
    },
    {
      "entropy": 1.8461510226130486,
      "epoch": 0.3731054690981347,
      "grad_norm": 6.8738298416137695,
      "learning_rate": 4.14170036357584e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8553184866905212,
      "num_tokens": 144707528.0,
      "step": 120360
    },
    {
      "entropy": 1.903184102475643,
      "epoch": 0.3731364682231844,
      "grad_norm": 7.4232025146484375,
      "learning_rate": 4.1415283181825895e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8506990998983384,
      "num_tokens": 144718816.0,
      "step": 120370
    },
    {
      "entropy": 1.9077295437455177,
      "epoch": 0.3731674673482341,
      "grad_norm": 8.14012622833252,
      "learning_rate": 4.1413562942277484e-06,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8307053163647652,
      "num_tokens": 144730281.0,
      "step": 120380
    },
    {
      "entropy": 1.8349354296922684,
      "epoch": 0.3731984664732838,
      "grad_norm": 8.339512825012207,
      "learning_rate": 4.141184291706868e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8565431639552117,
      "num_tokens": 144742205.0,
      "step": 120390
    },
    {
      "entropy": 1.8753593668341637,
      "epoch": 0.3732294655983335,
      "grad_norm": 9.116205215454102,
      "learning_rate": 4.141012310615495e-06,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8513768821954727,
      "num_tokens": 144753297.0,
      "step": 120400
    },
    {
      "entropy": 1.8986528918147088,
      "epoch": 0.3732604647233832,
      "grad_norm": 7.670609951019287,
      "learning_rate": 4.140840350949181e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8434357807040215,
      "num_tokens": 144764661.0,
      "step": 120410
    },
    {
      "entropy": 1.7499748080968858,
      "epoch": 0.37329146384843287,
      "grad_norm": 8.888646125793457,
      "learning_rate": 4.140668412703478e-06,
      "loss": 0.3657,
      "mean_token_accuracy": 0.862440787255764,
      "num_tokens": 144778571.0,
      "step": 120420
    },
    {
      "entropy": 1.76868616938591,
      "epoch": 0.3733224629734826,
      "grad_norm": 9.065146446228027,
      "learning_rate": 4.14049649587394e-06,
      "loss": 0.3731,
      "mean_token_accuracy": 0.8667063027620315,
      "num_tokens": 144791265.0,
      "step": 120430
    },
    {
      "entropy": 1.8670168176293374,
      "epoch": 0.37335346209853226,
      "grad_norm": 12.608692169189453,
      "learning_rate": 4.140324600456119e-06,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8426472395658493,
      "num_tokens": 144802626.0,
      "step": 120440
    },
    {
      "entropy": 1.7094080224633217,
      "epoch": 0.373384461223582,
      "grad_norm": 9.83113956451416,
      "learning_rate": 4.140152726445574e-06,
      "loss": 0.3783,
      "mean_token_accuracy": 0.8663820832967758,
      "num_tokens": 144816924.0,
      "step": 120450
    },
    {
      "entropy": 1.88383856266737,
      "epoch": 0.37341546034863166,
      "grad_norm": 8.397429466247559,
      "learning_rate": 4.1399808738378585e-06,
      "loss": 0.4993,
      "mean_token_accuracy": 0.8410907715559006,
      "num_tokens": 144829203.0,
      "step": 120460
    },
    {
      "entropy": 1.768805581331253,
      "epoch": 0.3734464594736814,
      "grad_norm": 4.13175106048584,
      "learning_rate": 4.139809042628535e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.8638075411319732,
      "num_tokens": 144841621.0,
      "step": 120470
    },
    {
      "entropy": 1.869265778362751,
      "epoch": 0.37347745859873105,
      "grad_norm": 10.438434600830078,
      "learning_rate": 4.139637232813159e-06,
      "loss": 0.5482,
      "mean_token_accuracy": 0.830617117881775,
      "num_tokens": 144853808.0,
      "step": 120480
    },
    {
      "entropy": 1.8952225580811501,
      "epoch": 0.3735084577237808,
      "grad_norm": 7.919762134552002,
      "learning_rate": 4.139465444387294e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.851693132519722,
      "num_tokens": 144865079.0,
      "step": 120490
    },
    {
      "entropy": 1.7868720442056656,
      "epoch": 0.37353945684883044,
      "grad_norm": 8.96243953704834,
      "learning_rate": 4.139293677346502e-06,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8544852897524834,
      "num_tokens": 144878075.0,
      "step": 120500
    },
    {
      "entropy": 1.8284976735711098,
      "epoch": 0.3735704559738801,
      "grad_norm": 8.363344192504883,
      "learning_rate": 4.139121931686345e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.85543752014637,
      "num_tokens": 144890326.0,
      "step": 120510
    },
    {
      "entropy": 1.870447462797165,
      "epoch": 0.37360145509892984,
      "grad_norm": 7.410390853881836,
      "learning_rate": 4.13895020740239e-06,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8517989814281464,
      "num_tokens": 144902157.0,
      "step": 120520
    },
    {
      "entropy": 1.800754214823246,
      "epoch": 0.3736324542239795,
      "grad_norm": 4.441434383392334,
      "learning_rate": 4.138778504490201e-06,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8548402667045594,
      "num_tokens": 144915102.0,
      "step": 120530
    },
    {
      "entropy": 1.8087909892201424,
      "epoch": 0.37366345334902923,
      "grad_norm": 7.617093086242676,
      "learning_rate": 4.138606822945347e-06,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8564250499010087,
      "num_tokens": 144927518.0,
      "step": 120540
    },
    {
      "entropy": 1.8769372016191483,
      "epoch": 0.3736944524740789,
      "grad_norm": 8.41362476348877,
      "learning_rate": 4.138435162763395e-06,
      "loss": 0.4812,
      "mean_token_accuracy": 0.8450376495718956,
      "num_tokens": 144939489.0,
      "step": 120550
    },
    {
      "entropy": 1.8020145073533058,
      "epoch": 0.3737254515991286,
      "grad_norm": 9.323637008666992,
      "learning_rate": 4.138263523939916e-06,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8448305040597915,
      "num_tokens": 144952661.0,
      "step": 120560
    },
    {
      "entropy": 1.84602689743042,
      "epoch": 0.3737564507241783,
      "grad_norm": 8.894927024841309,
      "learning_rate": 4.138091906470481e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8503866240382194,
      "num_tokens": 144964533.0,
      "step": 120570
    },
    {
      "entropy": 1.7196414768695831,
      "epoch": 0.373787449849228,
      "grad_norm": 4.194610118865967,
      "learning_rate": 4.137920310350664e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8610586509108543,
      "num_tokens": 144978621.0,
      "step": 120580
    },
    {
      "entropy": 1.8428490251302718,
      "epoch": 0.3738184489742777,
      "grad_norm": 4.064561367034912,
      "learning_rate": 4.137748735576036e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8580682083964348,
      "num_tokens": 144990822.0,
      "step": 120590
    },
    {
      "entropy": 1.8954029515385629,
      "epoch": 0.3738494480993274,
      "grad_norm": 8.305089950561523,
      "learning_rate": 4.137577182142173e-06,
      "loss": 0.511,
      "mean_token_accuracy": 0.8431953400373459,
      "num_tokens": 145002259.0,
      "step": 120600
    },
    {
      "entropy": 1.7457600995898246,
      "epoch": 0.3738804472243771,
      "grad_norm": 9.313241958618164,
      "learning_rate": 4.137405650044653e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8654039755463601,
      "num_tokens": 145016219.0,
      "step": 120610
    },
    {
      "entropy": 1.8425143510103226,
      "epoch": 0.3739114463494268,
      "grad_norm": 7.391794204711914,
      "learning_rate": 4.137234139279052e-06,
      "loss": 0.4997,
      "mean_token_accuracy": 0.8488540187478065,
      "num_tokens": 145028857.0,
      "step": 120620
    },
    {
      "entropy": 1.7952088013291359,
      "epoch": 0.37394244547447647,
      "grad_norm": 9.129136085510254,
      "learning_rate": 4.13706264984095e-06,
      "loss": 0.3903,
      "mean_token_accuracy": 0.8728530630469322,
      "num_tokens": 145041106.0,
      "step": 120630
    },
    {
      "entropy": 1.9029112622141837,
      "epoch": 0.3739734445995262,
      "grad_norm": 7.712813377380371,
      "learning_rate": 4.136891181725925e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8385331735014916,
      "num_tokens": 145052581.0,
      "step": 120640
    },
    {
      "entropy": 1.8401472687721252,
      "epoch": 0.37400444372457586,
      "grad_norm": 9.776946067810059,
      "learning_rate": 4.136719734929561e-06,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8453155905008316,
      "num_tokens": 145064652.0,
      "step": 120650
    },
    {
      "entropy": 1.815567010641098,
      "epoch": 0.3740354428496256,
      "grad_norm": 9.336750984191895,
      "learning_rate": 4.136548309447441e-06,
      "loss": 0.434,
      "mean_token_accuracy": 0.8552684485912323,
      "num_tokens": 145077064.0,
      "step": 120660
    },
    {
      "entropy": 1.8268222376704215,
      "epoch": 0.37406644197467526,
      "grad_norm": 8.70843505859375,
      "learning_rate": 4.136376905275147e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8503636911511421,
      "num_tokens": 145089585.0,
      "step": 120670
    },
    {
      "entropy": 1.8132446378469467,
      "epoch": 0.374097441099725,
      "grad_norm": 8.954428672790527,
      "learning_rate": 4.1362055224082654e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8630944639444351,
      "num_tokens": 145101852.0,
      "step": 120680
    },
    {
      "entropy": 1.8896322906017304,
      "epoch": 0.37412844022477465,
      "grad_norm": 6.831114768981934,
      "learning_rate": 4.136034160842382e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.84442989975214,
      "num_tokens": 145113578.0,
      "step": 120690
    },
    {
      "entropy": 1.8678228601813316,
      "epoch": 0.3741594393498244,
      "grad_norm": 11.830345153808594,
      "learning_rate": 4.135862820573086e-06,
      "loss": 0.4919,
      "mean_token_accuracy": 0.8500665947794914,
      "num_tokens": 145125041.0,
      "step": 120700
    },
    {
      "entropy": 1.7717063777148723,
      "epoch": 0.37419043847487404,
      "grad_norm": 2.549680233001709,
      "learning_rate": 4.135691501595966e-06,
      "loss": 0.3644,
      "mean_token_accuracy": 0.8615061908960342,
      "num_tokens": 145138854.0,
      "step": 120710
    },
    {
      "entropy": 1.8354929715394974,
      "epoch": 0.37422143759992377,
      "grad_norm": 3.5742509365081787,
      "learning_rate": 4.1355202039066126e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8558676704764366,
      "num_tokens": 145150497.0,
      "step": 120720
    },
    {
      "entropy": 1.8448586538434029,
      "epoch": 0.37425243672497344,
      "grad_norm": 8.689122200012207,
      "learning_rate": 4.135348927500618e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8484915763139724,
      "num_tokens": 145162443.0,
      "step": 120730
    },
    {
      "entropy": 1.8981183648109436,
      "epoch": 0.37428343585002316,
      "grad_norm": 8.72005844116211,
      "learning_rate": 4.135177672373573e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8429804682731629,
      "num_tokens": 145174166.0,
      "step": 120740
    },
    {
      "entropy": 1.9157382816076278,
      "epoch": 0.37431443497507283,
      "grad_norm": 3.737830400466919,
      "learning_rate": 4.135006438521074e-06,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8430665746331215,
      "num_tokens": 145185172.0,
      "step": 120750
    },
    {
      "entropy": 1.9020906642079354,
      "epoch": 0.3743454341001225,
      "grad_norm": 7.51852560043335,
      "learning_rate": 4.134835225938717e-06,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8433557212352752,
      "num_tokens": 145196378.0,
      "step": 120760
    },
    {
      "entropy": 1.8677937388420105,
      "epoch": 0.3743764332251722,
      "grad_norm": 3.965975284576416,
      "learning_rate": 4.134664034622098e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8616374880075455,
      "num_tokens": 145207527.0,
      "step": 120770
    },
    {
      "entropy": 1.8505346715450286,
      "epoch": 0.3744074323502219,
      "grad_norm": 4.419517993927002,
      "learning_rate": 4.134492864566814e-06,
      "loss": 0.4014,
      "mean_token_accuracy": 0.8657907411456108,
      "num_tokens": 145219338.0,
      "step": 120780
    },
    {
      "entropy": 1.7792392835021018,
      "epoch": 0.3744384314752716,
      "grad_norm": 4.039781093597412,
      "learning_rate": 4.134321715768466e-06,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8500527799129486,
      "num_tokens": 145232885.0,
      "step": 120790
    },
    {
      "entropy": 1.870184451341629,
      "epoch": 0.3744694306003213,
      "grad_norm": 3.6062183380126953,
      "learning_rate": 4.134150588222653e-06,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8536598861217499,
      "num_tokens": 145244298.0,
      "step": 120800
    },
    {
      "entropy": 1.7577032819390297,
      "epoch": 0.374500429725371,
      "grad_norm": 3.414278030395508,
      "learning_rate": 4.13397948192498e-06,
      "loss": 0.391,
      "mean_token_accuracy": 0.8604967936873436,
      "num_tokens": 145257392.0,
      "step": 120810
    },
    {
      "entropy": 1.8007031008601189,
      "epoch": 0.3745314288504207,
      "grad_norm": 4.170672416687012,
      "learning_rate": 4.133808396871046e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8616224005818367,
      "num_tokens": 145270196.0,
      "step": 120820
    },
    {
      "entropy": 1.9025542974472045,
      "epoch": 0.3745624279754704,
      "grad_norm": 7.770366191864014,
      "learning_rate": 4.133637333056459e-06,
      "loss": 0.4694,
      "mean_token_accuracy": 0.8512863472104073,
      "num_tokens": 145281494.0,
      "step": 120830
    },
    {
      "entropy": 1.8452581293880939,
      "epoch": 0.37459342710052007,
      "grad_norm": 8.3046293258667,
      "learning_rate": 4.1334662904768234e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8516708582639694,
      "num_tokens": 145294224.0,
      "step": 120840
    },
    {
      "entropy": 1.7171327024698257,
      "epoch": 0.3746244262255698,
      "grad_norm": 3.3472342491149902,
      "learning_rate": 4.1332952691277465e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.8607262581586838,
      "num_tokens": 145308338.0,
      "step": 120850
    },
    {
      "entropy": 1.9564589619636537,
      "epoch": 0.37465542535061946,
      "grad_norm": 7.92958402633667,
      "learning_rate": 4.133124269004837e-06,
      "loss": 0.5199,
      "mean_token_accuracy": 0.8395091354846954,
      "num_tokens": 145319683.0,
      "step": 120860
    },
    {
      "entropy": 1.8489731594920158,
      "epoch": 0.3746864244756692,
      "grad_norm": 8.664722442626953,
      "learning_rate": 4.132953290103704e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8534695088863373,
      "num_tokens": 145332281.0,
      "step": 120870
    },
    {
      "entropy": 1.8806229501962661,
      "epoch": 0.37471742360071886,
      "grad_norm": 9.805295944213867,
      "learning_rate": 4.132782332419957e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8593287214636802,
      "num_tokens": 145343989.0,
      "step": 120880
    },
    {
      "entropy": 1.8413976445794105,
      "epoch": 0.3747484227257686,
      "grad_norm": 3.8515052795410156,
      "learning_rate": 4.13261139594921e-06,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8467993855476379,
      "num_tokens": 145356520.0,
      "step": 120890
    },
    {
      "entropy": 1.9287862733006478,
      "epoch": 0.37477942185081825,
      "grad_norm": 8.918717384338379,
      "learning_rate": 4.132440480687076e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.854530693590641,
      "num_tokens": 145367592.0,
      "step": 120900
    },
    {
      "entropy": 1.909818661212921,
      "epoch": 0.374810420975868,
      "grad_norm": 11.631319999694824,
      "learning_rate": 4.13226958662917e-06,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8506747603416442,
      "num_tokens": 145378561.0,
      "step": 120910
    },
    {
      "entropy": 1.8912340737879276,
      "epoch": 0.37484142010091764,
      "grad_norm": 9.222434997558594,
      "learning_rate": 4.132098713771106e-06,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8446800738573075,
      "num_tokens": 145390493.0,
      "step": 120920
    },
    {
      "entropy": 1.891737161576748,
      "epoch": 0.37487241922596737,
      "grad_norm": 9.400554656982422,
      "learning_rate": 4.131927862108504e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8476646468043327,
      "num_tokens": 145401638.0,
      "step": 120930
    },
    {
      "entropy": 1.8229492217302323,
      "epoch": 0.37490341835101704,
      "grad_norm": 8.848627090454102,
      "learning_rate": 4.13175703163698e-06,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8519455164670944,
      "num_tokens": 145414244.0,
      "step": 120940
    },
    {
      "entropy": 1.8781214132905006,
      "epoch": 0.37493441747606676,
      "grad_norm": 9.196588516235352,
      "learning_rate": 4.131586222352156e-06,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8479282841086387,
      "num_tokens": 145425934.0,
      "step": 120950
    },
    {
      "entropy": 1.8320988327264787,
      "epoch": 0.37496541660111643,
      "grad_norm": 8.844073295593262,
      "learning_rate": 4.131415434249652e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8547179445624351,
      "num_tokens": 145438172.0,
      "step": 120960
    },
    {
      "entropy": 1.8629610985517502,
      "epoch": 0.37499641572616615,
      "grad_norm": 4.892116546630859,
      "learning_rate": 4.131244667325089e-06,
      "loss": 0.441,
      "mean_token_accuracy": 0.8539141744375229,
      "num_tokens": 145449705.0,
      "step": 120970
    },
    {
      "entropy": 1.8123175263404847,
      "epoch": 0.3750274148512158,
      "grad_norm": 3.7633779048919678,
      "learning_rate": 4.131073921574093e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8505129277706146,
      "num_tokens": 145463148.0,
      "step": 120980
    },
    {
      "entropy": 1.9299910217523575,
      "epoch": 0.3750584139762655,
      "grad_norm": 10.863778114318848,
      "learning_rate": 4.130903196992287e-06,
      "loss": 0.5118,
      "mean_token_accuracy": 0.8390725657343865,
      "num_tokens": 145474645.0,
      "step": 120990
    },
    {
      "entropy": 1.8196517676115036,
      "epoch": 0.3750894131013152,
      "grad_norm": 4.3333210945129395,
      "learning_rate": 4.1307324935752964e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8509810373187066,
      "num_tokens": 145487342.0,
      "step": 121000
    },
    {
      "entropy": 1.8983831241726876,
      "epoch": 0.3751204122263649,
      "grad_norm": 3.389133930206299,
      "learning_rate": 4.130561811318752e-06,
      "loss": 0.5603,
      "mean_token_accuracy": 0.839329156279564,
      "num_tokens": 145499254.0,
      "step": 121010
    },
    {
      "entropy": 1.7906238347291947,
      "epoch": 0.3751514113514146,
      "grad_norm": 8.32974624633789,
      "learning_rate": 4.1303911502182784e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8561747640371322,
      "num_tokens": 145511933.0,
      "step": 121020
    },
    {
      "entropy": 1.943038222193718,
      "epoch": 0.3751824104764643,
      "grad_norm": 8.438972473144531,
      "learning_rate": 4.130220510269507e-06,
      "loss": 0.4776,
      "mean_token_accuracy": 0.8507806181907653,
      "num_tokens": 145522873.0,
      "step": 121030
    },
    {
      "entropy": 1.9753375679254532,
      "epoch": 0.375213409601514,
      "grad_norm": 6.82116174697876,
      "learning_rate": 4.1300498914680705e-06,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8418274581432342,
      "num_tokens": 145533983.0,
      "step": 121040
    },
    {
      "entropy": 1.8580348297953606,
      "epoch": 0.37524440872656367,
      "grad_norm": 9.025100708007812,
      "learning_rate": 4.129879293809599e-06,
      "loss": 0.4782,
      "mean_token_accuracy": 0.840544268488884,
      "num_tokens": 145546293.0,
      "step": 121050
    },
    {
      "entropy": 1.8665331095457076,
      "epoch": 0.3752754078516134,
      "grad_norm": 4.244259834289551,
      "learning_rate": 4.129708717289727e-06,
      "loss": 0.4255,
      "mean_token_accuracy": 0.856557646393776,
      "num_tokens": 145557776.0,
      "step": 121060
    },
    {
      "entropy": 1.8694847792387008,
      "epoch": 0.37530640697666307,
      "grad_norm": 8.377169609069824,
      "learning_rate": 4.12953816190409e-06,
      "loss": 0.4456,
      "mean_token_accuracy": 0.857828502357006,
      "num_tokens": 145569154.0,
      "step": 121070
    },
    {
      "entropy": 1.7808043763041497,
      "epoch": 0.3753374061017128,
      "grad_norm": 4.1414570808410645,
      "learning_rate": 4.1293676276483235e-06,
      "loss": 0.3795,
      "mean_token_accuracy": 0.8570607796311378,
      "num_tokens": 145583082.0,
      "step": 121080
    },
    {
      "entropy": 1.9220749616622925,
      "epoch": 0.37536840522676246,
      "grad_norm": 8.00906753540039,
      "learning_rate": 4.1291971145180645e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8506890282034874,
      "num_tokens": 145594524.0,
      "step": 121090
    },
    {
      "entropy": 1.9267291516065597,
      "epoch": 0.3753994043518122,
      "grad_norm": 8.382123947143555,
      "learning_rate": 4.129026622508953e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8558269619941712,
      "num_tokens": 145605078.0,
      "step": 121100
    },
    {
      "entropy": 1.7520041272044182,
      "epoch": 0.37543040347686185,
      "grad_norm": 8.029170989990234,
      "learning_rate": 4.128856151616628e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8598437607288361,
      "num_tokens": 145618913.0,
      "step": 121110
    },
    {
      "entropy": 1.8347510501742363,
      "epoch": 0.3754614026019116,
      "grad_norm": 9.550764083862305,
      "learning_rate": 4.128685701836731e-06,
      "loss": 0.4574,
      "mean_token_accuracy": 0.8477082312107086,
      "num_tokens": 145631433.0,
      "step": 121120
    },
    {
      "entropy": 1.8798089861869811,
      "epoch": 0.37549240172696124,
      "grad_norm": 8.3656005859375,
      "learning_rate": 4.128515273164905e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.8539829984307289,
      "num_tokens": 145642408.0,
      "step": 121130
    },
    {
      "entropy": 1.7175431214272976,
      "epoch": 0.37552340085201097,
      "grad_norm": 8.398488998413086,
      "learning_rate": 4.128344865596793e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.8567556262016296,
      "num_tokens": 145656156.0,
      "step": 121140
    },
    {
      "entropy": 1.7876494243741035,
      "epoch": 0.37555439997706064,
      "grad_norm": 9.021903038024902,
      "learning_rate": 4.12817447912804e-06,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8566848263144493,
      "num_tokens": 145667977.0,
      "step": 121150
    },
    {
      "entropy": 1.7991645127534865,
      "epoch": 0.37558539910211036,
      "grad_norm": 8.89357852935791,
      "learning_rate": 4.128004113754292e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8592987477779388,
      "num_tokens": 145680773.0,
      "step": 121160
    },
    {
      "entropy": 1.8806771039962769,
      "epoch": 0.37561639822716003,
      "grad_norm": 7.232542514801025,
      "learning_rate": 4.127833769471198e-06,
      "loss": 0.5171,
      "mean_token_accuracy": 0.8432414412498475,
      "num_tokens": 145692228.0,
      "step": 121170
    },
    {
      "entropy": 1.8988723203539848,
      "epoch": 0.37564739735220976,
      "grad_norm": 9.036412239074707,
      "learning_rate": 4.127663446274406e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8560154557228088,
      "num_tokens": 145703505.0,
      "step": 121180
    },
    {
      "entropy": 1.876578164100647,
      "epoch": 0.3756783964772594,
      "grad_norm": 8.154003143310547,
      "learning_rate": 4.1274931441595645e-06,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8526700213551521,
      "num_tokens": 145714806.0,
      "step": 121190
    },
    {
      "entropy": 1.831437037885189,
      "epoch": 0.37570939560230915,
      "grad_norm": 8.418209075927734,
      "learning_rate": 4.1273228631223275e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8543977767229081,
      "num_tokens": 145726566.0,
      "step": 121200
    },
    {
      "entropy": 1.7670956924557686,
      "epoch": 0.3757403947273588,
      "grad_norm": 8.465004920959473,
      "learning_rate": 4.1271526031583445e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8540250107645988,
      "num_tokens": 145739210.0,
      "step": 121210
    },
    {
      "entropy": 1.7944332778453826,
      "epoch": 0.37577139385240854,
      "grad_norm": 8.313122749328613,
      "learning_rate": 4.126982364263272e-06,
      "loss": 0.467,
      "mean_token_accuracy": 0.8499791830778122,
      "num_tokens": 145751513.0,
      "step": 121220
    },
    {
      "entropy": 1.8116312876343728,
      "epoch": 0.3758023929774582,
      "grad_norm": 3.7453999519348145,
      "learning_rate": 4.126812146432764e-06,
      "loss": 0.3914,
      "mean_token_accuracy": 0.8603421568870544,
      "num_tokens": 145763916.0,
      "step": 121230
    },
    {
      "entropy": 1.9280547127127647,
      "epoch": 0.3758333921025079,
      "grad_norm": 8.784612655639648,
      "learning_rate": 4.126641949662477e-06,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8478071898221969,
      "num_tokens": 145774772.0,
      "step": 121240
    },
    {
      "entropy": 1.9599653363227845,
      "epoch": 0.3758643912275576,
      "grad_norm": 9.786565780639648,
      "learning_rate": 4.126471773948068e-06,
      "loss": 0.5111,
      "mean_token_accuracy": 0.8443946361541748,
      "num_tokens": 145785993.0,
      "step": 121250
    },
    {
      "entropy": 1.8358087345957756,
      "epoch": 0.3758953903526073,
      "grad_norm": 8.388903617858887,
      "learning_rate": 4.126301619285196e-06,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8493312045931816,
      "num_tokens": 145798151.0,
      "step": 121260
    },
    {
      "entropy": 1.8409698456525803,
      "epoch": 0.375926389477657,
      "grad_norm": 9.696656227111816,
      "learning_rate": 4.126131485669522e-06,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8447024121880531,
      "num_tokens": 145810340.0,
      "step": 121270
    },
    {
      "entropy": 1.9255991280078888,
      "epoch": 0.37595738860270667,
      "grad_norm": 9.015790939331055,
      "learning_rate": 4.125961373096706e-06,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8415984019637108,
      "num_tokens": 145821505.0,
      "step": 121280
    },
    {
      "entropy": 1.9274761855602265,
      "epoch": 0.3759883877277564,
      "grad_norm": 7.261139392852783,
      "learning_rate": 4.125791281562412e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.853644534945488,
      "num_tokens": 145832550.0,
      "step": 121290
    },
    {
      "entropy": 1.838864254951477,
      "epoch": 0.37601938685280606,
      "grad_norm": 4.556879043579102,
      "learning_rate": 4.125621211062301e-06,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8556353718042373,
      "num_tokens": 145844415.0,
      "step": 121300
    },
    {
      "entropy": 1.9384997293353081,
      "epoch": 0.3760503859778558,
      "grad_norm": 6.320328712463379,
      "learning_rate": 4.125451161592041e-06,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8399920925498009,
      "num_tokens": 145855907.0,
      "step": 121310
    },
    {
      "entropy": 1.8460743427276611,
      "epoch": 0.37608138510290545,
      "grad_norm": 8.746512413024902,
      "learning_rate": 4.125281133147298e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.865053367614746,
      "num_tokens": 145867251.0,
      "step": 121320
    },
    {
      "entropy": 1.8502772450447083,
      "epoch": 0.3761123842279552,
      "grad_norm": 4.59258508682251,
      "learning_rate": 4.125111125723738e-06,
      "loss": 0.4174,
      "mean_token_accuracy": 0.8603545278310776,
      "num_tokens": 145879356.0,
      "step": 121330
    },
    {
      "entropy": 1.8849507763981819,
      "epoch": 0.37614338335300485,
      "grad_norm": 7.458291530609131,
      "learning_rate": 4.124941139317031e-06,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8583846285939216,
      "num_tokens": 145891689.0,
      "step": 121340
    },
    {
      "entropy": 1.8615009844303132,
      "epoch": 0.37617438247805457,
      "grad_norm": 8.805126190185547,
      "learning_rate": 4.124771173922848e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8532016187906265,
      "num_tokens": 145903559.0,
      "step": 121350
    },
    {
      "entropy": 1.81989781036973,
      "epoch": 0.37620538160310424,
      "grad_norm": 7.289465427398682,
      "learning_rate": 4.1246012295368575e-06,
      "loss": 0.425,
      "mean_token_accuracy": 0.8597825333476067,
      "num_tokens": 145916356.0,
      "step": 121360
    },
    {
      "entropy": 1.8147448897361755,
      "epoch": 0.37623638072815396,
      "grad_norm": 7.9857497215271,
      "learning_rate": 4.124431306154735e-06,
      "loss": 0.423,
      "mean_token_accuracy": 0.8589125216007233,
      "num_tokens": 145928998.0,
      "step": 121370
    },
    {
      "entropy": 1.9180129885673523,
      "epoch": 0.37626737985320363,
      "grad_norm": 9.49499797821045,
      "learning_rate": 4.124261403772152e-06,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8353124216198922,
      "num_tokens": 145940800.0,
      "step": 121380
    },
    {
      "entropy": 1.920377929508686,
      "epoch": 0.37629837897825336,
      "grad_norm": 9.39726448059082,
      "learning_rate": 4.124091522384785e-06,
      "loss": 0.5386,
      "mean_token_accuracy": 0.834184144437313,
      "num_tokens": 145952846.0,
      "step": 121390
    },
    {
      "entropy": 1.9227221325039863,
      "epoch": 0.376329378103303,
      "grad_norm": 8.036456108093262,
      "learning_rate": 4.1239216619883106e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.85260841101408,
      "num_tokens": 145964633.0,
      "step": 121400
    },
    {
      "entropy": 1.880432690680027,
      "epoch": 0.37636037722835275,
      "grad_norm": 9.900726318359375,
      "learning_rate": 4.123751822578405e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8541592597961426,
      "num_tokens": 145976465.0,
      "step": 121410
    },
    {
      "entropy": 1.845744264125824,
      "epoch": 0.3763913763534024,
      "grad_norm": 8.883930206298828,
      "learning_rate": 4.123582004150748e-06,
      "loss": 0.4641,
      "mean_token_accuracy": 0.8476935580372811,
      "num_tokens": 145988619.0,
      "step": 121420
    },
    {
      "entropy": 1.8941200897097588,
      "epoch": 0.37642237547845214,
      "grad_norm": 7.9326910972595215,
      "learning_rate": 4.123412206701019e-06,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8494398832321167,
      "num_tokens": 145999813.0,
      "step": 121430
    },
    {
      "entropy": 1.950559636950493,
      "epoch": 0.3764533746035018,
      "grad_norm": 7.880441665649414,
      "learning_rate": 4.1232424302249e-06,
      "loss": 0.5265,
      "mean_token_accuracy": 0.8358272299170494,
      "num_tokens": 146010044.0,
      "step": 121440
    },
    {
      "entropy": 1.8937064185738564,
      "epoch": 0.37648437372855154,
      "grad_norm": 7.3297438621521,
      "learning_rate": 4.123072674718073e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8427557229995728,
      "num_tokens": 146021669.0,
      "step": 121450
    },
    {
      "entropy": 1.7439885810017586,
      "epoch": 0.3765153728536012,
      "grad_norm": 2.4287400245666504,
      "learning_rate": 4.122902940176222e-06,
      "loss": 0.3597,
      "mean_token_accuracy": 0.8698983415961266,
      "num_tokens": 146035437.0,
      "step": 121460
    },
    {
      "entropy": 1.8418650522828102,
      "epoch": 0.37654637197865093,
      "grad_norm": 8.513551712036133,
      "learning_rate": 4.122733226595032e-06,
      "loss": 0.434,
      "mean_token_accuracy": 0.856321020424366,
      "num_tokens": 146047669.0,
      "step": 121470
    },
    {
      "entropy": 1.897158458828926,
      "epoch": 0.3765773711037006,
      "grad_norm": 8.884608268737793,
      "learning_rate": 4.122563533970189e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8471438035368919,
      "num_tokens": 146058873.0,
      "step": 121480
    },
    {
      "entropy": 1.8239916279911994,
      "epoch": 0.37660837022875027,
      "grad_norm": 7.655562877655029,
      "learning_rate": 4.122393862297381e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8447648495435714,
      "num_tokens": 146071178.0,
      "step": 121490
    },
    {
      "entropy": 1.8941697224974632,
      "epoch": 0.3766393693538,
      "grad_norm": 3.823326349258423,
      "learning_rate": 4.122224211572297e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8591320559382438,
      "num_tokens": 146082489.0,
      "step": 121500
    },
    {
      "entropy": 1.8416035562753676,
      "epoch": 0.37667036847884966,
      "grad_norm": 7.697302341461182,
      "learning_rate": 4.122054581790626e-06,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8484307572245597,
      "num_tokens": 146094870.0,
      "step": 121510
    },
    {
      "entropy": 1.7722488626837731,
      "epoch": 0.3767013676038994,
      "grad_norm": 6.725828170776367,
      "learning_rate": 4.121884972948061e-06,
      "loss": 0.3802,
      "mean_token_accuracy": 0.8639165312051773,
      "num_tokens": 146108574.0,
      "step": 121520
    },
    {
      "entropy": 1.8973381593823433,
      "epoch": 0.37673236672894905,
      "grad_norm": 8.68867301940918,
      "learning_rate": 4.121715385040292e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8411045700311661,
      "num_tokens": 146120800.0,
      "step": 121530
    },
    {
      "entropy": 1.929002758860588,
      "epoch": 0.3767633658539988,
      "grad_norm": 7.320843696594238,
      "learning_rate": 4.1215458180630136e-06,
      "loss": 0.4931,
      "mean_token_accuracy": 0.8451986953616142,
      "num_tokens": 146132050.0,
      "step": 121540
    },
    {
      "entropy": 1.850935024023056,
      "epoch": 0.37679436497904845,
      "grad_norm": 3.584969997406006,
      "learning_rate": 4.121376272011922e-06,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8525848358869552,
      "num_tokens": 146144075.0,
      "step": 121550
    },
    {
      "entropy": 1.8184596046805381,
      "epoch": 0.37682536410409817,
      "grad_norm": 8.334314346313477,
      "learning_rate": 4.121206746882713e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8496242105960846,
      "num_tokens": 146156909.0,
      "step": 121560
    },
    {
      "entropy": 1.845918568968773,
      "epoch": 0.37685636322914784,
      "grad_norm": 7.1232805252075195,
      "learning_rate": 4.121037242671082e-06,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8491025045514107,
      "num_tokens": 146169187.0,
      "step": 121570
    },
    {
      "entropy": 1.8747231513261795,
      "epoch": 0.37688736235419756,
      "grad_norm": 8.403467178344727,
      "learning_rate": 4.12086775937273e-06,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8543328627943992,
      "num_tokens": 146181009.0,
      "step": 121580
    },
    {
      "entropy": 1.9061955615878106,
      "epoch": 0.37691836147924723,
      "grad_norm": 9.928679466247559,
      "learning_rate": 4.120698296983356e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8431771412491799,
      "num_tokens": 146192524.0,
      "step": 121590
    },
    {
      "entropy": 1.8643248841166495,
      "epoch": 0.37694936060429696,
      "grad_norm": 6.862342834472656,
      "learning_rate": 4.120528855498663e-06,
      "loss": 0.3942,
      "mean_token_accuracy": 0.8685376390814781,
      "num_tokens": 146204768.0,
      "step": 121600
    },
    {
      "entropy": 1.8592729791998863,
      "epoch": 0.3769803597293466,
      "grad_norm": 8.456074714660645,
      "learning_rate": 4.12035943491435e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8440333098173142,
      "num_tokens": 146217180.0,
      "step": 121610
    },
    {
      "entropy": 1.801856230199337,
      "epoch": 0.37701135885439635,
      "grad_norm": 4.020124435424805,
      "learning_rate": 4.120190035226123e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.859604449570179,
      "num_tokens": 146230138.0,
      "step": 121620
    },
    {
      "entropy": 1.8506782591342925,
      "epoch": 0.377042357979446,
      "grad_norm": 7.4072041511535645,
      "learning_rate": 4.120020656429685e-06,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8540009498596192,
      "num_tokens": 146242786.0,
      "step": 121630
    },
    {
      "entropy": 1.914864283800125,
      "epoch": 0.37707335710449574,
      "grad_norm": 7.623665809631348,
      "learning_rate": 4.119851298520746e-06,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8540341034531593,
      "num_tokens": 146253407.0,
      "step": 121640
    },
    {
      "entropy": 1.9302624300122262,
      "epoch": 0.3771043562295454,
      "grad_norm": 9.16896915435791,
      "learning_rate": 4.119681961495008e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8566766142845154,
      "num_tokens": 146264715.0,
      "step": 121650
    },
    {
      "entropy": 1.8666496634483338,
      "epoch": 0.37713535535459514,
      "grad_norm": 3.110356092453003,
      "learning_rate": 4.119512645348184e-06,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8514965951442719,
      "num_tokens": 146276968.0,
      "step": 121660
    },
    {
      "entropy": 1.8799085780978202,
      "epoch": 0.3771663544796448,
      "grad_norm": 4.057291030883789,
      "learning_rate": 4.119343350075981e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8615951225161552,
      "num_tokens": 146288576.0,
      "step": 121670
    },
    {
      "entropy": 1.818842075765133,
      "epoch": 0.37719735360469453,
      "grad_norm": 10.600461959838867,
      "learning_rate": 4.119174075674112e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.852581399679184,
      "num_tokens": 146300167.0,
      "step": 121680
    },
    {
      "entropy": 1.8106214493513106,
      "epoch": 0.3772283527297442,
      "grad_norm": 3.652275323867798,
      "learning_rate": 4.119004822138288e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.854650741815567,
      "num_tokens": 146312797.0,
      "step": 121690
    },
    {
      "entropy": 1.783032974600792,
      "epoch": 0.3772593518547939,
      "grad_norm": 2.3988730907440186,
      "learning_rate": 4.118835589464222e-06,
      "loss": 0.3963,
      "mean_token_accuracy": 0.8633334279060364,
      "num_tokens": 146325854.0,
      "step": 121700
    },
    {
      "entropy": 1.8655279122292996,
      "epoch": 0.3772903509798436,
      "grad_norm": 3.8433516025543213,
      "learning_rate": 4.118666377647631e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.8409992963075638,
      "num_tokens": 146338473.0,
      "step": 121710
    },
    {
      "entropy": 1.9374163687229156,
      "epoch": 0.3773213501048933,
      "grad_norm": 7.3235249519348145,
      "learning_rate": 4.118497186684229e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8448733389377594,
      "num_tokens": 146349721.0,
      "step": 121720
    },
    {
      "entropy": 1.8571631446480752,
      "epoch": 0.377352349229943,
      "grad_norm": 4.183323860168457,
      "learning_rate": 4.118328016569734e-06,
      "loss": 0.47,
      "mean_token_accuracy": 0.8527215197682381,
      "num_tokens": 146362284.0,
      "step": 121730
    },
    {
      "entropy": 1.9317335307598114,
      "epoch": 0.37738334835499265,
      "grad_norm": 7.954257965087891,
      "learning_rate": 4.118158867299864e-06,
      "loss": 0.537,
      "mean_token_accuracy": 0.835789829492569,
      "num_tokens": 146373357.0,
      "step": 121740
    },
    {
      "entropy": 1.8835597395896913,
      "epoch": 0.3774143474800424,
      "grad_norm": 8.603527069091797,
      "learning_rate": 4.117989738870338e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8524842366576195,
      "num_tokens": 146384815.0,
      "step": 121750
    },
    {
      "entropy": 1.8527083411812781,
      "epoch": 0.37744534660509205,
      "grad_norm": 15.195161819458008,
      "learning_rate": 4.117820631276879e-06,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8620884269475937,
      "num_tokens": 146396638.0,
      "step": 121760
    },
    {
      "entropy": 1.8090012684464454,
      "epoch": 0.37747634573014177,
      "grad_norm": 9.310328483581543,
      "learning_rate": 4.117651544515207e-06,
      "loss": 0.3812,
      "mean_token_accuracy": 0.8572055116295815,
      "num_tokens": 146409088.0,
      "step": 121770
    },
    {
      "entropy": 1.8714495450258255,
      "epoch": 0.37750734485519144,
      "grad_norm": 8.089369773864746,
      "learning_rate": 4.117482478581047e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.8630315214395523,
      "num_tokens": 146421044.0,
      "step": 121780
    },
    {
      "entropy": 1.8696634009480477,
      "epoch": 0.37753834398024116,
      "grad_norm": 4.10288667678833,
      "learning_rate": 4.117313433470123e-06,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8499879941344262,
      "num_tokens": 146432640.0,
      "step": 121790
    },
    {
      "entropy": 1.8695901393890382,
      "epoch": 0.37756934310529083,
      "grad_norm": 7.528837203979492,
      "learning_rate": 4.11714440917816e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8490506067872048,
      "num_tokens": 146444477.0,
      "step": 121800
    },
    {
      "entropy": 1.8624702721834183,
      "epoch": 0.37760034223034056,
      "grad_norm": 7.586561679840088,
      "learning_rate": 4.116975405700887e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.853815546631813,
      "num_tokens": 146457039.0,
      "step": 121810
    },
    {
      "entropy": 1.917418046295643,
      "epoch": 0.3776313413553902,
      "grad_norm": 7.109292030334473,
      "learning_rate": 4.116806423034029e-06,
      "loss": 0.4573,
      "mean_token_accuracy": 0.854109600186348,
      "num_tokens": 146469106.0,
      "step": 121820
    },
    {
      "entropy": 1.8877900682389737,
      "epoch": 0.37766234048043995,
      "grad_norm": 2.658243417739868,
      "learning_rate": 4.116637461173319e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8463023707270623,
      "num_tokens": 146482224.0,
      "step": 121830
    },
    {
      "entropy": 1.9093093752861023,
      "epoch": 0.3776933396054896,
      "grad_norm": 7.953685283660889,
      "learning_rate": 4.116468520114486e-06,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8459949418902397,
      "num_tokens": 146494016.0,
      "step": 121840
    },
    {
      "entropy": 1.8482676953077317,
      "epoch": 0.37772433873053934,
      "grad_norm": 5.989774703979492,
      "learning_rate": 4.116299599853262e-06,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8606431394815445,
      "num_tokens": 146507063.0,
      "step": 121850
    },
    {
      "entropy": 1.6848430618643762,
      "epoch": 0.377755337855589,
      "grad_norm": 9.330160140991211,
      "learning_rate": 4.11613070038538e-06,
      "loss": 0.3608,
      "mean_token_accuracy": 0.874959084391594,
      "num_tokens": 146521265.0,
      "step": 121860
    },
    {
      "entropy": 1.8134713634848594,
      "epoch": 0.37778633698063874,
      "grad_norm": 9.139564514160156,
      "learning_rate": 4.115961821706575e-06,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8542681246995926,
      "num_tokens": 146533625.0,
      "step": 121870
    },
    {
      "entropy": 1.8329403929412365,
      "epoch": 0.3778173361056884,
      "grad_norm": 2.2994585037231445,
      "learning_rate": 4.115792963812582e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8562861248850823,
      "num_tokens": 146546747.0,
      "step": 121880
    },
    {
      "entropy": 1.8927996829152107,
      "epoch": 0.37784833523073813,
      "grad_norm": 8.54615306854248,
      "learning_rate": 4.115624126699139e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8517010971903801,
      "num_tokens": 146559042.0,
      "step": 121890
    },
    {
      "entropy": 1.878790758550167,
      "epoch": 0.3778793343557878,
      "grad_norm": 10.510698318481445,
      "learning_rate": 4.1154553103619835e-06,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8425199687480927,
      "num_tokens": 146570240.0,
      "step": 121900
    },
    {
      "entropy": 1.8623479381203651,
      "epoch": 0.3779103334808375,
      "grad_norm": 8.871630668640137,
      "learning_rate": 4.115286514796853e-06,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8559099569916725,
      "num_tokens": 146583049.0,
      "step": 121910
    },
    {
      "entropy": 1.8767142370343208,
      "epoch": 0.3779413326058872,
      "grad_norm": 3.710120677947998,
      "learning_rate": 4.115117739999491e-06,
      "loss": 0.5138,
      "mean_token_accuracy": 0.847675909101963,
      "num_tokens": 146595103.0,
      "step": 121920
    },
    {
      "entropy": 1.960743510723114,
      "epoch": 0.3779723317309369,
      "grad_norm": 7.444425582885742,
      "learning_rate": 4.114948985965637e-06,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8415925487875938,
      "num_tokens": 146605922.0,
      "step": 121930
    },
    {
      "entropy": 1.8839008912444115,
      "epoch": 0.3780033308559866,
      "grad_norm": 9.009261131286621,
      "learning_rate": 4.114780252691036e-06,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8500188961625099,
      "num_tokens": 146617318.0,
      "step": 121940
    },
    {
      "entropy": 1.89123537838459,
      "epoch": 0.3780343299810363,
      "grad_norm": 7.541473865509033,
      "learning_rate": 4.11461154017143e-06,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8585240036249161,
      "num_tokens": 146628643.0,
      "step": 121950
    },
    {
      "entropy": 1.9025135144591332,
      "epoch": 0.378065329106086,
      "grad_norm": 6.690996170043945,
      "learning_rate": 4.114442848402565e-06,
      "loss": 0.5119,
      "mean_token_accuracy": 0.8481958448886872,
      "num_tokens": 146640271.0,
      "step": 121960
    },
    {
      "entropy": 1.956543865799904,
      "epoch": 0.3780963282311357,
      "grad_norm": 8.179557800292969,
      "learning_rate": 4.1142741773801894e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8471057340502739,
      "num_tokens": 146650293.0,
      "step": 121970
    },
    {
      "entropy": 1.8972681686282158,
      "epoch": 0.37812732735618537,
      "grad_norm": 8.407889366149902,
      "learning_rate": 4.1141055271000485e-06,
      "loss": 0.4785,
      "mean_token_accuracy": 0.845854164659977,
      "num_tokens": 146662226.0,
      "step": 121980
    },
    {
      "entropy": 1.9091035678982735,
      "epoch": 0.37815832648123504,
      "grad_norm": 7.07802152633667,
      "learning_rate": 4.113936897557893e-06,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8480123952031136,
      "num_tokens": 146674239.0,
      "step": 121990
    },
    {
      "entropy": 1.906779918074608,
      "epoch": 0.37818932560628477,
      "grad_norm": 10.43696403503418,
      "learning_rate": 4.113768288749473e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8478205010294915,
      "num_tokens": 146686512.0,
      "step": 122000
    },
    {
      "entropy": 1.9320502176880836,
      "epoch": 0.37822032473133443,
      "grad_norm": 9.740495681762695,
      "learning_rate": 4.113599700670539e-06,
      "loss": 0.4757,
      "mean_token_accuracy": 0.8459899872541428,
      "num_tokens": 146697859.0,
      "step": 122010
    },
    {
      "entropy": 1.9591059625148772,
      "epoch": 0.37825132385638416,
      "grad_norm": 8.180789947509766,
      "learning_rate": 4.113431133316846e-06,
      "loss": 0.4797,
      "mean_token_accuracy": 0.8437874868512154,
      "num_tokens": 146709083.0,
      "step": 122020
    },
    {
      "entropy": 1.802576979994774,
      "epoch": 0.3782823229814338,
      "grad_norm": 9.705636024475098,
      "learning_rate": 4.113262586684146e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8546473920345307,
      "num_tokens": 146722550.0,
      "step": 122030
    },
    {
      "entropy": 1.8501103572547435,
      "epoch": 0.37831332210648355,
      "grad_norm": 7.220241069793701,
      "learning_rate": 4.113094060768193e-06,
      "loss": 0.4619,
      "mean_token_accuracy": 0.8407644376158714,
      "num_tokens": 146736130.0,
      "step": 122040
    },
    {
      "entropy": 1.847523419559002,
      "epoch": 0.3783443212315332,
      "grad_norm": 7.426035404205322,
      "learning_rate": 4.112925555564747e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8541855216026306,
      "num_tokens": 146748100.0,
      "step": 122050
    },
    {
      "entropy": 1.7743535205721854,
      "epoch": 0.37837532035658294,
      "grad_norm": 4.075657367706299,
      "learning_rate": 4.112757071069562e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8489448487758636,
      "num_tokens": 146760893.0,
      "step": 122060
    },
    {
      "entropy": 1.853407160937786,
      "epoch": 0.3784063194816326,
      "grad_norm": 10.630528450012207,
      "learning_rate": 4.1125886072784e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8410938188433648,
      "num_tokens": 146773477.0,
      "step": 122070
    },
    {
      "entropy": 1.831977953016758,
      "epoch": 0.37843731860668234,
      "grad_norm": 9.587197303771973,
      "learning_rate": 4.112420164187019e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.856691500544548,
      "num_tokens": 146784991.0,
      "step": 122080
    },
    {
      "entropy": 1.8454594373703004,
      "epoch": 0.378468317731732,
      "grad_norm": 8.84549331665039,
      "learning_rate": 4.11225174179118e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8553299933671952,
      "num_tokens": 146797158.0,
      "step": 122090
    },
    {
      "entropy": 1.889687070250511,
      "epoch": 0.37849931685678173,
      "grad_norm": 8.103165626525879,
      "learning_rate": 4.1120833400866465e-06,
      "loss": 0.4901,
      "mean_token_accuracy": 0.8467686668038368,
      "num_tokens": 146808540.0,
      "step": 122100
    },
    {
      "entropy": 1.7636478379368783,
      "epoch": 0.3785303159818314,
      "grad_norm": 8.541621208190918,
      "learning_rate": 4.111914959069182e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8589817076921463,
      "num_tokens": 146822024.0,
      "step": 122110
    },
    {
      "entropy": 1.8948402941226958,
      "epoch": 0.3785613151068811,
      "grad_norm": 7.753279685974121,
      "learning_rate": 4.111746598734551e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8480223119258881,
      "num_tokens": 146833733.0,
      "step": 122120
    },
    {
      "entropy": 1.8317367523908614,
      "epoch": 0.3785923142319308,
      "grad_norm": 9.554505348205566,
      "learning_rate": 4.11157825907852e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.853913576900959,
      "num_tokens": 146846482.0,
      "step": 122130
    },
    {
      "entropy": 1.8885809898376464,
      "epoch": 0.3786233133569805,
      "grad_norm": 8.104804039001465,
      "learning_rate": 4.111409940096856e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8428619146347046,
      "num_tokens": 146857645.0,
      "step": 122140
    },
    {
      "entropy": 1.9016202330589294,
      "epoch": 0.3786543124820302,
      "grad_norm": 8.457221031188965,
      "learning_rate": 4.111241641785328e-06,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8547043219208718,
      "num_tokens": 146868691.0,
      "step": 122150
    },
    {
      "entropy": 1.8326540157198905,
      "epoch": 0.3786853116070799,
      "grad_norm": 8.81411361694336,
      "learning_rate": 4.111073364139704e-06,
      "loss": 0.4304,
      "mean_token_accuracy": 0.8510780692100525,
      "num_tokens": 146881779.0,
      "step": 122160
    },
    {
      "entropy": 1.8807669505476952,
      "epoch": 0.3787163107321296,
      "grad_norm": 6.89314603805542,
      "learning_rate": 4.110905107155758e-06,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8529252395033836,
      "num_tokens": 146893089.0,
      "step": 122170
    },
    {
      "entropy": 1.8773617178201676,
      "epoch": 0.3787473098571793,
      "grad_norm": 9.011570930480957,
      "learning_rate": 4.110736870829259e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8588549971580506,
      "num_tokens": 146904814.0,
      "step": 122180
    },
    {
      "entropy": 1.8098601162433625,
      "epoch": 0.378778308982229,
      "grad_norm": 4.172004222869873,
      "learning_rate": 4.110568655155984e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8595098748803138,
      "num_tokens": 146917001.0,
      "step": 122190
    },
    {
      "entropy": 1.9448596596717835,
      "epoch": 0.3788093081072787,
      "grad_norm": 9.438422203063965,
      "learning_rate": 4.110400460131704e-06,
      "loss": 0.5126,
      "mean_token_accuracy": 0.8462162002921104,
      "num_tokens": 146928045.0,
      "step": 122200
    },
    {
      "entropy": 1.8534026801586152,
      "epoch": 0.37884030723232837,
      "grad_norm": 7.917478561401367,
      "learning_rate": 4.110232285752197e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8487755730748177,
      "num_tokens": 146939455.0,
      "step": 122210
    },
    {
      "entropy": 1.88424232006073,
      "epoch": 0.3788713063573781,
      "grad_norm": 9.62708568572998,
      "learning_rate": 4.110064132013238e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8592986583709716,
      "num_tokens": 146951083.0,
      "step": 122220
    },
    {
      "entropy": 1.9062071546912194,
      "epoch": 0.37890230548242776,
      "grad_norm": 7.392994403839111,
      "learning_rate": 4.109895998910608e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8547699064016342,
      "num_tokens": 146962391.0,
      "step": 122230
    },
    {
      "entropy": 1.8592015653848648,
      "epoch": 0.37893330460747743,
      "grad_norm": 8.539904594421387,
      "learning_rate": 4.109727886440085e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8498929470777512,
      "num_tokens": 146974481.0,
      "step": 122240
    },
    {
      "entropy": 1.8191335678100586,
      "epoch": 0.37896430373252715,
      "grad_norm": 8.98744010925293,
      "learning_rate": 4.109559794597449e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8560615256428719,
      "num_tokens": 146986678.0,
      "step": 122250
    },
    {
      "entropy": 1.8230956450104714,
      "epoch": 0.3789953028575768,
      "grad_norm": 5.308132171630859,
      "learning_rate": 4.109391723378483e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8582112580537796,
      "num_tokens": 146999024.0,
      "step": 122260
    },
    {
      "entropy": 1.890698865056038,
      "epoch": 0.37902630198262655,
      "grad_norm": 8.043314933776855,
      "learning_rate": 4.109223672778969e-06,
      "loss": 0.4737,
      "mean_token_accuracy": 0.851884001493454,
      "num_tokens": 147011194.0,
      "step": 122270
    },
    {
      "entropy": 1.8781690523028374,
      "epoch": 0.3790573011076762,
      "grad_norm": 3.952792167663574,
      "learning_rate": 4.1090556427946905e-06,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8536352217197418,
      "num_tokens": 147022928.0,
      "step": 122280
    },
    {
      "entropy": 1.8398768171668052,
      "epoch": 0.37908830023272594,
      "grad_norm": 4.229760646820068,
      "learning_rate": 4.108887633421435e-06,
      "loss": 0.4094,
      "mean_token_accuracy": 0.8582636058330536,
      "num_tokens": 147035202.0,
      "step": 122290
    },
    {
      "entropy": 1.8784176617860795,
      "epoch": 0.3791192993577756,
      "grad_norm": 4.562318801879883,
      "learning_rate": 4.10871964465499e-06,
      "loss": 0.3901,
      "mean_token_accuracy": 0.8587144047021866,
      "num_tokens": 147047405.0,
      "step": 122300
    },
    {
      "entropy": 1.8330136485397817,
      "epoch": 0.37915029848282533,
      "grad_norm": 7.896421432495117,
      "learning_rate": 4.108551676491141e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8706734910607338,
      "num_tokens": 147060914.0,
      "step": 122310
    },
    {
      "entropy": 1.9075976356863975,
      "epoch": 0.379181297607875,
      "grad_norm": 9.339140892028809,
      "learning_rate": 4.108383728925676e-06,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8520648717880249,
      "num_tokens": 147072427.0,
      "step": 122320
    },
    {
      "entropy": 1.8779551640152932,
      "epoch": 0.3792122967329247,
      "grad_norm": 8.027301788330078,
      "learning_rate": 4.108215801954389e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8487935498356819,
      "num_tokens": 147085063.0,
      "step": 122330
    },
    {
      "entropy": 1.9310110673308372,
      "epoch": 0.3792432958579744,
      "grad_norm": 8.90977668762207,
      "learning_rate": 4.108047895573069e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8506239086389542,
      "num_tokens": 147096264.0,
      "step": 122340
    },
    {
      "entropy": 1.9274428904056549,
      "epoch": 0.3792742949830241,
      "grad_norm": 8.3607759475708,
      "learning_rate": 4.107880009777509e-06,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8582290455698967,
      "num_tokens": 147108064.0,
      "step": 122350
    },
    {
      "entropy": 1.931325614452362,
      "epoch": 0.3793052941080738,
      "grad_norm": 7.63694429397583,
      "learning_rate": 4.1077121445635036e-06,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8499775350093841,
      "num_tokens": 147120037.0,
      "step": 122360
    },
    {
      "entropy": 1.8099196195602416,
      "epoch": 0.3793362932331235,
      "grad_norm": 3.8295421600341797,
      "learning_rate": 4.107544299926848e-06,
      "loss": 0.4353,
      "mean_token_accuracy": 0.8536288380622864,
      "num_tokens": 147132561.0,
      "step": 122370
    },
    {
      "entropy": 1.8952490270137787,
      "epoch": 0.3793672923581732,
      "grad_norm": 3.9062981605529785,
      "learning_rate": 4.107376475863337e-06,
      "loss": 0.4945,
      "mean_token_accuracy": 0.845606592297554,
      "num_tokens": 147144257.0,
      "step": 122380
    },
    {
      "entropy": 1.8194826439023017,
      "epoch": 0.3793982914832229,
      "grad_norm": 2.3818023204803467,
      "learning_rate": 4.10720867236877e-06,
      "loss": 0.5514,
      "mean_token_accuracy": 0.8439658433198929,
      "num_tokens": 147158022.0,
      "step": 122390
    },
    {
      "entropy": 1.8628859788179397,
      "epoch": 0.3794292906082726,
      "grad_norm": 9.438950538635254,
      "learning_rate": 4.107040889438944e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8520619332790375,
      "num_tokens": 147171213.0,
      "step": 122400
    },
    {
      "entropy": 1.8283814415335655,
      "epoch": 0.3794602897333223,
      "grad_norm": 8.620655059814453,
      "learning_rate": 4.10687312706966e-06,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8559184551239014,
      "num_tokens": 147183776.0,
      "step": 122410
    },
    {
      "entropy": 1.9575203657150269,
      "epoch": 0.37949128885837197,
      "grad_norm": 9.457103729248047,
      "learning_rate": 4.106705385256718e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8479184478521347,
      "num_tokens": 147194326.0,
      "step": 122420
    },
    {
      "entropy": 1.832918418943882,
      "epoch": 0.3795222879834217,
      "grad_norm": 4.032721519470215,
      "learning_rate": 4.106537663995922e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.857351616024971,
      "num_tokens": 147207032.0,
      "step": 122430
    },
    {
      "entropy": 1.7993885815143584,
      "epoch": 0.37955328710847136,
      "grad_norm": 4.224924087524414,
      "learning_rate": 4.106369963283075e-06,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8557244941592217,
      "num_tokens": 147220135.0,
      "step": 122440
    },
    {
      "entropy": 1.8836615636944771,
      "epoch": 0.3795842862335211,
      "grad_norm": 7.664379596710205,
      "learning_rate": 4.106202283113981e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8581904530525207,
      "num_tokens": 147232750.0,
      "step": 122450
    },
    {
      "entropy": 1.9607893019914626,
      "epoch": 0.37961528535857075,
      "grad_norm": 8.335966110229492,
      "learning_rate": 4.106034623484447e-06,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8552387475967407,
      "num_tokens": 147244068.0,
      "step": 122460
    },
    {
      "entropy": 1.8356236606836318,
      "epoch": 0.3796462844836205,
      "grad_norm": 6.946023941040039,
      "learning_rate": 4.105866984390278e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8576385736465454,
      "num_tokens": 147256513.0,
      "step": 122470
    },
    {
      "entropy": 1.9411771401762963,
      "epoch": 0.37967728360867015,
      "grad_norm": 7.646897792816162,
      "learning_rate": 4.105699365827284e-06,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8406426265835762,
      "num_tokens": 147267715.0,
      "step": 122480
    },
    {
      "entropy": 1.8944382056593896,
      "epoch": 0.3797082827337198,
      "grad_norm": 2.604159355163574,
      "learning_rate": 4.105531767791274e-06,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8485357999801636,
      "num_tokens": 147279090.0,
      "step": 122490
    },
    {
      "entropy": 1.8582454279065133,
      "epoch": 0.37973928185876954,
      "grad_norm": 7.28471565246582,
      "learning_rate": 4.105364190278059e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8588834837079048,
      "num_tokens": 147291484.0,
      "step": 122500
    },
    {
      "entropy": 1.9929537415504455,
      "epoch": 0.3797702809838192,
      "grad_norm": 10.385540962219238,
      "learning_rate": 4.105196633283452e-06,
      "loss": 0.5533,
      "mean_token_accuracy": 0.8405931279063225,
      "num_tokens": 147302573.0,
      "step": 122510
    },
    {
      "entropy": 1.8923503875732421,
      "epoch": 0.37980128010886893,
      "grad_norm": 8.61189079284668,
      "learning_rate": 4.1050290968032635e-06,
      "loss": 0.441,
      "mean_token_accuracy": 0.8514659553766251,
      "num_tokens": 147314906.0,
      "step": 122520
    },
    {
      "entropy": 1.8648876428604126,
      "epoch": 0.3798322792339186,
      "grad_norm": 8.03481388092041,
      "learning_rate": 4.10486158083331e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8421670094132423,
      "num_tokens": 147327317.0,
      "step": 122530
    },
    {
      "entropy": 1.780473567545414,
      "epoch": 0.3798632783589683,
      "grad_norm": 8.313706398010254,
      "learning_rate": 4.104694085369405e-06,
      "loss": 0.3914,
      "mean_token_accuracy": 0.8570443943142891,
      "num_tokens": 147340984.0,
      "step": 122540
    },
    {
      "entropy": 1.9261644035577774,
      "epoch": 0.379894277484018,
      "grad_norm": 7.854968070983887,
      "learning_rate": 4.104526610407367e-06,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8521699935197831,
      "num_tokens": 147352717.0,
      "step": 122550
    },
    {
      "entropy": 1.8588500022888184,
      "epoch": 0.3799252766090677,
      "grad_norm": 3.6017353534698486,
      "learning_rate": 4.104359155943014e-06,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8543045312166214,
      "num_tokens": 147365365.0,
      "step": 122560
    },
    {
      "entropy": 1.8084479227662087,
      "epoch": 0.3799562757341174,
      "grad_norm": 3.5251076221466064,
      "learning_rate": 4.104191721972163e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8611663445830345,
      "num_tokens": 147378567.0,
      "step": 122570
    },
    {
      "entropy": 1.849086406826973,
      "epoch": 0.3799872748591671,
      "grad_norm": 4.066585063934326,
      "learning_rate": 4.104024308490636e-06,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8478951990604401,
      "num_tokens": 147390692.0,
      "step": 122580
    },
    {
      "entropy": 1.900680673122406,
      "epoch": 0.3800182739842168,
      "grad_norm": 10.499907493591309,
      "learning_rate": 4.103856915494254e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8512833088636398,
      "num_tokens": 147402469.0,
      "step": 122590
    },
    {
      "entropy": 1.8157877206802369,
      "epoch": 0.3800492731092665,
      "grad_norm": 8.330013275146484,
      "learning_rate": 4.1036895429788395e-06,
      "loss": 0.4072,
      "mean_token_accuracy": 0.8638851836323738,
      "num_tokens": 147415369.0,
      "step": 122600
    },
    {
      "entropy": 1.9261372715234757,
      "epoch": 0.3800802722343162,
      "grad_norm": 8.978737831115723,
      "learning_rate": 4.103522190940217e-06,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8390441820025444,
      "num_tokens": 147426432.0,
      "step": 122610
    },
    {
      "entropy": 1.9050227865576743,
      "epoch": 0.3801112713593659,
      "grad_norm": 8.126078605651855,
      "learning_rate": 4.10335485937421e-06,
      "loss": 0.5085,
      "mean_token_accuracy": 0.844034643471241,
      "num_tokens": 147437847.0,
      "step": 122620
    },
    {
      "entropy": 1.9048843890428544,
      "epoch": 0.38014227048441557,
      "grad_norm": 7.828693389892578,
      "learning_rate": 4.103187548276646e-06,
      "loss": 0.4734,
      "mean_token_accuracy": 0.8543620184063911,
      "num_tokens": 147449041.0,
      "step": 122630
    },
    {
      "entropy": 1.8402218729257585,
      "epoch": 0.3801732696094653,
      "grad_norm": 8.167677879333496,
      "learning_rate": 4.103020257643353e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8610726460814476,
      "num_tokens": 147461307.0,
      "step": 122640
    },
    {
      "entropy": 1.8394718445837497,
      "epoch": 0.38020426873451496,
      "grad_norm": 4.112548828125,
      "learning_rate": 4.1028529874701575e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8556538373231888,
      "num_tokens": 147474356.0,
      "step": 122650
    },
    {
      "entropy": 1.833746100962162,
      "epoch": 0.3802352678595647,
      "grad_norm": 6.661306858062744,
      "learning_rate": 4.102685737752892e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.855821980535984,
      "num_tokens": 147487546.0,
      "step": 122660
    },
    {
      "entropy": 1.8778820380568504,
      "epoch": 0.38026626698461435,
      "grad_norm": 7.601857662200928,
      "learning_rate": 4.102518508487384e-06,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8578482121229172,
      "num_tokens": 147500179.0,
      "step": 122670
    },
    {
      "entropy": 1.8702596746385098,
      "epoch": 0.3802972661096641,
      "grad_norm": 8.159707069396973,
      "learning_rate": 4.10235129966947e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8527193248271943,
      "num_tokens": 147512342.0,
      "step": 122680
    },
    {
      "entropy": 1.9300275981426238,
      "epoch": 0.38032826523471375,
      "grad_norm": 9.76501750946045,
      "learning_rate": 4.10218411129498e-06,
      "loss": 0.5027,
      "mean_token_accuracy": 0.842798325419426,
      "num_tokens": 147523781.0,
      "step": 122690
    },
    {
      "entropy": 1.8886654764413833,
      "epoch": 0.38035926435976347,
      "grad_norm": 8.383800506591797,
      "learning_rate": 4.10201694335975e-06,
      "loss": 0.5191,
      "mean_token_accuracy": 0.8469223186373711,
      "num_tokens": 147536454.0,
      "step": 122700
    },
    {
      "entropy": 1.9181490674614907,
      "epoch": 0.38039026348481314,
      "grad_norm": 7.91489315032959,
      "learning_rate": 4.1018497958596145e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8422677874565124,
      "num_tokens": 147548041.0,
      "step": 122710
    },
    {
      "entropy": 1.7913452178239821,
      "epoch": 0.3804212626098628,
      "grad_norm": 8.690677642822266,
      "learning_rate": 4.1016826687904125e-06,
      "loss": 0.4008,
      "mean_token_accuracy": 0.8631450653076171,
      "num_tokens": 147561955.0,
      "step": 122720
    },
    {
      "entropy": 1.8757226839661598,
      "epoch": 0.38045226173491253,
      "grad_norm": 8.690422058105469,
      "learning_rate": 4.10151556214798e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8484616339206695,
      "num_tokens": 147574345.0,
      "step": 122730
    },
    {
      "entropy": 1.9869503617286681,
      "epoch": 0.3804832608599622,
      "grad_norm": 9.068852424621582,
      "learning_rate": 4.101348475928157e-06,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8462048456072807,
      "num_tokens": 147584992.0,
      "step": 122740
    },
    {
      "entropy": 1.8192487761378289,
      "epoch": 0.3805142599850119,
      "grad_norm": 8.221050262451172,
      "learning_rate": 4.101181410126785e-06,
      "loss": 0.4065,
      "mean_token_accuracy": 0.8553620144724846,
      "num_tokens": 147598434.0,
      "step": 122750
    },
    {
      "entropy": 1.9016779504716397,
      "epoch": 0.3805452591100616,
      "grad_norm": 8.561354637145996,
      "learning_rate": 4.101014364739705e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8509173199534417,
      "num_tokens": 147610587.0,
      "step": 122760
    },
    {
      "entropy": 1.9240503638982773,
      "epoch": 0.3805762582351113,
      "grad_norm": 8.848665237426758,
      "learning_rate": 4.100847339762759e-06,
      "loss": 0.481,
      "mean_token_accuracy": 0.8474395364522934,
      "num_tokens": 147622419.0,
      "step": 122770
    },
    {
      "entropy": 1.9462111860513687,
      "epoch": 0.380607257360161,
      "grad_norm": 8.207417488098145,
      "learning_rate": 4.100680335191792e-06,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8494761645793915,
      "num_tokens": 147632828.0,
      "step": 122780
    },
    {
      "entropy": 1.8405832841992378,
      "epoch": 0.3806382564852107,
      "grad_norm": 4.011960506439209,
      "learning_rate": 4.1005133510226495e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8583981603384018,
      "num_tokens": 147645476.0,
      "step": 122790
    },
    {
      "entropy": 1.9420069836080074,
      "epoch": 0.3806692556102604,
      "grad_norm": 6.821649074554443,
      "learning_rate": 4.100346387251177e-06,
      "loss": 0.5369,
      "mean_token_accuracy": 0.8337976217269898,
      "num_tokens": 147657157.0,
      "step": 122800
    },
    {
      "entropy": 1.9372777387499809,
      "epoch": 0.3807002547353101,
      "grad_norm": 8.612418174743652,
      "learning_rate": 4.100179443873223e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8544177770614624,
      "num_tokens": 147668847.0,
      "step": 122810
    },
    {
      "entropy": 1.840282154083252,
      "epoch": 0.3807312538603598,
      "grad_norm": 4.8995041847229,
      "learning_rate": 4.100012520884635e-06,
      "loss": 0.5072,
      "mean_token_accuracy": 0.8471588909626007,
      "num_tokens": 147682236.0,
      "step": 122820
    },
    {
      "entropy": 1.9169215068221093,
      "epoch": 0.3807622529854095,
      "grad_norm": 8.391762733459473,
      "learning_rate": 4.0998456182812634e-06,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8483142048120499,
      "num_tokens": 147693794.0,
      "step": 122830
    },
    {
      "entropy": 1.8828682228922844,
      "epoch": 0.38079325211045917,
      "grad_norm": 3.6466331481933594,
      "learning_rate": 4.099678736058961e-06,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8551005318760871,
      "num_tokens": 147706128.0,
      "step": 122840
    },
    {
      "entropy": 1.894105489552021,
      "epoch": 0.3808242512355089,
      "grad_norm": 9.506867408752441,
      "learning_rate": 4.0995118742135785e-06,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8543181657791138,
      "num_tokens": 147717697.0,
      "step": 122850
    },
    {
      "entropy": 1.8872636377811431,
      "epoch": 0.38085525036055856,
      "grad_norm": 9.389772415161133,
      "learning_rate": 4.09934503274097e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8546540632843971,
      "num_tokens": 147730381.0,
      "step": 122860
    },
    {
      "entropy": 1.8735971391201018,
      "epoch": 0.3808862494856083,
      "grad_norm": 10.220447540283203,
      "learning_rate": 4.09917821163699e-06,
      "loss": 0.448,
      "mean_token_accuracy": 0.8552726373076439,
      "num_tokens": 147742566.0,
      "step": 122870
    },
    {
      "entropy": 1.8962013125419617,
      "epoch": 0.38091724861065795,
      "grad_norm": 7.584221363067627,
      "learning_rate": 4.099011410897494e-06,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8471159264445305,
      "num_tokens": 147755055.0,
      "step": 122880
    },
    {
      "entropy": 1.9144986018538475,
      "epoch": 0.3809482477357077,
      "grad_norm": 8.150506019592285,
      "learning_rate": 4.098844630518339e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8621457502245903,
      "num_tokens": 147766218.0,
      "step": 122890
    },
    {
      "entropy": 1.8816324099898338,
      "epoch": 0.38097924686075735,
      "grad_norm": 8.89142894744873,
      "learning_rate": 4.0986778704953845e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8532285436987876,
      "num_tokens": 147778543.0,
      "step": 122900
    },
    {
      "entropy": 1.9272763848304748,
      "epoch": 0.3810102459858071,
      "grad_norm": 8.55495548248291,
      "learning_rate": 4.098511130824489e-06,
      "loss": 0.4988,
      "mean_token_accuracy": 0.845673693716526,
      "num_tokens": 147789726.0,
      "step": 122910
    },
    {
      "entropy": 1.7462919235229493,
      "epoch": 0.38104124511085674,
      "grad_norm": 9.728035926818848,
      "learning_rate": 4.0983444115015134e-06,
      "loss": 0.3724,
      "mean_token_accuracy": 0.8587867051362992,
      "num_tokens": 147804271.0,
      "step": 122920
    },
    {
      "entropy": 1.8114602223038674,
      "epoch": 0.38107224423590647,
      "grad_norm": 5.929993152618408,
      "learning_rate": 4.098177712522317e-06,
      "loss": 0.3789,
      "mean_token_accuracy": 0.860235296189785,
      "num_tokens": 147817072.0,
      "step": 122930
    },
    {
      "entropy": 1.846523255109787,
      "epoch": 0.38110324336095613,
      "grad_norm": 8.399463653564453,
      "learning_rate": 4.098011033882767e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8544706001877784,
      "num_tokens": 147830099.0,
      "step": 122940
    },
    {
      "entropy": 1.9178942322731019,
      "epoch": 0.38113424248600586,
      "grad_norm": 4.426633358001709,
      "learning_rate": 4.0978443755787265e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8501872152090073,
      "num_tokens": 147841437.0,
      "step": 122950
    },
    {
      "entropy": 1.881843727827072,
      "epoch": 0.3811652416110555,
      "grad_norm": 4.205018043518066,
      "learning_rate": 4.097677737606057e-06,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8623296022415161,
      "num_tokens": 147852742.0,
      "step": 122960
    },
    {
      "entropy": 1.8286479495465755,
      "epoch": 0.3811962407361052,
      "grad_norm": 10.341126441955566,
      "learning_rate": 4.09751111996063e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8560968965291977,
      "num_tokens": 147865179.0,
      "step": 122970
    },
    {
      "entropy": 1.8264545932412148,
      "epoch": 0.3812272398611549,
      "grad_norm": 8.994978904724121,
      "learning_rate": 4.09734452263831e-06,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8656197428703308,
      "num_tokens": 147877567.0,
      "step": 122980
    },
    {
      "entropy": 1.9247392192482948,
      "epoch": 0.3812582389862046,
      "grad_norm": 7.663559913635254,
      "learning_rate": 4.097177945634967e-06,
      "loss": 0.4954,
      "mean_token_accuracy": 0.8439255699515342,
      "num_tokens": 147889528.0,
      "step": 122990
    },
    {
      "entropy": 1.8423993363976479,
      "epoch": 0.3812892381112543,
      "grad_norm": 3.762521266937256,
      "learning_rate": 4.0970113889464705e-06,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8593509882688523,
      "num_tokens": 147902458.0,
      "step": 123000
    },
    {
      "entropy": 1.9567664802074431,
      "epoch": 0.381320237236304,
      "grad_norm": 8.625659942626953,
      "learning_rate": 4.096844852568692e-06,
      "loss": 0.5569,
      "mean_token_accuracy": 0.8298911958932876,
      "num_tokens": 147913729.0,
      "step": 123010
    },
    {
      "entropy": 1.8710352510213852,
      "epoch": 0.3813512363613537,
      "grad_norm": 8.000286102294922,
      "learning_rate": 4.096678336497503e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8588433533906936,
      "num_tokens": 147925548.0,
      "step": 123020
    },
    {
      "entropy": 1.9936707019805908,
      "epoch": 0.3813822354864034,
      "grad_norm": 9.606999397277832,
      "learning_rate": 4.09651184072878e-06,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8335342779755592,
      "num_tokens": 147936901.0,
      "step": 123030
    },
    {
      "entropy": 1.9054792627692223,
      "epoch": 0.3814132346114531,
      "grad_norm": 9.011368751525879,
      "learning_rate": 4.096345365258394e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8416555240750313,
      "num_tokens": 147948698.0,
      "step": 123040
    },
    {
      "entropy": 1.9143636167049407,
      "epoch": 0.38144423373650277,
      "grad_norm": 9.330082893371582,
      "learning_rate": 4.096178910082223e-06,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8574577882885933,
      "num_tokens": 147960109.0,
      "step": 123050
    },
    {
      "entropy": 1.8986399203538895,
      "epoch": 0.3814752328615525,
      "grad_norm": 8.78416919708252,
      "learning_rate": 4.096012475196143e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8498771920800209,
      "num_tokens": 147972193.0,
      "step": 123060
    },
    {
      "entropy": 1.8185602709650994,
      "epoch": 0.38150623198660216,
      "grad_norm": 8.202901840209961,
      "learning_rate": 4.095846060596033e-06,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8550837740302086,
      "num_tokens": 147985079.0,
      "step": 123070
    },
    {
      "entropy": 1.9409471869468689,
      "epoch": 0.3815372311116519,
      "grad_norm": 7.777408123016357,
      "learning_rate": 4.095679666277773e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8446749314665795,
      "num_tokens": 147996703.0,
      "step": 123080
    },
    {
      "entropy": 1.9127967476844787,
      "epoch": 0.38156823023670156,
      "grad_norm": 8.144055366516113,
      "learning_rate": 4.095513292237241e-06,
      "loss": 0.5223,
      "mean_token_accuracy": 0.8405533611774445,
      "num_tokens": 148007728.0,
      "step": 123090
    },
    {
      "entropy": 1.8132119834423066,
      "epoch": 0.3815992293617513,
      "grad_norm": 4.817179203033447,
      "learning_rate": 4.095346938470322e-06,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8394502356648446,
      "num_tokens": 148021176.0,
      "step": 123100
    },
    {
      "entropy": 1.8567951753735543,
      "epoch": 0.38163022848680095,
      "grad_norm": 8.94315242767334,
      "learning_rate": 4.095180604972897e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8610012695193291,
      "num_tokens": 148033775.0,
      "step": 123110
    },
    {
      "entropy": 1.7960670098662377,
      "epoch": 0.3816612276118507,
      "grad_norm": 8.07936954498291,
      "learning_rate": 4.095014291740849e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8699286490678787,
      "num_tokens": 148046444.0,
      "step": 123120
    },
    {
      "entropy": 1.880557769536972,
      "epoch": 0.38169222673690034,
      "grad_norm": 8.235962867736816,
      "learning_rate": 4.094847998770066e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.859109228849411,
      "num_tokens": 148058765.0,
      "step": 123130
    },
    {
      "entropy": 1.8710905313491821,
      "epoch": 0.38172322586195007,
      "grad_norm": 8.953729629516602,
      "learning_rate": 4.094681726056433e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8571123152971267,
      "num_tokens": 148071128.0,
      "step": 123140
    },
    {
      "entropy": 1.8630672127008439,
      "epoch": 0.38175422498699974,
      "grad_norm": 8.592804908752441,
      "learning_rate": 4.094515473595838e-06,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8458248943090438,
      "num_tokens": 148082981.0,
      "step": 123150
    },
    {
      "entropy": 1.900773896276951,
      "epoch": 0.38178522411204946,
      "grad_norm": 7.337060928344727,
      "learning_rate": 4.0943492413841685e-06,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8392305105924607,
      "num_tokens": 148095109.0,
      "step": 123160
    },
    {
      "entropy": 1.827962300926447,
      "epoch": 0.38181622323709913,
      "grad_norm": 7.826770305633545,
      "learning_rate": 4.094183029417316e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.8655768454074859,
      "num_tokens": 148108413.0,
      "step": 123170
    },
    {
      "entropy": 1.8747796788811684,
      "epoch": 0.38184722236214885,
      "grad_norm": 8.268677711486816,
      "learning_rate": 4.0940168376911705e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8500054642558098,
      "num_tokens": 148120152.0,
      "step": 123180
    },
    {
      "entropy": 1.8802484720945358,
      "epoch": 0.3818782214871985,
      "grad_norm": 9.56521987915039,
      "learning_rate": 4.093850666201625e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8550689592957497,
      "num_tokens": 148131543.0,
      "step": 123190
    },
    {
      "entropy": 1.9417632609605788,
      "epoch": 0.38190922061224825,
      "grad_norm": 8.511123657226562,
      "learning_rate": 4.093684514944573e-06,
      "loss": 0.5105,
      "mean_token_accuracy": 0.8440429598093033,
      "num_tokens": 148142246.0,
      "step": 123200
    },
    {
      "entropy": 1.862280185520649,
      "epoch": 0.3819402197372979,
      "grad_norm": 9.24006175994873,
      "learning_rate": 4.093518383915908e-06,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8556727141141891,
      "num_tokens": 148154604.0,
      "step": 123210
    },
    {
      "entropy": 1.90358504652977,
      "epoch": 0.3819712188623476,
      "grad_norm": 7.789550304412842,
      "learning_rate": 4.093352273111527e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8538018524646759,
      "num_tokens": 148166038.0,
      "step": 123220
    },
    {
      "entropy": 1.8469286099076272,
      "epoch": 0.3820022179873973,
      "grad_norm": 8.740273475646973,
      "learning_rate": 4.093186182527327e-06,
      "loss": 0.4235,
      "mean_token_accuracy": 0.858376681804657,
      "num_tokens": 148178188.0,
      "step": 123230
    },
    {
      "entropy": 1.8784661173820496,
      "epoch": 0.382033217112447,
      "grad_norm": 8.926239013671875,
      "learning_rate": 4.093020112159205e-06,
      "loss": 0.616,
      "mean_token_accuracy": 0.8364622831344605,
      "num_tokens": 148190734.0,
      "step": 123240
    },
    {
      "entropy": 1.9453864216804504,
      "epoch": 0.3820642162374967,
      "grad_norm": 7.907076358795166,
      "learning_rate": 4.092854062003061e-06,
      "loss": 0.47,
      "mean_token_accuracy": 0.8546721920371055,
      "num_tokens": 148201633.0,
      "step": 123250
    },
    {
      "entropy": 1.8376804277300836,
      "epoch": 0.38209521536254637,
      "grad_norm": 9.020318031311035,
      "learning_rate": 4.092688032054795e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8542071163654328,
      "num_tokens": 148214535.0,
      "step": 123260
    },
    {
      "entropy": 1.8767757460474968,
      "epoch": 0.3821262144875961,
      "grad_norm": 8.229818344116211,
      "learning_rate": 4.092522022310309e-06,
      "loss": 0.5102,
      "mean_token_accuracy": 0.8497373566031456,
      "num_tokens": 148226463.0,
      "step": 123270
    },
    {
      "entropy": 1.8976902469992638,
      "epoch": 0.38215721361264576,
      "grad_norm": 8.769594192504883,
      "learning_rate": 4.092356032765506e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8375331163406372,
      "num_tokens": 148238620.0,
      "step": 123280
    },
    {
      "entropy": 1.9701337844133378,
      "epoch": 0.3821882127376955,
      "grad_norm": 9.727083206176758,
      "learning_rate": 4.092190063416288e-06,
      "loss": 0.5527,
      "mean_token_accuracy": 0.8314969301223755,
      "num_tokens": 148249082.0,
      "step": 123290
    },
    {
      "entropy": 1.8411318197846414,
      "epoch": 0.38221921186274516,
      "grad_norm": 7.282608985900879,
      "learning_rate": 4.0920241142585636e-06,
      "loss": 0.39,
      "mean_token_accuracy": 0.8594763651490211,
      "num_tokens": 148261567.0,
      "step": 123300
    },
    {
      "entropy": 1.9043711677193642,
      "epoch": 0.3822502109877949,
      "grad_norm": 7.864800453186035,
      "learning_rate": 4.091858185288235e-06,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8414738133549691,
      "num_tokens": 148273132.0,
      "step": 123310
    },
    {
      "entropy": 1.809215198457241,
      "epoch": 0.38228121011284455,
      "grad_norm": 8.225022315979004,
      "learning_rate": 4.091692276501213e-06,
      "loss": 0.3809,
      "mean_token_accuracy": 0.8564652130007744,
      "num_tokens": 148286238.0,
      "step": 123320
    },
    {
      "entropy": 1.7878570154309272,
      "epoch": 0.3823122092378943,
      "grad_norm": 4.167350769042969,
      "learning_rate": 4.0915263878934044e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8619178295135498,
      "num_tokens": 148298922.0,
      "step": 123330
    },
    {
      "entropy": 1.850788153707981,
      "epoch": 0.38234320836294394,
      "grad_norm": 7.613214492797852,
      "learning_rate": 4.091360519460719e-06,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8597221150994301,
      "num_tokens": 148311479.0,
      "step": 123340
    },
    {
      "entropy": 1.8307570219039917,
      "epoch": 0.38237420748799367,
      "grad_norm": 4.395393371582031,
      "learning_rate": 4.0911946711990686e-06,
      "loss": 0.4586,
      "mean_token_accuracy": 0.8481859222054482,
      "num_tokens": 148324806.0,
      "step": 123350
    },
    {
      "entropy": 1.9396772772073745,
      "epoch": 0.38240520661304334,
      "grad_norm": 3.719348430633545,
      "learning_rate": 4.091028843104365e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8467726603150367,
      "num_tokens": 148336424.0,
      "step": 123360
    },
    {
      "entropy": 1.8743967324495316,
      "epoch": 0.38243620573809306,
      "grad_norm": 3.7305939197540283,
      "learning_rate": 4.090863035172519e-06,
      "loss": 0.443,
      "mean_token_accuracy": 0.8555470928549767,
      "num_tokens": 148348774.0,
      "step": 123370
    },
    {
      "entropy": 1.9252464964985847,
      "epoch": 0.38246720486314273,
      "grad_norm": 8.94243335723877,
      "learning_rate": 4.090697247399448e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8391307726502418,
      "num_tokens": 148360798.0,
      "step": 123380
    },
    {
      "entropy": 1.9863820880651475,
      "epoch": 0.38249820398819245,
      "grad_norm": 9.898303031921387,
      "learning_rate": 4.090531479781067e-06,
      "loss": 0.5399,
      "mean_token_accuracy": 0.8426146388053894,
      "num_tokens": 148371596.0,
      "step": 123390
    },
    {
      "entropy": 1.8762628570199014,
      "epoch": 0.3825292031132421,
      "grad_norm": 7.574060440063477,
      "learning_rate": 4.0903657323132925e-06,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8638320744037629,
      "num_tokens": 148383122.0,
      "step": 123400
    },
    {
      "entropy": 1.9100775212049483,
      "epoch": 0.38256020223829185,
      "grad_norm": 4.106917858123779,
      "learning_rate": 4.0902000049920414e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8347994074225425,
      "num_tokens": 148394952.0,
      "step": 123410
    },
    {
      "entropy": 1.9066940754652024,
      "epoch": 0.3825912013633415,
      "grad_norm": 7.716053009033203,
      "learning_rate": 4.090034297813234e-06,
      "loss": 0.4522,
      "mean_token_accuracy": 0.849234452843666,
      "num_tokens": 148407262.0,
      "step": 123420
    },
    {
      "entropy": 1.7800323203206063,
      "epoch": 0.38262220048839124,
      "grad_norm": 4.275322914123535,
      "learning_rate": 4.089868610772788e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8655282810330391,
      "num_tokens": 148421221.0,
      "step": 123430
    },
    {
      "entropy": 1.823851725459099,
      "epoch": 0.3826531996134409,
      "grad_norm": 8.498555183410645,
      "learning_rate": 4.089702943866628e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.8620473012328148,
      "num_tokens": 148434376.0,
      "step": 123440
    },
    {
      "entropy": 1.9421150475740432,
      "epoch": 0.38268419873849063,
      "grad_norm": 8.584688186645508,
      "learning_rate": 4.0895372970906745e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.8448976293206215,
      "num_tokens": 148445096.0,
      "step": 123450
    },
    {
      "entropy": 1.8550694674253463,
      "epoch": 0.3827151978635403,
      "grad_norm": 8.82845401763916,
      "learning_rate": 4.089371670440852e-06,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8511300772428513,
      "num_tokens": 148456927.0,
      "step": 123460
    },
    {
      "entropy": 1.92039655148983,
      "epoch": 0.38274619698858997,
      "grad_norm": 8.373238563537598,
      "learning_rate": 4.089206063913085e-06,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8491540059447289,
      "num_tokens": 148468499.0,
      "step": 123470
    },
    {
      "entropy": 1.8469054311513902,
      "epoch": 0.3827771961136397,
      "grad_norm": 8.064623832702637,
      "learning_rate": 4.089040477503299e-06,
      "loss": 0.4028,
      "mean_token_accuracy": 0.8639935791492462,
      "num_tokens": 148480307.0,
      "step": 123480
    },
    {
      "entropy": 1.961165651679039,
      "epoch": 0.38280819523868936,
      "grad_norm": 7.094374656677246,
      "learning_rate": 4.088874911207421e-06,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8461368843913079,
      "num_tokens": 148490982.0,
      "step": 123490
    },
    {
      "entropy": 1.8972523733973503,
      "epoch": 0.3828391943637391,
      "grad_norm": 9.41102123260498,
      "learning_rate": 4.08870936502138e-06,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8459867879748344,
      "num_tokens": 148503085.0,
      "step": 123500
    },
    {
      "entropy": 1.8196463674306869,
      "epoch": 0.38287019348878876,
      "grad_norm": 8.395617485046387,
      "learning_rate": 4.088543838941105e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8633744567632675,
      "num_tokens": 148515132.0,
      "step": 123510
    },
    {
      "entropy": 1.933710703253746,
      "epoch": 0.3829011926138385,
      "grad_norm": 9.440218925476074,
      "learning_rate": 4.088378332962527e-06,
      "loss": 0.5686,
      "mean_token_accuracy": 0.8320772811770439,
      "num_tokens": 148526393.0,
      "step": 123520
    },
    {
      "entropy": 1.8306830495595932,
      "epoch": 0.38293219173888815,
      "grad_norm": 3.874952554702759,
      "learning_rate": 4.088212847081577e-06,
      "loss": 0.4136,
      "mean_token_accuracy": 0.8493449553847313,
      "num_tokens": 148539616.0,
      "step": 123530
    },
    {
      "entropy": 1.8871184498071671,
      "epoch": 0.3829631908639379,
      "grad_norm": 4.710858345031738,
      "learning_rate": 4.0880473812941876e-06,
      "loss": 0.4812,
      "mean_token_accuracy": 0.848838959634304,
      "num_tokens": 148551814.0,
      "step": 123540
    },
    {
      "entropy": 1.8882265403866767,
      "epoch": 0.38299418998898754,
      "grad_norm": 7.315601825714111,
      "learning_rate": 4.087881935596294e-06,
      "loss": 0.4169,
      "mean_token_accuracy": 0.8602377876639367,
      "num_tokens": 148563466.0,
      "step": 123550
    },
    {
      "entropy": 1.9142208576202393,
      "epoch": 0.38302518911403727,
      "grad_norm": 8.4950590133667,
      "learning_rate": 4.08771650998383e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8584097474813461,
      "num_tokens": 148575212.0,
      "step": 123560
    },
    {
      "entropy": 1.9166872769594192,
      "epoch": 0.38305618823908694,
      "grad_norm": 8.157064437866211,
      "learning_rate": 4.087551104452733e-06,
      "loss": 0.5462,
      "mean_token_accuracy": 0.8363913476467133,
      "num_tokens": 148587345.0,
      "step": 123570
    },
    {
      "entropy": 1.852193793654442,
      "epoch": 0.38308718736413666,
      "grad_norm": 5.7400031089782715,
      "learning_rate": 4.08738571899894e-06,
      "loss": 0.4275,
      "mean_token_accuracy": 0.8614915952086448,
      "num_tokens": 148599766.0,
      "step": 123580
    },
    {
      "entropy": 1.6889735147356988,
      "epoch": 0.38311818648918633,
      "grad_norm": 9.48442554473877,
      "learning_rate": 4.08722035361839e-06,
      "loss": 0.3805,
      "mean_token_accuracy": 0.8619127184152603,
      "num_tokens": 148613775.0,
      "step": 123590
    },
    {
      "entropy": 1.8661672458052636,
      "epoch": 0.38314918561423605,
      "grad_norm": 3.9152071475982666,
      "learning_rate": 4.087055008307023e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8527839332818985,
      "num_tokens": 148625883.0,
      "step": 123600
    },
    {
      "entropy": 1.9174211084842683,
      "epoch": 0.3831801847392857,
      "grad_norm": 4.289289951324463,
      "learning_rate": 4.086889683060778e-06,
      "loss": 0.466,
      "mean_token_accuracy": 0.8489474713802337,
      "num_tokens": 148637944.0,
      "step": 123610
    },
    {
      "entropy": 1.9397675469517708,
      "epoch": 0.38321118386433545,
      "grad_norm": 8.902761459350586,
      "learning_rate": 4.086724377875599e-06,
      "loss": 0.5397,
      "mean_token_accuracy": 0.8369091704487801,
      "num_tokens": 148649564.0,
      "step": 123620
    },
    {
      "entropy": 1.8931380987167359,
      "epoch": 0.3832421829893851,
      "grad_norm": 7.973553657531738,
      "learning_rate": 4.08655909274743e-06,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8456487700343132,
      "num_tokens": 148661427.0,
      "step": 123630
    },
    {
      "entropy": 1.779283882677555,
      "epoch": 0.38327318211443484,
      "grad_norm": 9.01836109161377,
      "learning_rate": 4.086393827672212e-06,
      "loss": 0.449,
      "mean_token_accuracy": 0.8560737937688827,
      "num_tokens": 148675109.0,
      "step": 123640
    },
    {
      "entropy": 1.899433010816574,
      "epoch": 0.3833041812394845,
      "grad_norm": 9.137697219848633,
      "learning_rate": 4.086228582645893e-06,
      "loss": 0.445,
      "mean_token_accuracy": 0.8533209547400474,
      "num_tokens": 148686499.0,
      "step": 123650
    },
    {
      "entropy": 1.9183125644922256,
      "epoch": 0.38333518036453423,
      "grad_norm": 8.239258766174316,
      "learning_rate": 4.08606335766442e-06,
      "loss": 0.4644,
      "mean_token_accuracy": 0.8569030150771141,
      "num_tokens": 148697749.0,
      "step": 123660
    },
    {
      "entropy": 1.7693776428699493,
      "epoch": 0.3833661794895839,
      "grad_norm": 3.299410343170166,
      "learning_rate": 4.08589815272374e-06,
      "loss": 0.3586,
      "mean_token_accuracy": 0.8623643219470978,
      "num_tokens": 148712318.0,
      "step": 123670
    },
    {
      "entropy": 1.9084981709718705,
      "epoch": 0.3833971786146336,
      "grad_norm": 7.353041172027588,
      "learning_rate": 4.085732967819801e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.847068789601326,
      "num_tokens": 148724494.0,
      "step": 123680
    },
    {
      "entropy": 1.9519029945135116,
      "epoch": 0.3834281777396833,
      "grad_norm": 8.309041976928711,
      "learning_rate": 4.085567802948554e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8656853973865509,
      "num_tokens": 148735242.0,
      "step": 123690
    },
    {
      "entropy": 1.8672381609678268,
      "epoch": 0.383459176864733,
      "grad_norm": 8.3343505859375,
      "learning_rate": 4.085402658105951e-06,
      "loss": 0.4244,
      "mean_token_accuracy": 0.8520313948392868,
      "num_tokens": 148747370.0,
      "step": 123700
    },
    {
      "entropy": 1.9503202199935914,
      "epoch": 0.3834901759897827,
      "grad_norm": 7.8455729484558105,
      "learning_rate": 4.085237533287944e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8526298850774765,
      "num_tokens": 148757963.0,
      "step": 123710
    },
    {
      "entropy": 1.861949661374092,
      "epoch": 0.38352117511483236,
      "grad_norm": 3.735707998275757,
      "learning_rate": 4.085072428490485e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8578638583421707,
      "num_tokens": 148770099.0,
      "step": 123720
    },
    {
      "entropy": 1.8470420002937318,
      "epoch": 0.3835521742398821,
      "grad_norm": 10.177236557006836,
      "learning_rate": 4.0849073437095295e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8429055437445641,
      "num_tokens": 148782367.0,
      "step": 123730
    },
    {
      "entropy": 1.9221158519387245,
      "epoch": 0.38358317336493175,
      "grad_norm": 4.895500183105469,
      "learning_rate": 4.0847422789410344e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8465067788958549,
      "num_tokens": 148794003.0,
      "step": 123740
    },
    {
      "entropy": 1.970751866698265,
      "epoch": 0.3836141724899815,
      "grad_norm": 7.53320837020874,
      "learning_rate": 4.084577234180957e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8521478682756424,
      "num_tokens": 148804935.0,
      "step": 123750
    },
    {
      "entropy": 1.7943205565214158,
      "epoch": 0.38364517161503114,
      "grad_norm": 6.433713436126709,
      "learning_rate": 4.084412209425253e-06,
      "loss": 0.3881,
      "mean_token_accuracy": 0.8569139391183853,
      "num_tokens": 148818096.0,
      "step": 123760
    },
    {
      "entropy": 1.7551213905215264,
      "epoch": 0.38367617074008087,
      "grad_norm": 7.526129245758057,
      "learning_rate": 4.084247204669883e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.8625258266925812,
      "num_tokens": 148831912.0,
      "step": 123770
    },
    {
      "entropy": 1.862623292207718,
      "epoch": 0.38370716986513054,
      "grad_norm": 8.402950286865234,
      "learning_rate": 4.084082219910807e-06,
      "loss": 0.3984,
      "mean_token_accuracy": 0.8564006343483925,
      "num_tokens": 148844690.0,
      "step": 123780
    },
    {
      "entropy": 1.90057223290205,
      "epoch": 0.38373816899018026,
      "grad_norm": 3.392810344696045,
      "learning_rate": 4.083917255143988e-06,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8512577295303345,
      "num_tokens": 148856761.0,
      "step": 123790
    },
    {
      "entropy": 1.9783919692039489,
      "epoch": 0.38376916811522993,
      "grad_norm": 9.549758911132812,
      "learning_rate": 4.083752310365388e-06,
      "loss": 0.5152,
      "mean_token_accuracy": 0.8406032085418701,
      "num_tokens": 148867423.0,
      "step": 123800
    },
    {
      "entropy": 1.8232928544282914,
      "epoch": 0.38380016724027965,
      "grad_norm": 4.337226867675781,
      "learning_rate": 4.083587385570969e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.852597390115261,
      "num_tokens": 148879925.0,
      "step": 123810
    },
    {
      "entropy": 1.8824557334184646,
      "epoch": 0.3838311663653293,
      "grad_norm": 7.782412528991699,
      "learning_rate": 4.083422480756698e-06,
      "loss": 0.4413,
      "mean_token_accuracy": 0.856006121635437,
      "num_tokens": 148891322.0,
      "step": 123820
    },
    {
      "entropy": 1.903056225925684,
      "epoch": 0.38386216549037905,
      "grad_norm": 8.749157905578613,
      "learning_rate": 4.083257595918541e-06,
      "loss": 0.454,
      "mean_token_accuracy": 0.845898973941803,
      "num_tokens": 148903673.0,
      "step": 123830
    },
    {
      "entropy": 1.9123461306095124,
      "epoch": 0.3838931646154287,
      "grad_norm": 8.949573516845703,
      "learning_rate": 4.083092731052464e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8477382779121398,
      "num_tokens": 148915982.0,
      "step": 123840
    },
    {
      "entropy": 1.867120423913002,
      "epoch": 0.38392416374047844,
      "grad_norm": 8.769627571105957,
      "learning_rate": 4.082927886154436e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8450190261006355,
      "num_tokens": 148928151.0,
      "step": 123850
    },
    {
      "entropy": 1.9295039504766465,
      "epoch": 0.3839551628655281,
      "grad_norm": 10.065016746520996,
      "learning_rate": 4.0827630612204275e-06,
      "loss": 0.5505,
      "mean_token_accuracy": 0.8388926431536674,
      "num_tokens": 148939030.0,
      "step": 123860
    },
    {
      "entropy": 1.8576544880867005,
      "epoch": 0.38398616199057783,
      "grad_norm": 11.196900367736816,
      "learning_rate": 4.082598256246408e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8566042557358742,
      "num_tokens": 148950841.0,
      "step": 123870
    },
    {
      "entropy": 1.8984556332230569,
      "epoch": 0.3840171611156275,
      "grad_norm": 9.907042503356934,
      "learning_rate": 4.082433471228349e-06,
      "loss": 0.5076,
      "mean_token_accuracy": 0.8443589746952057,
      "num_tokens": 148962371.0,
      "step": 123880
    },
    {
      "entropy": 1.7866959184408189,
      "epoch": 0.3840481602406772,
      "grad_norm": 7.650696754455566,
      "learning_rate": 4.082268706162224e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8526209697127343,
      "num_tokens": 148975931.0,
      "step": 123890
    },
    {
      "entropy": 1.8315860643982886,
      "epoch": 0.3840791593657269,
      "grad_norm": 2.8190009593963623,
      "learning_rate": 4.082103961044008e-06,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8488441839814186,
      "num_tokens": 148989418.0,
      "step": 123900
    },
    {
      "entropy": 1.803328700363636,
      "epoch": 0.3841101584907766,
      "grad_norm": 8.294083595275879,
      "learning_rate": 4.081939235869675e-06,
      "loss": 0.4175,
      "mean_token_accuracy": 0.8606008633971214,
      "num_tokens": 149003217.0,
      "step": 123910
    },
    {
      "entropy": 1.8937855035066604,
      "epoch": 0.3841411576158263,
      "grad_norm": 8.08438777923584,
      "learning_rate": 4.081774530635201e-06,
      "loss": 0.511,
      "mean_token_accuracy": 0.8416438445448875,
      "num_tokens": 149014790.0,
      "step": 123920
    },
    {
      "entropy": 1.9089516907930375,
      "epoch": 0.384172156740876,
      "grad_norm": 8.53679370880127,
      "learning_rate": 4.081609845336565e-06,
      "loss": 0.4647,
      "mean_token_accuracy": 0.848609185218811,
      "num_tokens": 149025831.0,
      "step": 123930
    },
    {
      "entropy": 1.7949181117117405,
      "epoch": 0.3842031558659257,
      "grad_norm": 4.385751247406006,
      "learning_rate": 4.0814451799697436e-06,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8539316862821579,
      "num_tokens": 149038872.0,
      "step": 123940
    },
    {
      "entropy": 1.8139029011130332,
      "epoch": 0.3842341549909754,
      "grad_norm": 3.748425006866455,
      "learning_rate": 4.081280534530718e-06,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8566805422306061,
      "num_tokens": 149051871.0,
      "step": 123950
    },
    {
      "entropy": 1.7684552431106568,
      "epoch": 0.3842651541160251,
      "grad_norm": 10.310443878173828,
      "learning_rate": 4.081115909015469e-06,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8510794118046761,
      "num_tokens": 149065114.0,
      "step": 123960
    },
    {
      "entropy": 1.8360266655683517,
      "epoch": 0.38429615324107474,
      "grad_norm": 9.245149612426758,
      "learning_rate": 4.080951303419979e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.844825841486454,
      "num_tokens": 149077418.0,
      "step": 123970
    },
    {
      "entropy": 1.7973667308688164,
      "epoch": 0.38432715236612447,
      "grad_norm": 8.100232124328613,
      "learning_rate": 4.0807867177402305e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8598975196480751,
      "num_tokens": 149090213.0,
      "step": 123980
    },
    {
      "entropy": 1.8818108469247818,
      "epoch": 0.38435815149117414,
      "grad_norm": 8.730313301086426,
      "learning_rate": 4.080622151972207e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8592151805758477,
      "num_tokens": 149102306.0,
      "step": 123990
    },
    {
      "entropy": 1.9396295577287674,
      "epoch": 0.38438915061622386,
      "grad_norm": 8.592621803283691,
      "learning_rate": 4.080457606111895e-06,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8406242683529854,
      "num_tokens": 149114284.0,
      "step": 124000
    },
    {
      "entropy": 1.8134768947958946,
      "epoch": 0.38442014974127353,
      "grad_norm": 8.235865592956543,
      "learning_rate": 4.080293080155281e-06,
      "loss": 0.422,
      "mean_token_accuracy": 0.8475548833608627,
      "num_tokens": 149127288.0,
      "step": 124010
    },
    {
      "entropy": 1.7447491750121116,
      "epoch": 0.38445114886632326,
      "grad_norm": 8.99509334564209,
      "learning_rate": 4.080128574098353e-06,
      "loss": 0.3478,
      "mean_token_accuracy": 0.8710707902908326,
      "num_tokens": 149140350.0,
      "step": 124020
    },
    {
      "entropy": 1.9281997457146645,
      "epoch": 0.3844821479913729,
      "grad_norm": 9.224446296691895,
      "learning_rate": 4.0799640879370986e-06,
      "loss": 0.5329,
      "mean_token_accuracy": 0.8362716525793076,
      "num_tokens": 149151526.0,
      "step": 124030
    },
    {
      "entropy": 1.907413001358509,
      "epoch": 0.38451314711642265,
      "grad_norm": 8.881566047668457,
      "learning_rate": 4.079799621667508e-06,
      "loss": 0.5119,
      "mean_token_accuracy": 0.841308769583702,
      "num_tokens": 149162744.0,
      "step": 124040
    },
    {
      "entropy": 1.764004696905613,
      "epoch": 0.3845441462414723,
      "grad_norm": 8.439988136291504,
      "learning_rate": 4.079635175285575e-06,
      "loss": 0.4012,
      "mean_token_accuracy": 0.8685743331909179,
      "num_tokens": 149175813.0,
      "step": 124050
    },
    {
      "entropy": 1.8820173025131226,
      "epoch": 0.38457514536652204,
      "grad_norm": 7.626965522766113,
      "learning_rate": 4.079470748787288e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8496981009840965,
      "num_tokens": 149188028.0,
      "step": 124060
    },
    {
      "entropy": 1.9048197478055955,
      "epoch": 0.3846061444915717,
      "grad_norm": 4.152190685272217,
      "learning_rate": 4.079306342168641e-06,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8482609540224075,
      "num_tokens": 149199701.0,
      "step": 124070
    },
    {
      "entropy": 1.9153966382145882,
      "epoch": 0.38463714361662144,
      "grad_norm": 7.222431659698486,
      "learning_rate": 4.079141955425631e-06,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8490887343883514,
      "num_tokens": 149211482.0,
      "step": 124080
    },
    {
      "entropy": 1.7790709897875785,
      "epoch": 0.3846681427416711,
      "grad_norm": 3.140486001968384,
      "learning_rate": 4.0789775885542525e-06,
      "loss": 0.417,
      "mean_token_accuracy": 0.8636253029108047,
      "num_tokens": 149224774.0,
      "step": 124090
    },
    {
      "entropy": 1.9251772895455361,
      "epoch": 0.38469914186672083,
      "grad_norm": 4.184020519256592,
      "learning_rate": 4.0788132415505e-06,
      "loss": 0.4724,
      "mean_token_accuracy": 0.8495764017105103,
      "num_tokens": 149236068.0,
      "step": 124100
    },
    {
      "entropy": 1.8736867174506187,
      "epoch": 0.3847301409917705,
      "grad_norm": 9.371009826660156,
      "learning_rate": 4.078648914410375e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8543329164385796,
      "num_tokens": 149248502.0,
      "step": 124110
    },
    {
      "entropy": 1.8708575084805488,
      "epoch": 0.3847611401168202,
      "grad_norm": 3.769465684890747,
      "learning_rate": 4.078484607129874e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8484385311603546,
      "num_tokens": 149260240.0,
      "step": 124120
    },
    {
      "entropy": 1.9300208121538163,
      "epoch": 0.3847921392418699,
      "grad_norm": 7.285602569580078,
      "learning_rate": 4.078320319704997e-06,
      "loss": 0.492,
      "mean_token_accuracy": 0.8484133720397949,
      "num_tokens": 149271397.0,
      "step": 124130
    },
    {
      "entropy": 1.8357270821928977,
      "epoch": 0.3848231383669196,
      "grad_norm": 3.4529945850372314,
      "learning_rate": 4.078156052131747e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8534119576215744,
      "num_tokens": 149283833.0,
      "step": 124140
    },
    {
      "entropy": 1.9487372577190398,
      "epoch": 0.3848541374919693,
      "grad_norm": 7.477880001068115,
      "learning_rate": 4.0779918044061244e-06,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8423938781023026,
      "num_tokens": 149294571.0,
      "step": 124150
    },
    {
      "entropy": 1.8556806325912476,
      "epoch": 0.384885136617019,
      "grad_norm": 4.033416748046875,
      "learning_rate": 4.077827576524136e-06,
      "loss": 0.4794,
      "mean_token_accuracy": 0.84943907558918,
      "num_tokens": 149307020.0,
      "step": 124160
    },
    {
      "entropy": 1.877730706334114,
      "epoch": 0.3849161357420687,
      "grad_norm": 8.740376472473145,
      "learning_rate": 4.077663368481781e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.8621541142463685,
      "num_tokens": 149318207.0,
      "step": 124170
    },
    {
      "entropy": 1.9129869103431703,
      "epoch": 0.3849471348671184,
      "grad_norm": 8.78154468536377,
      "learning_rate": 4.07749918027507e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8474374935030937,
      "num_tokens": 149329638.0,
      "step": 124180
    },
    {
      "entropy": 1.8814658731222154,
      "epoch": 0.38497813399216807,
      "grad_norm": 8.555291175842285,
      "learning_rate": 4.077335011900008e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8533140555024147,
      "num_tokens": 149341341.0,
      "step": 124190
    },
    {
      "entropy": 1.8951398521661758,
      "epoch": 0.38500913311721774,
      "grad_norm": 7.886125564575195,
      "learning_rate": 4.077170863352603e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8445818021893501,
      "num_tokens": 149353637.0,
      "step": 124200
    },
    {
      "entropy": 1.8607033982872963,
      "epoch": 0.38504013224226746,
      "grad_norm": 6.102626323699951,
      "learning_rate": 4.0770067346288645e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8472836226224899,
      "num_tokens": 149365535.0,
      "step": 124210
    },
    {
      "entropy": 1.9797284454107285,
      "epoch": 0.38507113136731713,
      "grad_norm": 8.020270347595215,
      "learning_rate": 4.076842625724803e-06,
      "loss": 0.5767,
      "mean_token_accuracy": 0.8357729539275169,
      "num_tokens": 149376980.0,
      "step": 124220
    },
    {
      "entropy": 1.8790086820721625,
      "epoch": 0.38510213049236686,
      "grad_norm": 8.278203010559082,
      "learning_rate": 4.076678536636429e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8577861353754997,
      "num_tokens": 149388865.0,
      "step": 124230
    },
    {
      "entropy": 1.8905736938118936,
      "epoch": 0.3851331296174165,
      "grad_norm": 9.991707801818848,
      "learning_rate": 4.076514467359756e-06,
      "loss": 0.4826,
      "mean_token_accuracy": 0.857266665995121,
      "num_tokens": 149400675.0,
      "step": 124240
    },
    {
      "entropy": 1.9448306143283844,
      "epoch": 0.38516412874246625,
      "grad_norm": 8.666476249694824,
      "learning_rate": 4.076350417890796e-06,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8469108313322067,
      "num_tokens": 149411540.0,
      "step": 124250
    },
    {
      "entropy": 1.8579439014196395,
      "epoch": 0.3851951278675159,
      "grad_norm": 4.379420280456543,
      "learning_rate": 4.076186388225566e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.8480482071638107,
      "num_tokens": 149424172.0,
      "step": 124260
    },
    {
      "entropy": 1.7416525185108185,
      "epoch": 0.38522612699256564,
      "grad_norm": 4.337442874908447,
      "learning_rate": 4.076022378360081e-06,
      "loss": 0.3671,
      "mean_token_accuracy": 0.87484562844038,
      "num_tokens": 149438737.0,
      "step": 124270
    },
    {
      "entropy": 1.8429612591862679,
      "epoch": 0.3852571261176153,
      "grad_norm": 3.8294622898101807,
      "learning_rate": 4.075858388290356e-06,
      "loss": 0.4457,
      "mean_token_accuracy": 0.855830529332161,
      "num_tokens": 149451375.0,
      "step": 124280
    },
    {
      "entropy": 1.8703719988465308,
      "epoch": 0.38528812524266504,
      "grad_norm": 8.686786651611328,
      "learning_rate": 4.075694418012412e-06,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8558905765414238,
      "num_tokens": 149463577.0,
      "step": 124290
    },
    {
      "entropy": 1.9531716659665108,
      "epoch": 0.3853191243677147,
      "grad_norm": 10.40597915649414,
      "learning_rate": 4.075530467522267e-06,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8417243003845215,
      "num_tokens": 149474986.0,
      "step": 124300
    },
    {
      "entropy": 1.9056412920355796,
      "epoch": 0.38535012349276443,
      "grad_norm": 9.468035697937012,
      "learning_rate": 4.0753665368159415e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8571117267012596,
      "num_tokens": 149486586.0,
      "step": 124310
    },
    {
      "entropy": 1.8688831850886345,
      "epoch": 0.3853811226178141,
      "grad_norm": 7.645292282104492,
      "learning_rate": 4.0752026258894575e-06,
      "loss": 0.4171,
      "mean_token_accuracy": 0.8533623024821282,
      "num_tokens": 149499014.0,
      "step": 124320
    },
    {
      "entropy": 1.8921454787254333,
      "epoch": 0.3854121217428638,
      "grad_norm": 3.7551565170288086,
      "learning_rate": 4.0750387347388356e-06,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8515683576464653,
      "num_tokens": 149511579.0,
      "step": 124330
    },
    {
      "entropy": 1.933222909271717,
      "epoch": 0.3854431208679135,
      "grad_norm": 9.027485847473145,
      "learning_rate": 4.074874863360102e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8418224409222603,
      "num_tokens": 149522694.0,
      "step": 124340
    },
    {
      "entropy": 1.88301939368248,
      "epoch": 0.3854741199929632,
      "grad_norm": 7.786672115325928,
      "learning_rate": 4.074711011749281e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8608944162726402,
      "num_tokens": 149534481.0,
      "step": 124350
    },
    {
      "entropy": 1.870416359603405,
      "epoch": 0.3855051191180129,
      "grad_norm": 7.477562427520752,
      "learning_rate": 4.0745471799023966e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8581738561391831,
      "num_tokens": 149546190.0,
      "step": 124360
    },
    {
      "entropy": 1.8137023776769639,
      "epoch": 0.3855361182430626,
      "grad_norm": 8.037935256958008,
      "learning_rate": 4.074383367815478e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.855655774474144,
      "num_tokens": 149559214.0,
      "step": 124370
    },
    {
      "entropy": 1.890501284599304,
      "epoch": 0.3855671173681123,
      "grad_norm": 4.326991558074951,
      "learning_rate": 4.074219575484553e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8478201061487198,
      "num_tokens": 149570897.0,
      "step": 124380
    },
    {
      "entropy": 1.8549866631627083,
      "epoch": 0.385598116493162,
      "grad_norm": 9.21172046661377,
      "learning_rate": 4.07405580290565e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8404435843229294,
      "num_tokens": 149583130.0,
      "step": 124390
    },
    {
      "entropy": 1.7634894296526908,
      "epoch": 0.38562911561821167,
      "grad_norm": 2.975172996520996,
      "learning_rate": 4.073892050074802e-06,
      "loss": 0.3846,
      "mean_token_accuracy": 0.8556309074163437,
      "num_tokens": 149597170.0,
      "step": 124400
    },
    {
      "entropy": 1.9400777205824853,
      "epoch": 0.3856601147432614,
      "grad_norm": 4.058889389038086,
      "learning_rate": 4.073728316988036e-06,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8479943916201591,
      "num_tokens": 149608439.0,
      "step": 124410
    },
    {
      "entropy": 1.9313759356737137,
      "epoch": 0.38569111386831106,
      "grad_norm": 9.538460731506348,
      "learning_rate": 4.073564603641389e-06,
      "loss": 0.5035,
      "mean_token_accuracy": 0.8419632911682129,
      "num_tokens": 149619530.0,
      "step": 124420
    },
    {
      "entropy": 1.8018861994147302,
      "epoch": 0.3857221129933608,
      "grad_norm": 3.8183743953704834,
      "learning_rate": 4.073400910030892e-06,
      "loss": 0.3968,
      "mean_token_accuracy": 0.8584095433354377,
      "num_tokens": 149632347.0,
      "step": 124430
    },
    {
      "entropy": 1.8204200729727744,
      "epoch": 0.38575311211841046,
      "grad_norm": 4.118322849273682,
      "learning_rate": 4.073237236152582e-06,
      "loss": 0.39,
      "mean_token_accuracy": 0.8663435325026512,
      "num_tokens": 149645135.0,
      "step": 124440
    },
    {
      "entropy": 1.8887554615736009,
      "epoch": 0.3857841112434601,
      "grad_norm": 2.7827911376953125,
      "learning_rate": 4.073073582002494e-06,
      "loss": 0.465,
      "mean_token_accuracy": 0.8540782734751702,
      "num_tokens": 149657595.0,
      "step": 124450
    },
    {
      "entropy": 1.8850025668740273,
      "epoch": 0.38581511036850985,
      "grad_norm": 8.43443489074707,
      "learning_rate": 4.0729099475766645e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8449831783771515,
      "num_tokens": 149669584.0,
      "step": 124460
    },
    {
      "entropy": 1.873554064333439,
      "epoch": 0.3858461094935595,
      "grad_norm": 7.510732173919678,
      "learning_rate": 4.072746332871133e-06,
      "loss": 0.4565,
      "mean_token_accuracy": 0.8536177009344101,
      "num_tokens": 149681728.0,
      "step": 124470
    },
    {
      "entropy": 1.9078126922249794,
      "epoch": 0.38587710861860924,
      "grad_norm": 7.718918800354004,
      "learning_rate": 4.072582737881938e-06,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8453854739665985,
      "num_tokens": 149693031.0,
      "step": 124480
    },
    {
      "entropy": 1.8811730653047563,
      "epoch": 0.3859081077436589,
      "grad_norm": 7.272776126861572,
      "learning_rate": 4.0724191626051195e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8460710808634758,
      "num_tokens": 149704960.0,
      "step": 124490
    },
    {
      "entropy": 1.930656510591507,
      "epoch": 0.38593910686870864,
      "grad_norm": 7.962609767913818,
      "learning_rate": 4.07225560703672e-06,
      "loss": 0.5241,
      "mean_token_accuracy": 0.8429863288998604,
      "num_tokens": 149716392.0,
      "step": 124500
    },
    {
      "entropy": 1.9329011127352715,
      "epoch": 0.3859701059937583,
      "grad_norm": 9.818818092346191,
      "learning_rate": 4.072092071172782e-06,
      "loss": 0.5077,
      "mean_token_accuracy": 0.8381221756339073,
      "num_tokens": 149727625.0,
      "step": 124510
    },
    {
      "entropy": 1.841926720738411,
      "epoch": 0.38600110511880803,
      "grad_norm": 9.0508394241333,
      "learning_rate": 4.071928555009349e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8424027681350708,
      "num_tokens": 149740643.0,
      "step": 124520
    },
    {
      "entropy": 1.8008276581764222,
      "epoch": 0.3860321042438577,
      "grad_norm": 7.965822219848633,
      "learning_rate": 4.071765058542466e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8529940769076347,
      "num_tokens": 149753969.0,
      "step": 124530
    },
    {
      "entropy": 1.9138699561357497,
      "epoch": 0.3860631033689074,
      "grad_norm": 4.05026388168335,
      "learning_rate": 4.07160158176818e-06,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8561643600463867,
      "num_tokens": 149765733.0,
      "step": 124540
    },
    {
      "entropy": 1.9274681612849236,
      "epoch": 0.3860941024939571,
      "grad_norm": 8.39354133605957,
      "learning_rate": 4.071438124682538e-06,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8445206269621849,
      "num_tokens": 149777474.0,
      "step": 124550
    },
    {
      "entropy": 1.8434103056788445,
      "epoch": 0.3861251016190068,
      "grad_norm": 3.958385944366455,
      "learning_rate": 4.071274687281586e-06,
      "loss": 0.4663,
      "mean_token_accuracy": 0.8522407054901123,
      "num_tokens": 149789629.0,
      "step": 124560
    },
    {
      "entropy": 1.9501070573925972,
      "epoch": 0.3861561007440565,
      "grad_norm": 3.8399457931518555,
      "learning_rate": 4.071111269561375e-06,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8507745549082756,
      "num_tokens": 149800781.0,
      "step": 124570
    },
    {
      "entropy": 1.8660836547613144,
      "epoch": 0.3861870998691062,
      "grad_norm": 3.3907036781311035,
      "learning_rate": 4.0709478715179555e-06,
      "loss": 0.441,
      "mean_token_accuracy": 0.852207650244236,
      "num_tokens": 149813258.0,
      "step": 124580
    },
    {
      "entropy": 1.9248910203576088,
      "epoch": 0.3862180989941559,
      "grad_norm": 3.793597459793091,
      "learning_rate": 4.070784493147379e-06,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8523493394255638,
      "num_tokens": 149824551.0,
      "step": 124590
    },
    {
      "entropy": 1.913597546517849,
      "epoch": 0.3862490981192056,
      "grad_norm": 8.392363548278809,
      "learning_rate": 4.0706211344457e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.846868097782135,
      "num_tokens": 149836161.0,
      "step": 124600
    },
    {
      "entropy": 1.8576731622219085,
      "epoch": 0.38628009724425527,
      "grad_norm": 8.112271308898926,
      "learning_rate": 4.070457795408968e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8608527705073357,
      "num_tokens": 149848097.0,
      "step": 124610
    },
    {
      "entropy": 1.8224573642015458,
      "epoch": 0.386311096369305,
      "grad_norm": 3.8690743446350098,
      "learning_rate": 4.0702944760332415e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8590767070651054,
      "num_tokens": 149860778.0,
      "step": 124620
    },
    {
      "entropy": 1.8621004924178124,
      "epoch": 0.38634209549435466,
      "grad_norm": 8.412699699401855,
      "learning_rate": 4.070131176314576e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8528893545269967,
      "num_tokens": 149872659.0,
      "step": 124630
    },
    {
      "entropy": 1.8653160884976387,
      "epoch": 0.3863730946194044,
      "grad_norm": 4.710209846496582,
      "learning_rate": 4.069967896249026e-06,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8521018907427788,
      "num_tokens": 149885265.0,
      "step": 124640
    },
    {
      "entropy": 1.8452029392123221,
      "epoch": 0.38640409374445406,
      "grad_norm": 7.326101779937744,
      "learning_rate": 4.069804635832654e-06,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8424705550074577,
      "num_tokens": 149897692.0,
      "step": 124650
    },
    {
      "entropy": 1.8934701785445214,
      "epoch": 0.3864350928695038,
      "grad_norm": 9.25935173034668,
      "learning_rate": 4.069641395061516e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.8364736765623093,
      "num_tokens": 149910077.0,
      "step": 124660
    },
    {
      "entropy": 1.900476099550724,
      "epoch": 0.38646609199455345,
      "grad_norm": 11.954851150512695,
      "learning_rate": 4.069478173931674e-06,
      "loss": 0.4723,
      "mean_token_accuracy": 0.8435856163501739,
      "num_tokens": 149921585.0,
      "step": 124670
    },
    {
      "entropy": 1.9049529254436492,
      "epoch": 0.3864970911196032,
      "grad_norm": 8.252809524536133,
      "learning_rate": 4.069314972439188e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8509178534150124,
      "num_tokens": 149933471.0,
      "step": 124680
    },
    {
      "entropy": 1.914153940975666,
      "epoch": 0.38652809024465284,
      "grad_norm": 8.16724681854248,
      "learning_rate": 4.0691517905801225e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8479727208614349,
      "num_tokens": 149944962.0,
      "step": 124690
    },
    {
      "entropy": 1.9179544657468797,
      "epoch": 0.3865590893697025,
      "grad_norm": 8.549003601074219,
      "learning_rate": 4.0689886283505405e-06,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8387503355741501,
      "num_tokens": 149956453.0,
      "step": 124700
    },
    {
      "entropy": 1.944225938618183,
      "epoch": 0.38659008849475224,
      "grad_norm": 8.267438888549805,
      "learning_rate": 4.068825485746507e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8414188399910927,
      "num_tokens": 149968058.0,
      "step": 124710
    },
    {
      "entropy": 1.8530795663595199,
      "epoch": 0.3866210876198019,
      "grad_norm": 3.6572959423065186,
      "learning_rate": 4.068662362764087e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8553397417068481,
      "num_tokens": 149980786.0,
      "step": 124720
    },
    {
      "entropy": 1.8606258913874627,
      "epoch": 0.38665208674485163,
      "grad_norm": 9.493535995483398,
      "learning_rate": 4.06849925939935e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8513687327504158,
      "num_tokens": 149994145.0,
      "step": 124730
    },
    {
      "entropy": 1.9374653339385985,
      "epoch": 0.3866830858699013,
      "grad_norm": 8.496316909790039,
      "learning_rate": 4.0683361756483615e-06,
      "loss": 0.514,
      "mean_token_accuracy": 0.8495217755436897,
      "num_tokens": 150005113.0,
      "step": 124740
    },
    {
      "entropy": 1.8425881132483481,
      "epoch": 0.386714084994951,
      "grad_norm": 8.17953872680664,
      "learning_rate": 4.068173111507192e-06,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8537844985723495,
      "num_tokens": 150018495.0,
      "step": 124750
    },
    {
      "entropy": 1.8482965901494026,
      "epoch": 0.3867450841200007,
      "grad_norm": 5.024632453918457,
      "learning_rate": 4.068010066971912e-06,
      "loss": 0.4146,
      "mean_token_accuracy": 0.8522414952516556,
      "num_tokens": 150031464.0,
      "step": 124760
    },
    {
      "entropy": 1.934024366736412,
      "epoch": 0.3867760832450504,
      "grad_norm": 8.86674976348877,
      "learning_rate": 4.067847042038591e-06,
      "loss": 0.4842,
      "mean_token_accuracy": 0.84689302444458,
      "num_tokens": 150043092.0,
      "step": 124770
    },
    {
      "entropy": 1.871715322136879,
      "epoch": 0.3868070823701001,
      "grad_norm": 3.58791184425354,
      "learning_rate": 4.067684036703305e-06,
      "loss": 0.4503,
      "mean_token_accuracy": 0.8503245860338211,
      "num_tokens": 150055771.0,
      "step": 124780
    },
    {
      "entropy": 1.833419594168663,
      "epoch": 0.3868380814951498,
      "grad_norm": 7.293220043182373,
      "learning_rate": 4.067521050962126e-06,
      "loss": 0.4072,
      "mean_token_accuracy": 0.8605907499790192,
      "num_tokens": 150068131.0,
      "step": 124790
    },
    {
      "entropy": 1.8409696131944657,
      "epoch": 0.3868690806201995,
      "grad_norm": 7.163329124450684,
      "learning_rate": 4.067358084811128e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.853056974709034,
      "num_tokens": 150080793.0,
      "step": 124800
    },
    {
      "entropy": 1.8847892254590988,
      "epoch": 0.3869000797452492,
      "grad_norm": 8.353142738342285,
      "learning_rate": 4.067195138246388e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8540623351931572,
      "num_tokens": 150093526.0,
      "step": 124810
    },
    {
      "entropy": 2.0080907315015795,
      "epoch": 0.3869310788702989,
      "grad_norm": 8.072431564331055,
      "learning_rate": 4.067032211263983e-06,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8375670999288559,
      "num_tokens": 150104134.0,
      "step": 124820
    },
    {
      "entropy": 1.8393555164337159,
      "epoch": 0.3869620779953486,
      "grad_norm": 2.448547840118408,
      "learning_rate": 4.06686930385999e-06,
      "loss": 0.431,
      "mean_token_accuracy": 0.8578300714492798,
      "num_tokens": 150116750.0,
      "step": 124830
    },
    {
      "entropy": 1.8307070322334766,
      "epoch": 0.38699307712039827,
      "grad_norm": 3.879143476486206,
      "learning_rate": 4.06670641603049e-06,
      "loss": 0.4256,
      "mean_token_accuracy": 0.8474024429917335,
      "num_tokens": 150130258.0,
      "step": 124840
    },
    {
      "entropy": 1.9666341543197632,
      "epoch": 0.387024076245448,
      "grad_norm": 7.913427352905273,
      "learning_rate": 4.066543547771561e-06,
      "loss": 0.5647,
      "mean_token_accuracy": 0.8228250458836556,
      "num_tokens": 150141397.0,
      "step": 124850
    },
    {
      "entropy": 1.9169279381632804,
      "epoch": 0.38705507537049766,
      "grad_norm": 7.764703750610352,
      "learning_rate": 4.066380699079287e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8474949941039085,
      "num_tokens": 150153315.0,
      "step": 124860
    },
    {
      "entropy": 1.9108419820666314,
      "epoch": 0.3870860744955474,
      "grad_norm": 8.689712524414062,
      "learning_rate": 4.066217869949748e-06,
      "loss": 0.4448,
      "mean_token_accuracy": 0.8523913621902466,
      "num_tokens": 150165774.0,
      "step": 124870
    },
    {
      "entropy": 1.9573454082012176,
      "epoch": 0.38711707362059705,
      "grad_norm": 7.767167091369629,
      "learning_rate": 4.0660550603790286e-06,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8440973252058029,
      "num_tokens": 150177549.0,
      "step": 124880
    },
    {
      "entropy": 1.8977736786007882,
      "epoch": 0.3871480727456468,
      "grad_norm": 4.0380425453186035,
      "learning_rate": 4.065892270363214e-06,
      "loss": 0.4515,
      "mean_token_accuracy": 0.8582114532589913,
      "num_tokens": 150189097.0,
      "step": 124890
    },
    {
      "entropy": 1.832365168631077,
      "epoch": 0.38717907187069645,
      "grad_norm": 4.158474922180176,
      "learning_rate": 4.06572949989839e-06,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8578600034117698,
      "num_tokens": 150202332.0,
      "step": 124900
    },
    {
      "entropy": 1.9247950717806817,
      "epoch": 0.38721007099574617,
      "grad_norm": 8.511116027832031,
      "learning_rate": 4.065566748980642e-06,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8546933025121689,
      "num_tokens": 150213722.0,
      "step": 124910
    },
    {
      "entropy": 1.8911121636629105,
      "epoch": 0.38724107012079584,
      "grad_norm": 7.204141616821289,
      "learning_rate": 4.065404017606059e-06,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8537615105509758,
      "num_tokens": 150225589.0,
      "step": 124920
    },
    {
      "entropy": 1.8968706488609315,
      "epoch": 0.38727206924584556,
      "grad_norm": 7.295899868011475,
      "learning_rate": 4.06524130577073e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8530194133520126,
      "num_tokens": 150236807.0,
      "step": 124930
    },
    {
      "entropy": 1.8373442202806474,
      "epoch": 0.38730306837089523,
      "grad_norm": 8.396409034729004,
      "learning_rate": 4.065078613470747e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8561060458421708,
      "num_tokens": 150249132.0,
      "step": 124940
    },
    {
      "entropy": 1.834753280878067,
      "epoch": 0.3873340674959449,
      "grad_norm": 8.531126022338867,
      "learning_rate": 4.0649159407021976e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8525745943188667,
      "num_tokens": 150261614.0,
      "step": 124950
    },
    {
      "entropy": 1.9869435787200929,
      "epoch": 0.3873650666209946,
      "grad_norm": 8.008123397827148,
      "learning_rate": 4.064753287461177e-06,
      "loss": 0.4939,
      "mean_token_accuracy": 0.8471731215715408,
      "num_tokens": 150272837.0,
      "step": 124960
    },
    {
      "entropy": 1.9212356433272362,
      "epoch": 0.3873960657460443,
      "grad_norm": 9.139792442321777,
      "learning_rate": 4.064590653743777e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8533764079213142,
      "num_tokens": 150284178.0,
      "step": 124970
    },
    {
      "entropy": 1.9715091735124588,
      "epoch": 0.387427064871094,
      "grad_norm": 7.504796504974365,
      "learning_rate": 4.0644280395460925e-06,
      "loss": 0.5083,
      "mean_token_accuracy": 0.8431293860077858,
      "num_tokens": 150295307.0,
      "step": 124980
    },
    {
      "entropy": 1.9665077716112136,
      "epoch": 0.3874580639961437,
      "grad_norm": 7.719574928283691,
      "learning_rate": 4.064265444864221e-06,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8392296627163887,
      "num_tokens": 150306726.0,
      "step": 124990
    },
    {
      "entropy": 1.8853733599185944,
      "epoch": 0.3874890631211934,
      "grad_norm": 7.892232894897461,
      "learning_rate": 4.064102869694256e-06,
      "loss": 0.494,
      "mean_token_accuracy": 0.84963990598917,
      "num_tokens": 150318729.0,
      "step": 125000
    },
    {
      "entropy": 1.8874075815081597,
      "epoch": 0.3875200622462431,
      "grad_norm": 8.320926666259766,
      "learning_rate": 4.063940314032298e-06,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8477603197097778,
      "num_tokens": 150329931.0,
      "step": 125010
    },
    {
      "entropy": 1.9008696138858796,
      "epoch": 0.3875510613712928,
      "grad_norm": 8.944568634033203,
      "learning_rate": 4.063777777874444e-06,
      "loss": 0.4467,
      "mean_token_accuracy": 0.8521299719810486,
      "num_tokens": 150341393.0,
      "step": 125020
    },
    {
      "entropy": 1.9297545284032822,
      "epoch": 0.3875820604963425,
      "grad_norm": 8.812226295471191,
      "learning_rate": 4.063615261216795e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8559518858790398,
      "num_tokens": 150352453.0,
      "step": 125030
    },
    {
      "entropy": 1.9456378057599069,
      "epoch": 0.3876130596213922,
      "grad_norm": 7.481411933898926,
      "learning_rate": 4.063452764055453e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8542866796255112,
      "num_tokens": 150363663.0,
      "step": 125040
    },
    {
      "entropy": 1.843149345368147,
      "epoch": 0.38764405874644187,
      "grad_norm": 7.64381742477417,
      "learning_rate": 4.063290286386518e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8573990285396575,
      "num_tokens": 150376498.0,
      "step": 125050
    },
    {
      "entropy": 1.926984567940235,
      "epoch": 0.3876750578714916,
      "grad_norm": 7.688129901885986,
      "learning_rate": 4.063127828206096e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8490166947245598,
      "num_tokens": 150387561.0,
      "step": 125060
    },
    {
      "entropy": 1.8717382565140723,
      "epoch": 0.38770605699654126,
      "grad_norm": 3.5536370277404785,
      "learning_rate": 4.062965389510289e-06,
      "loss": 0.4053,
      "mean_token_accuracy": 0.8601777657866478,
      "num_tokens": 150399917.0,
      "step": 125070
    },
    {
      "entropy": 1.8823929965496062,
      "epoch": 0.387737056121591,
      "grad_norm": 4.017459869384766,
      "learning_rate": 4.062802970295203e-06,
      "loss": 0.3831,
      "mean_token_accuracy": 0.8708006635308265,
      "num_tokens": 150412228.0,
      "step": 125080
    },
    {
      "entropy": 1.9448561638593673,
      "epoch": 0.38776805524664065,
      "grad_norm": 9.202611923217773,
      "learning_rate": 4.062640570556946e-06,
      "loss": 0.4823,
      "mean_token_accuracy": 0.842719966173172,
      "num_tokens": 150423234.0,
      "step": 125090
    },
    {
      "entropy": 1.9790111839771272,
      "epoch": 0.3877990543716904,
      "grad_norm": 8.097685813903809,
      "learning_rate": 4.062478190291623e-06,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8447407379746437,
      "num_tokens": 150434200.0,
      "step": 125100
    },
    {
      "entropy": 1.9440268695354461,
      "epoch": 0.38783005349674005,
      "grad_norm": 6.974407196044922,
      "learning_rate": 4.062315829495345e-06,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8498797222971917,
      "num_tokens": 150446535.0,
      "step": 125110
    },
    {
      "entropy": 1.7851212821900844,
      "epoch": 0.38786105262178977,
      "grad_norm": 3.838473320007324,
      "learning_rate": 4.062153488164221e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8568575367331505,
      "num_tokens": 150459866.0,
      "step": 125120
    },
    {
      "entropy": 1.8512066915631293,
      "epoch": 0.38789205174683944,
      "grad_norm": 8.369192123413086,
      "learning_rate": 4.0619911662943615e-06,
      "loss": 0.4093,
      "mean_token_accuracy": 0.8748730286955834,
      "num_tokens": 150471449.0,
      "step": 125130
    },
    {
      "entropy": 1.9556255728006362,
      "epoch": 0.38792305087188916,
      "grad_norm": 8.41415023803711,
      "learning_rate": 4.06182886388188e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8445400312542916,
      "num_tokens": 150482247.0,
      "step": 125140
    },
    {
      "entropy": 1.9603990465402603,
      "epoch": 0.38795404999693883,
      "grad_norm": 11.196761131286621,
      "learning_rate": 4.061666580922887e-06,
      "loss": 0.5023,
      "mean_token_accuracy": 0.842097707092762,
      "num_tokens": 150492674.0,
      "step": 125150
    },
    {
      "entropy": 1.9363124072551727,
      "epoch": 0.38798504912198856,
      "grad_norm": 7.762144565582275,
      "learning_rate": 4.0615043174135e-06,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8462110877037048,
      "num_tokens": 150503435.0,
      "step": 125160
    },
    {
      "entropy": 1.8757888361811639,
      "epoch": 0.3880160482470382,
      "grad_norm": 7.726643085479736,
      "learning_rate": 4.061342073349831e-06,
      "loss": 0.428,
      "mean_token_accuracy": 0.8505197063088417,
      "num_tokens": 150515945.0,
      "step": 125170
    },
    {
      "entropy": 1.903496977686882,
      "epoch": 0.38804704737208795,
      "grad_norm": 8.79697036743164,
      "learning_rate": 4.061179848727998e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8483191832900048,
      "num_tokens": 150527375.0,
      "step": 125180
    },
    {
      "entropy": 1.8405410438776015,
      "epoch": 0.3880780464971376,
      "grad_norm": 7.9003586769104,
      "learning_rate": 4.061017643544118e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.854458573460579,
      "num_tokens": 150540114.0,
      "step": 125190
    },
    {
      "entropy": 1.94074095338583,
      "epoch": 0.3881090456221873,
      "grad_norm": 7.372513294219971,
      "learning_rate": 4.06085545779431e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8478065714240074,
      "num_tokens": 150551950.0,
      "step": 125200
    },
    {
      "entropy": 1.9185340121388434,
      "epoch": 0.388140044747237,
      "grad_norm": 8.795543670654297,
      "learning_rate": 4.060693291474694e-06,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8613968342542648,
      "num_tokens": 150563662.0,
      "step": 125210
    },
    {
      "entropy": 1.780056294798851,
      "epoch": 0.3881710438722867,
      "grad_norm": 6.95949125289917,
      "learning_rate": 4.06053114458139e-06,
      "loss": 0.3812,
      "mean_token_accuracy": 0.8654720067977906,
      "num_tokens": 150576950.0,
      "step": 125220
    },
    {
      "entropy": 1.856193946301937,
      "epoch": 0.3882020429973364,
      "grad_norm": 6.373475074768066,
      "learning_rate": 4.060369017110518e-06,
      "loss": 0.4502,
      "mean_token_accuracy": 0.854848501086235,
      "num_tokens": 150589326.0,
      "step": 125230
    },
    {
      "entropy": 1.8003525391221047,
      "epoch": 0.3882330421223861,
      "grad_norm": 10.690299034118652,
      "learning_rate": 4.060206909058204e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8460852265357971,
      "num_tokens": 150602928.0,
      "step": 125240
    },
    {
      "entropy": 1.7654723808169366,
      "epoch": 0.3882640412474358,
      "grad_norm": 3.999009132385254,
      "learning_rate": 4.060044820420571e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8616338670253754,
      "num_tokens": 150616871.0,
      "step": 125250
    },
    {
      "entropy": 1.8826753467321395,
      "epoch": 0.38829504037248547,
      "grad_norm": 8.977141380310059,
      "learning_rate": 4.059882751193742e-06,
      "loss": 0.466,
      "mean_token_accuracy": 0.8482603415846824,
      "num_tokens": 150628654.0,
      "step": 125260
    },
    {
      "entropy": 1.904711978137493,
      "epoch": 0.3883260394975352,
      "grad_norm": 7.154592990875244,
      "learning_rate": 4.059720701373846e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8563105076551437,
      "num_tokens": 150639652.0,
      "step": 125270
    },
    {
      "entropy": 1.9130603343248367,
      "epoch": 0.38835703862258486,
      "grad_norm": 7.394834518432617,
      "learning_rate": 4.059558670957009e-06,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8381910189986229,
      "num_tokens": 150651907.0,
      "step": 125280
    },
    {
      "entropy": 1.8639254540205001,
      "epoch": 0.3883880377476346,
      "grad_norm": 7.72260856628418,
      "learning_rate": 4.059396659939359e-06,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8464089751243591,
      "num_tokens": 150664885.0,
      "step": 125290
    },
    {
      "entropy": 1.8498515799641608,
      "epoch": 0.38841903687268425,
      "grad_norm": 4.0547871589660645,
      "learning_rate": 4.059234668317025e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8645502269268036,
      "num_tokens": 150677100.0,
      "step": 125300
    },
    {
      "entropy": 1.844531959295273,
      "epoch": 0.388450035997734,
      "grad_norm": 8.105052947998047,
      "learning_rate": 4.059072696086137e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8442863315343857,
      "num_tokens": 150689825.0,
      "step": 125310
    },
    {
      "entropy": 1.894164504110813,
      "epoch": 0.38848103512278365,
      "grad_norm": 7.457192897796631,
      "learning_rate": 4.058910743242828e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8596911758184433,
      "num_tokens": 150701538.0,
      "step": 125320
    },
    {
      "entropy": 1.7797346115112305,
      "epoch": 0.38851203424783337,
      "grad_norm": 8.076257705688477,
      "learning_rate": 4.05874880978323e-06,
      "loss": 0.3842,
      "mean_token_accuracy": 0.8683399274945259,
      "num_tokens": 150713771.0,
      "step": 125330
    },
    {
      "entropy": 1.8412994965910912,
      "epoch": 0.38854303337288304,
      "grad_norm": 8.09342098236084,
      "learning_rate": 4.058586895703477e-06,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8563569858670235,
      "num_tokens": 150726360.0,
      "step": 125340
    },
    {
      "entropy": 1.8202768236398696,
      "epoch": 0.38857403249793276,
      "grad_norm": 8.33877182006836,
      "learning_rate": 4.058425000999703e-06,
      "loss": 0.3963,
      "mean_token_accuracy": 0.8674609377980232,
      "num_tokens": 150738959.0,
      "step": 125350
    },
    {
      "entropy": 1.8868903383612632,
      "epoch": 0.38860503162298243,
      "grad_norm": 6.489454746246338,
      "learning_rate": 4.058263125668045e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8557512402534485,
      "num_tokens": 150750974.0,
      "step": 125360
    },
    {
      "entropy": 1.8816938310861588,
      "epoch": 0.38863603074803216,
      "grad_norm": 3.8871240615844727,
      "learning_rate": 4.058101269704639e-06,
      "loss": 0.5132,
      "mean_token_accuracy": 0.8465664908289909,
      "num_tokens": 150763017.0,
      "step": 125370
    },
    {
      "entropy": 1.967480507493019,
      "epoch": 0.3886670298730818,
      "grad_norm": 8.47203254699707,
      "learning_rate": 4.0579394331056216e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.8347558289766311,
      "num_tokens": 150773877.0,
      "step": 125380
    },
    {
      "entropy": 1.846261352300644,
      "epoch": 0.38869802899813155,
      "grad_norm": 8.301435470581055,
      "learning_rate": 4.057777615867134e-06,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8551396802067757,
      "num_tokens": 150786304.0,
      "step": 125390
    },
    {
      "entropy": 1.8747814357280732,
      "epoch": 0.3887290281231812,
      "grad_norm": 4.218428134918213,
      "learning_rate": 4.057615817985316e-06,
      "loss": 0.4316,
      "mean_token_accuracy": 0.8523815110325813,
      "num_tokens": 150797931.0,
      "step": 125400
    },
    {
      "entropy": 1.90915547311306,
      "epoch": 0.38876002724823094,
      "grad_norm": 4.607741355895996,
      "learning_rate": 4.057454039456308e-06,
      "loss": 0.4806,
      "mean_token_accuracy": 0.8459027215838433,
      "num_tokens": 150809552.0,
      "step": 125410
    },
    {
      "entropy": 1.9121482491493225,
      "epoch": 0.3887910263732806,
      "grad_norm": 3.7354090213775635,
      "learning_rate": 4.057292280276254e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8472837626934051,
      "num_tokens": 150820587.0,
      "step": 125420
    },
    {
      "entropy": 1.8476616084575652,
      "epoch": 0.38882202549833034,
      "grad_norm": 8.451462745666504,
      "learning_rate": 4.057130540441295e-06,
      "loss": 0.4871,
      "mean_token_accuracy": 0.8486483857035637,
      "num_tokens": 150832835.0,
      "step": 125430
    },
    {
      "entropy": 1.8385225757956505,
      "epoch": 0.38885302462338,
      "grad_norm": 10.56385612487793,
      "learning_rate": 4.0569688199475765e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8505441263318062,
      "num_tokens": 150845471.0,
      "step": 125440
    },
    {
      "entropy": 1.8741918548941612,
      "epoch": 0.3888840237484297,
      "grad_norm": 7.1380085945129395,
      "learning_rate": 4.056807118791245e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8532653734087944,
      "num_tokens": 150857775.0,
      "step": 125450
    },
    {
      "entropy": 1.8643349304795265,
      "epoch": 0.3889150228734794,
      "grad_norm": 7.431187152862549,
      "learning_rate": 4.056645436968446e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8581862285733223,
      "num_tokens": 150869444.0,
      "step": 125460
    },
    {
      "entropy": 1.892384371161461,
      "epoch": 0.38894602199852907,
      "grad_norm": 5.618492603302002,
      "learning_rate": 4.056483774475327e-06,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8496564209461213,
      "num_tokens": 150881306.0,
      "step": 125470
    },
    {
      "entropy": 1.9383496299386025,
      "epoch": 0.3889770211235788,
      "grad_norm": 9.300127029418945,
      "learning_rate": 4.0563221313080375e-06,
      "loss": 0.482,
      "mean_token_accuracy": 0.8460114181041718,
      "num_tokens": 150892540.0,
      "step": 125480
    },
    {
      "entropy": 1.969851815700531,
      "epoch": 0.38900802024862846,
      "grad_norm": 10.012271881103516,
      "learning_rate": 4.056160507462727e-06,
      "loss": 0.5231,
      "mean_token_accuracy": 0.8429014101624489,
      "num_tokens": 150902915.0,
      "step": 125490
    },
    {
      "entropy": 1.801128013432026,
      "epoch": 0.3890390193736782,
      "grad_norm": 7.085505962371826,
      "learning_rate": 4.055998902935546e-06,
      "loss": 0.441,
      "mean_token_accuracy": 0.8595685541629792,
      "num_tokens": 150915697.0,
      "step": 125500
    },
    {
      "entropy": 1.899875347316265,
      "epoch": 0.38907001849872785,
      "grad_norm": 7.814671993255615,
      "learning_rate": 4.055837317722647e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8444872334599495,
      "num_tokens": 150927597.0,
      "step": 125510
    },
    {
      "entropy": 1.9567907482385636,
      "epoch": 0.3891010176237776,
      "grad_norm": 8.426041603088379,
      "learning_rate": 4.055675751820183e-06,
      "loss": 0.5164,
      "mean_token_accuracy": 0.8452389568090439,
      "num_tokens": 150938164.0,
      "step": 125520
    },
    {
      "entropy": 1.9653800070285796,
      "epoch": 0.38913201674882725,
      "grad_norm": 8.481525421142578,
      "learning_rate": 4.055514205224307e-06,
      "loss": 0.544,
      "mean_token_accuracy": 0.8337672173976898,
      "num_tokens": 150948930.0,
      "step": 125530
    },
    {
      "entropy": 1.8959671720862388,
      "epoch": 0.38916301587387697,
      "grad_norm": 3.948054075241089,
      "learning_rate": 4.055352677931176e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8531327813863754,
      "num_tokens": 150960961.0,
      "step": 125540
    },
    {
      "entropy": 1.934750607609749,
      "epoch": 0.38919401499892664,
      "grad_norm": 10.166702270507812,
      "learning_rate": 4.055191169936945e-06,
      "loss": 0.5107,
      "mean_token_accuracy": 0.8438746586441994,
      "num_tokens": 150971612.0,
      "step": 125550
    },
    {
      "entropy": 1.9563965529203415,
      "epoch": 0.38922501412397636,
      "grad_norm": 9.370492935180664,
      "learning_rate": 4.055029681237772e-06,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8403390854597091,
      "num_tokens": 150982858.0,
      "step": 125560
    },
    {
      "entropy": 1.8438422948122024,
      "epoch": 0.38925601324902603,
      "grad_norm": 7.086329460144043,
      "learning_rate": 4.054868211829815e-06,
      "loss": 0.3796,
      "mean_token_accuracy": 0.8686971604824066,
      "num_tokens": 150994866.0,
      "step": 125570
    },
    {
      "entropy": 1.858051958680153,
      "epoch": 0.38928701237407576,
      "grad_norm": 8.350471496582031,
      "learning_rate": 4.054706761709233e-06,
      "loss": 0.4018,
      "mean_token_accuracy": 0.8600917294621467,
      "num_tokens": 151007774.0,
      "step": 125580
    },
    {
      "entropy": 1.8484682857990264,
      "epoch": 0.3893180114991254,
      "grad_norm": 8.876520156860352,
      "learning_rate": 4.054545330872188e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.853037391602993,
      "num_tokens": 151020015.0,
      "step": 125590
    },
    {
      "entropy": 1.8667544916272163,
      "epoch": 0.38934901062417515,
      "grad_norm": 8.567416191101074,
      "learning_rate": 4.054383919314841e-06,
      "loss": 0.4612,
      "mean_token_accuracy": 0.8439093992114067,
      "num_tokens": 151033160.0,
      "step": 125600
    },
    {
      "entropy": 1.825586286187172,
      "epoch": 0.3893800097492248,
      "grad_norm": 6.282723426818848,
      "learning_rate": 4.054222527033354e-06,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8462097451090813,
      "num_tokens": 151046530.0,
      "step": 125610
    },
    {
      "entropy": 1.9695552736520767,
      "epoch": 0.38941100887427454,
      "grad_norm": 11.352642059326172,
      "learning_rate": 4.054061154023891e-06,
      "loss": 0.5256,
      "mean_token_accuracy": 0.8349577113986015,
      "num_tokens": 151057681.0,
      "step": 125620
    },
    {
      "entropy": 1.8230530142784118,
      "epoch": 0.3894420079993242,
      "grad_norm": 3.0121681690216064,
      "learning_rate": 4.053899800282617e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.8639827251434327,
      "num_tokens": 151070490.0,
      "step": 125630
    },
    {
      "entropy": 1.9242575734853744,
      "epoch": 0.38947300712437394,
      "grad_norm": 7.926018714904785,
      "learning_rate": 4.053738465805698e-06,
      "loss": 0.4306,
      "mean_token_accuracy": 0.8579060733318329,
      "num_tokens": 151083191.0,
      "step": 125640
    },
    {
      "entropy": 1.8758929014205932,
      "epoch": 0.3895040062494236,
      "grad_norm": 4.752248287200928,
      "learning_rate": 4.0535771505893e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8485000059008598,
      "num_tokens": 151096082.0,
      "step": 125650
    },
    {
      "entropy": 1.9501382157206535,
      "epoch": 0.38953500537447333,
      "grad_norm": 4.228458881378174,
      "learning_rate": 4.053415854629593e-06,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8403441533446312,
      "num_tokens": 151107333.0,
      "step": 125660
    },
    {
      "entropy": 1.985108458995819,
      "epoch": 0.389566004499523,
      "grad_norm": 9.141436576843262,
      "learning_rate": 4.053254577922745e-06,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8309772089123726,
      "num_tokens": 151118366.0,
      "step": 125670
    },
    {
      "entropy": 1.9115714818239211,
      "epoch": 0.3895970036245727,
      "grad_norm": 3.911010980606079,
      "learning_rate": 4.053093320464925e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8501477718353272,
      "num_tokens": 151130118.0,
      "step": 125680
    },
    {
      "entropy": 1.976603901386261,
      "epoch": 0.3896280027496224,
      "grad_norm": 8.365656852722168,
      "learning_rate": 4.0529320822523064e-06,
      "loss": 0.4979,
      "mean_token_accuracy": 0.8426365479826927,
      "num_tokens": 151140975.0,
      "step": 125690
    },
    {
      "entropy": 1.900614893436432,
      "epoch": 0.38965900187467206,
      "grad_norm": 8.18988037109375,
      "learning_rate": 4.05277086328106e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8563957184553146,
      "num_tokens": 151152889.0,
      "step": 125700
    },
    {
      "entropy": 1.9433655887842178,
      "epoch": 0.3896900009997218,
      "grad_norm": 7.366755962371826,
      "learning_rate": 4.05260966354736e-06,
      "loss": 0.4645,
      "mean_token_accuracy": 0.855003735423088,
      "num_tokens": 151164778.0,
      "step": 125710
    },
    {
      "entropy": 1.882521539926529,
      "epoch": 0.38972100012477146,
      "grad_norm": 7.148138523101807,
      "learning_rate": 4.05244848304738e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8558038398623466,
      "num_tokens": 151176942.0,
      "step": 125720
    },
    {
      "entropy": 1.8969984114170075,
      "epoch": 0.3897519992498212,
      "grad_norm": 6.991118907928467,
      "learning_rate": 4.052287321777295e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8517007067799568,
      "num_tokens": 151188748.0,
      "step": 125730
    },
    {
      "entropy": 1.8719248160719872,
      "epoch": 0.38978299837487085,
      "grad_norm": 8.159533500671387,
      "learning_rate": 4.052126179733283e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8400168597698212,
      "num_tokens": 151201409.0,
      "step": 125740
    },
    {
      "entropy": 1.8552237376570702,
      "epoch": 0.3898139974999206,
      "grad_norm": 4.326776504516602,
      "learning_rate": 4.051965056911522e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8551932379603386,
      "num_tokens": 151213645.0,
      "step": 125750
    },
    {
      "entropy": 1.88699069917202,
      "epoch": 0.38984499662497024,
      "grad_norm": 7.64647102355957,
      "learning_rate": 4.051803953308188e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8565610900521279,
      "num_tokens": 151225668.0,
      "step": 125760
    },
    {
      "entropy": 1.8817365534603596,
      "epoch": 0.38987599575001997,
      "grad_norm": 9.160236358642578,
      "learning_rate": 4.051642868919464e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8513475954532623,
      "num_tokens": 151238004.0,
      "step": 125770
    },
    {
      "entropy": 1.8975436851382255,
      "epoch": 0.38990699487506963,
      "grad_norm": 6.79841423034668,
      "learning_rate": 4.0514818037415285e-06,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8526737794280053,
      "num_tokens": 151250091.0,
      "step": 125780
    },
    {
      "entropy": 1.9469924926757813,
      "epoch": 0.38993799400011936,
      "grad_norm": 7.432028770446777,
      "learning_rate": 4.051320757770564e-06,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8524128124117851,
      "num_tokens": 151261967.0,
      "step": 125790
    },
    {
      "entropy": 1.9410534769296646,
      "epoch": 0.38996899312516903,
      "grad_norm": 3.4470081329345703,
      "learning_rate": 4.051159731002754e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.844953115284443,
      "num_tokens": 151273163.0,
      "step": 125800
    },
    {
      "entropy": 1.9425782278180121,
      "epoch": 0.38999999225021875,
      "grad_norm": 7.442099094390869,
      "learning_rate": 4.050998723434281e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.8562641754746437,
      "num_tokens": 151284496.0,
      "step": 125810
    },
    {
      "entropy": 1.8365588143467904,
      "epoch": 0.3900309913752684,
      "grad_norm": 8.772604942321777,
      "learning_rate": 4.050837735061332e-06,
      "loss": 0.454,
      "mean_token_accuracy": 0.8577765494585037,
      "num_tokens": 151297315.0,
      "step": 125820
    },
    {
      "entropy": 1.8861859440803528,
      "epoch": 0.39006199050031815,
      "grad_norm": 7.54157829284668,
      "learning_rate": 4.050676765880091e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8518304273486137,
      "num_tokens": 151309596.0,
      "step": 125830
    },
    {
      "entropy": 1.8863596022129059,
      "epoch": 0.3900929896253678,
      "grad_norm": 9.19207763671875,
      "learning_rate": 4.0505158158867485e-06,
      "loss": 0.4917,
      "mean_token_accuracy": 0.8433287471532822,
      "num_tokens": 151320769.0,
      "step": 125840
    },
    {
      "entropy": 1.9262786269187928,
      "epoch": 0.39012398875041754,
      "grad_norm": 5.372649192810059,
      "learning_rate": 4.050354885077489e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.8311650440096855,
      "num_tokens": 151332882.0,
      "step": 125850
    },
    {
      "entropy": 1.913481391966343,
      "epoch": 0.3901549878754672,
      "grad_norm": 8.025369644165039,
      "learning_rate": 4.050193973448504e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8461648374795914,
      "num_tokens": 151344485.0,
      "step": 125860
    },
    {
      "entropy": 1.9113824039697647,
      "epoch": 0.39018598700051693,
      "grad_norm": 2.482351779937744,
      "learning_rate": 4.050033080995983e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8611611127853394,
      "num_tokens": 151356413.0,
      "step": 125870
    },
    {
      "entropy": 1.9198512472212315,
      "epoch": 0.3902169861255666,
      "grad_norm": 9.45190143585205,
      "learning_rate": 4.049872207716118e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.8463770866394043,
      "num_tokens": 151369155.0,
      "step": 125880
    },
    {
      "entropy": 1.9116858139634132,
      "epoch": 0.3902479852506163,
      "grad_norm": 7.803056716918945,
      "learning_rate": 4.0497113536051e-06,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8526840418577194,
      "num_tokens": 151381458.0,
      "step": 125890
    },
    {
      "entropy": 1.8976224765181542,
      "epoch": 0.390278984375666,
      "grad_norm": 9.358246803283691,
      "learning_rate": 4.049550518659125e-06,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8442497715353966,
      "num_tokens": 151393539.0,
      "step": 125900
    },
    {
      "entropy": 1.8757586926221848,
      "epoch": 0.3903099835007157,
      "grad_norm": 7.887486457824707,
      "learning_rate": 4.049389702874385e-06,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8386203497648239,
      "num_tokens": 151406150.0,
      "step": 125910
    },
    {
      "entropy": 1.851169802248478,
      "epoch": 0.3903409826257654,
      "grad_norm": 9.607558250427246,
      "learning_rate": 4.049228906247078e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8545200616121292,
      "num_tokens": 151419052.0,
      "step": 125920
    },
    {
      "entropy": 1.8620378583669663,
      "epoch": 0.39037198175081506,
      "grad_norm": 4.9543986320495605,
      "learning_rate": 4.049068128773399e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8555585101246834,
      "num_tokens": 151431366.0,
      "step": 125930
    },
    {
      "entropy": 1.9239520370960235,
      "epoch": 0.3904029808758648,
      "grad_norm": 7.769740104675293,
      "learning_rate": 4.048907370449547e-06,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8550285518169403,
      "num_tokens": 151442721.0,
      "step": 125940
    },
    {
      "entropy": 1.9248721554875374,
      "epoch": 0.39043398000091445,
      "grad_norm": 7.393733978271484,
      "learning_rate": 4.048746631271719e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.8553394272923469,
      "num_tokens": 151454600.0,
      "step": 125950
    },
    {
      "entropy": 1.8826577708125114,
      "epoch": 0.3904649791259642,
      "grad_norm": 8.594537734985352,
      "learning_rate": 4.048585911236118e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8587727382779121,
      "num_tokens": 151466942.0,
      "step": 125960
    },
    {
      "entropy": 1.8130075439810753,
      "epoch": 0.39049597825101384,
      "grad_norm": 9.427589416503906,
      "learning_rate": 4.048425210338942e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8479527696967125,
      "num_tokens": 151480312.0,
      "step": 125970
    },
    {
      "entropy": 1.9348285168409347,
      "epoch": 0.39052697737606357,
      "grad_norm": 8.240906715393066,
      "learning_rate": 4.048264528576393e-06,
      "loss": 0.5186,
      "mean_token_accuracy": 0.8450803935527802,
      "num_tokens": 151491045.0,
      "step": 125980
    },
    {
      "entropy": 1.8613590627908707,
      "epoch": 0.39055797650111324,
      "grad_norm": 4.4133734703063965,
      "learning_rate": 4.048103865944676e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.8640340849757194,
      "num_tokens": 151502174.0,
      "step": 125990
    },
    {
      "entropy": 1.905706176161766,
      "epoch": 0.39058897562616296,
      "grad_norm": 7.541008949279785,
      "learning_rate": 4.047943222439993e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8492152616381645,
      "num_tokens": 151513749.0,
      "step": 126000
    },
    {
      "entropy": 1.88614112585783,
      "epoch": 0.39061997475121263,
      "grad_norm": 8.197542190551758,
      "learning_rate": 4.047782598058551e-06,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8495625197887421,
      "num_tokens": 151525450.0,
      "step": 126010
    },
    {
      "entropy": 1.8722525179386138,
      "epoch": 0.39065097387626235,
      "grad_norm": 7.363889217376709,
      "learning_rate": 4.047621992796555e-06,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8617817178368569,
      "num_tokens": 151538034.0,
      "step": 126020
    },
    {
      "entropy": 1.9224463373422622,
      "epoch": 0.390681973001312,
      "grad_norm": 9.672123908996582,
      "learning_rate": 4.047461406650214e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8521727249026299,
      "num_tokens": 151549821.0,
      "step": 126030
    },
    {
      "entropy": 1.9509896606206893,
      "epoch": 0.39071297212636175,
      "grad_norm": 7.809655666351318,
      "learning_rate": 4.047300839615734e-06,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8400021120905876,
      "num_tokens": 151561202.0,
      "step": 126040
    },
    {
      "entropy": 1.8582612618803978,
      "epoch": 0.3907439712514114,
      "grad_norm": 4.203096866607666,
      "learning_rate": 4.047140291689325e-06,
      "loss": 0.4059,
      "mean_token_accuracy": 0.864464844763279,
      "num_tokens": 151574760.0,
      "step": 126050
    },
    {
      "entropy": 1.9256421521306037,
      "epoch": 0.39077497037646114,
      "grad_norm": 7.765346050262451,
      "learning_rate": 4.0469797628671996e-06,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8578554153442383,
      "num_tokens": 151586713.0,
      "step": 126060
    },
    {
      "entropy": 1.9531617119908333,
      "epoch": 0.3908059695015108,
      "grad_norm": 7.535767078399658,
      "learning_rate": 4.046819253145565e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8484017476439476,
      "num_tokens": 151598618.0,
      "step": 126070
    },
    {
      "entropy": 1.8639555156230927,
      "epoch": 0.39083696862656053,
      "grad_norm": 8.342696189880371,
      "learning_rate": 4.046658762520638e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8571187525987625,
      "num_tokens": 151610588.0,
      "step": 126080
    },
    {
      "entropy": 1.8678423702716827,
      "epoch": 0.3908679677516102,
      "grad_norm": 3.639899730682373,
      "learning_rate": 4.046498290988629e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8643327265977859,
      "num_tokens": 151622625.0,
      "step": 126090
    },
    {
      "entropy": 1.8767968013882637,
      "epoch": 0.3908989668766599,
      "grad_norm": 7.925841808319092,
      "learning_rate": 4.046337838545756e-06,
      "loss": 0.4044,
      "mean_token_accuracy": 0.8563542291522026,
      "num_tokens": 151636178.0,
      "step": 126100
    },
    {
      "entropy": 1.8535062327980996,
      "epoch": 0.3909299660017096,
      "grad_norm": 6.453927516937256,
      "learning_rate": 4.046177405188231e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8552742913365364,
      "num_tokens": 151649219.0,
      "step": 126110
    },
    {
      "entropy": 1.9538506001234055,
      "epoch": 0.3909609651267593,
      "grad_norm": 8.822364807128906,
      "learning_rate": 4.046016990912272e-06,
      "loss": 0.4634,
      "mean_token_accuracy": 0.8550143092870712,
      "num_tokens": 151660943.0,
      "step": 126120
    },
    {
      "entropy": 1.8665287435054778,
      "epoch": 0.390991964251809,
      "grad_norm": 8.066899299621582,
      "learning_rate": 4.045856595714099e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8569189369678497,
      "num_tokens": 151673423.0,
      "step": 126130
    },
    {
      "entropy": 1.9262668624520303,
      "epoch": 0.3910229633768587,
      "grad_norm": 7.15204381942749,
      "learning_rate": 4.045696219589927e-06,
      "loss": 0.5058,
      "mean_token_accuracy": 0.847204127907753,
      "num_tokens": 151684997.0,
      "step": 126140
    },
    {
      "entropy": 1.9528811901807785,
      "epoch": 0.3910539625019084,
      "grad_norm": 8.068526268005371,
      "learning_rate": 4.045535862535979e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8466339007019996,
      "num_tokens": 151696214.0,
      "step": 126150
    },
    {
      "entropy": 1.8696447387337685,
      "epoch": 0.3910849616269581,
      "grad_norm": 8.606704711914062,
      "learning_rate": 4.045375524548474e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.852445213496685,
      "num_tokens": 151708632.0,
      "step": 126160
    },
    {
      "entropy": 1.899414698779583,
      "epoch": 0.3911159607520078,
      "grad_norm": 7.6843342781066895,
      "learning_rate": 4.0452152056236355e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8635346978902817,
      "num_tokens": 151720657.0,
      "step": 126170
    },
    {
      "entropy": 1.8455950021743774,
      "epoch": 0.39114695987705744,
      "grad_norm": 8.676742553710938,
      "learning_rate": 4.045054905757685e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8539397269487381,
      "num_tokens": 151732951.0,
      "step": 126180
    },
    {
      "entropy": 1.9040052101016045,
      "epoch": 0.39117795900210717,
      "grad_norm": 3.939152956008911,
      "learning_rate": 4.044894624946848e-06,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8455827251076699,
      "num_tokens": 151744428.0,
      "step": 126190
    },
    {
      "entropy": 1.840182974934578,
      "epoch": 0.39120895812715684,
      "grad_norm": 7.910426616668701,
      "learning_rate": 4.044734363187349e-06,
      "loss": 0.4163,
      "mean_token_accuracy": 0.8497889280319214,
      "num_tokens": 151757200.0,
      "step": 126200
    },
    {
      "entropy": 1.9612287282943726,
      "epoch": 0.39123995725220656,
      "grad_norm": 8.55207347869873,
      "learning_rate": 4.044574120475414e-06,
      "loss": 0.5217,
      "mean_token_accuracy": 0.8453647062182427,
      "num_tokens": 151768054.0,
      "step": 126210
    },
    {
      "entropy": 1.925032651424408,
      "epoch": 0.39127095637725623,
      "grad_norm": 10.712833404541016,
      "learning_rate": 4.044413896807269e-06,
      "loss": 0.5053,
      "mean_token_accuracy": 0.845679797232151,
      "num_tokens": 151778718.0,
      "step": 126220
    },
    {
      "entropy": 1.9098077476024629,
      "epoch": 0.39130195550230595,
      "grad_norm": 10.31223201751709,
      "learning_rate": 4.044253692179145e-06,
      "loss": 0.471,
      "mean_token_accuracy": 0.8485254973173142,
      "num_tokens": 151789936.0,
      "step": 126230
    },
    {
      "entropy": 1.926163762807846,
      "epoch": 0.3913329546273556,
      "grad_norm": 7.671618938446045,
      "learning_rate": 4.044093506587268e-06,
      "loss": 0.4995,
      "mean_token_accuracy": 0.8456138476729393,
      "num_tokens": 151800690.0,
      "step": 126240
    },
    {
      "entropy": 1.9659644454717635,
      "epoch": 0.39136395375240535,
      "grad_norm": 10.632147789001465,
      "learning_rate": 4.043933340027872e-06,
      "loss": 0.5037,
      "mean_token_accuracy": 0.8462962001562119,
      "num_tokens": 151811555.0,
      "step": 126250
    },
    {
      "entropy": 1.9211021691560746,
      "epoch": 0.391394952877455,
      "grad_norm": 8.475778579711914,
      "learning_rate": 4.043773192497186e-06,
      "loss": 0.455,
      "mean_token_accuracy": 0.8569527640938759,
      "num_tokens": 151823099.0,
      "step": 126260
    },
    {
      "entropy": 1.8699581772089005,
      "epoch": 0.39142595200250474,
      "grad_norm": 7.7866530418396,
      "learning_rate": 4.043613063991444e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8597669988870621,
      "num_tokens": 151834750.0,
      "step": 126270
    },
    {
      "entropy": 1.9547498047351837,
      "epoch": 0.3914569511275544,
      "grad_norm": 8.300743103027344,
      "learning_rate": 4.043452954506876e-06,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8422427833080292,
      "num_tokens": 151846263.0,
      "step": 126280
    },
    {
      "entropy": 1.8979168817400933,
      "epoch": 0.39148795025260413,
      "grad_norm": 9.36401653289795,
      "learning_rate": 4.04329286403972e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8518133416771889,
      "num_tokens": 151857975.0,
      "step": 126290
    },
    {
      "entropy": 1.8685284316539765,
      "epoch": 0.3915189493776538,
      "grad_norm": 7.125948429107666,
      "learning_rate": 4.043132792586211e-06,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8564907148480415,
      "num_tokens": 151869148.0,
      "step": 126300
    },
    {
      "entropy": 1.932071290910244,
      "epoch": 0.3915499485027035,
      "grad_norm": 8.5045804977417,
      "learning_rate": 4.042972740142585e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.85165533721447,
      "num_tokens": 151880669.0,
      "step": 126310
    },
    {
      "entropy": 1.898838460445404,
      "epoch": 0.3915809476277532,
      "grad_norm": 8.528064727783203,
      "learning_rate": 4.04281270670508e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8440765485167503,
      "num_tokens": 151893003.0,
      "step": 126320
    },
    {
      "entropy": 1.8209847897291183,
      "epoch": 0.3916119467528029,
      "grad_norm": 9.340559959411621,
      "learning_rate": 4.042652692269934e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.859174670279026,
      "num_tokens": 151905874.0,
      "step": 126330
    },
    {
      "entropy": 1.95557002723217,
      "epoch": 0.3916429458778526,
      "grad_norm": 8.08029842376709,
      "learning_rate": 4.042492696833388e-06,
      "loss": 0.5362,
      "mean_token_accuracy": 0.8356934756040573,
      "num_tokens": 151917270.0,
      "step": 126340
    },
    {
      "entropy": 1.8023017689585685,
      "epoch": 0.3916739450029023,
      "grad_norm": 8.126782417297363,
      "learning_rate": 4.042332720391681e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8570797294378281,
      "num_tokens": 151931119.0,
      "step": 126350
    },
    {
      "entropy": 1.881139089167118,
      "epoch": 0.391704944127952,
      "grad_norm": 8.571240425109863,
      "learning_rate": 4.042172762941057e-06,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8508111834526062,
      "num_tokens": 151942409.0,
      "step": 126360
    },
    {
      "entropy": 1.8666364967823028,
      "epoch": 0.3917359432530017,
      "grad_norm": 8.6209077835083,
      "learning_rate": 4.042012824477757e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8415684968233108,
      "num_tokens": 151954323.0,
      "step": 126370
    },
    {
      "entropy": 1.8725038409233092,
      "epoch": 0.3917669423780514,
      "grad_norm": 7.019345760345459,
      "learning_rate": 4.041852904998025e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.850992114841938,
      "num_tokens": 151967314.0,
      "step": 126380
    },
    {
      "entropy": 1.8260181456804276,
      "epoch": 0.3917979415031011,
      "grad_norm": 8.401409149169922,
      "learning_rate": 4.0416930044981064e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.84242385327816,
      "num_tokens": 151980943.0,
      "step": 126390
    },
    {
      "entropy": 1.9128731414675713,
      "epoch": 0.39182894062815077,
      "grad_norm": 3.6636550426483154,
      "learning_rate": 4.041533122974248e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8557316899299622,
      "num_tokens": 151992446.0,
      "step": 126400
    },
    {
      "entropy": 1.8801569387316703,
      "epoch": 0.3918599397532005,
      "grad_norm": 4.1143107414245605,
      "learning_rate": 4.041373260422696e-06,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8517079591751099,
      "num_tokens": 152004499.0,
      "step": 126410
    },
    {
      "entropy": 1.9254588589072228,
      "epoch": 0.39189093887825016,
      "grad_norm": 7.596971035003662,
      "learning_rate": 4.041213416839698e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8518746435642243,
      "num_tokens": 152016295.0,
      "step": 126420
    },
    {
      "entropy": 1.9142033576965332,
      "epoch": 0.39192193800329983,
      "grad_norm": 7.351873397827148,
      "learning_rate": 4.0410535922215046e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8495833858847618,
      "num_tokens": 152027558.0,
      "step": 126430
    },
    {
      "entropy": 1.892107328772545,
      "epoch": 0.39195293712834955,
      "grad_norm": 9.62601089477539,
      "learning_rate": 4.040893786564364e-06,
      "loss": 0.4711,
      "mean_token_accuracy": 0.8505282923579216,
      "num_tokens": 152039402.0,
      "step": 126440
    },
    {
      "entropy": 1.8616481512784957,
      "epoch": 0.3919839362533992,
      "grad_norm": 7.576359748840332,
      "learning_rate": 4.040733999864529e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8497416257858277,
      "num_tokens": 152051794.0,
      "step": 126450
    },
    {
      "entropy": 1.7416065499186515,
      "epoch": 0.39201493537844895,
      "grad_norm": 4.21736478805542,
      "learning_rate": 4.0405742321182505e-06,
      "loss": 0.3715,
      "mean_token_accuracy": 0.857705582678318,
      "num_tokens": 152065713.0,
      "step": 126460
    },
    {
      "entropy": 1.8679188892245293,
      "epoch": 0.3920459345034986,
      "grad_norm": 7.051849365234375,
      "learning_rate": 4.040414483321783e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8559218123555183,
      "num_tokens": 152078414.0,
      "step": 126470
    },
    {
      "entropy": 1.9041266560554504,
      "epoch": 0.39207693362854834,
      "grad_norm": 8.104182243347168,
      "learning_rate": 4.0402547534713795e-06,
      "loss": 0.4943,
      "mean_token_accuracy": 0.8375504925847054,
      "num_tokens": 152090227.0,
      "step": 126480
    },
    {
      "entropy": 1.9575203120708466,
      "epoch": 0.392107932753598,
      "grad_norm": 7.981137752532959,
      "learning_rate": 4.0400950425632965e-06,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8419002950191498,
      "num_tokens": 152101040.0,
      "step": 126490
    },
    {
      "entropy": 1.907607588171959,
      "epoch": 0.39213893187864773,
      "grad_norm": 8.032891273498535,
      "learning_rate": 4.03993535059379e-06,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8457751244306564,
      "num_tokens": 152113002.0,
      "step": 126500
    },
    {
      "entropy": 1.8676989868283271,
      "epoch": 0.3921699310036974,
      "grad_norm": 8.83267879486084,
      "learning_rate": 4.039775677559117e-06,
      "loss": 0.4944,
      "mean_token_accuracy": 0.839282539486885,
      "num_tokens": 152125801.0,
      "step": 126510
    },
    {
      "entropy": 1.943473118543625,
      "epoch": 0.3922009301287471,
      "grad_norm": 9.639033317565918,
      "learning_rate": 4.0396160234555365e-06,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8399049505591393,
      "num_tokens": 152137319.0,
      "step": 126520
    },
    {
      "entropy": 1.885578916966915,
      "epoch": 0.3922319292537968,
      "grad_norm": 4.101627349853516,
      "learning_rate": 4.039456388279307e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8486464694142342,
      "num_tokens": 152148855.0,
      "step": 126530
    },
    {
      "entropy": 1.8974051237106324,
      "epoch": 0.3922629283788465,
      "grad_norm": 4.128299713134766,
      "learning_rate": 4.03929677202669e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8376821115612983,
      "num_tokens": 152161206.0,
      "step": 126540
    },
    {
      "entropy": 1.8535172313451767,
      "epoch": 0.3922939275038962,
      "grad_norm": 8.3539400100708,
      "learning_rate": 4.0391371746939465e-06,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8649252921342849,
      "num_tokens": 152172699.0,
      "step": 126550
    },
    {
      "entropy": 1.954978096485138,
      "epoch": 0.3923249266289459,
      "grad_norm": 3.532893180847168,
      "learning_rate": 4.03897759627734e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8465725928544998,
      "num_tokens": 152184753.0,
      "step": 126560
    },
    {
      "entropy": 1.9255331978201866,
      "epoch": 0.3923559257539956,
      "grad_norm": 10.25304126739502,
      "learning_rate": 4.038818036773132e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8339841410517692,
      "num_tokens": 152196175.0,
      "step": 126570
    },
    {
      "entropy": 1.893292284011841,
      "epoch": 0.3923869248790453,
      "grad_norm": 4.092190742492676,
      "learning_rate": 4.038658496177589e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8404751896858216,
      "num_tokens": 152208309.0,
      "step": 126580
    },
    {
      "entropy": 1.892499789595604,
      "epoch": 0.392417924004095,
      "grad_norm": 7.2047438621521,
      "learning_rate": 4.0384989744869764e-06,
      "loss": 0.4035,
      "mean_token_accuracy": 0.8644706353545188,
      "num_tokens": 152220743.0,
      "step": 126590
    },
    {
      "entropy": 1.8051872000098228,
      "epoch": 0.3924489231291447,
      "grad_norm": 8.140718460083008,
      "learning_rate": 4.03833947169756e-06,
      "loss": 0.4284,
      "mean_token_accuracy": 0.857630018889904,
      "num_tokens": 152234128.0,
      "step": 126600
    },
    {
      "entropy": 1.9284190341830254,
      "epoch": 0.39247992225419437,
      "grad_norm": 8.732551574707031,
      "learning_rate": 4.038179987805609e-06,
      "loss": 0.4658,
      "mean_token_accuracy": 0.8481160834431648,
      "num_tokens": 152245832.0,
      "step": 126610
    },
    {
      "entropy": 1.8666399344801903,
      "epoch": 0.3925109213792441,
      "grad_norm": 5.154988765716553,
      "learning_rate": 4.038020522807391e-06,
      "loss": 0.4379,
      "mean_token_accuracy": 0.8664086773991585,
      "num_tokens": 152258281.0,
      "step": 126620
    },
    {
      "entropy": 1.9185226619243623,
      "epoch": 0.39254192050429376,
      "grad_norm": 8.254905700683594,
      "learning_rate": 4.037861076699175e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8595322445034981,
      "num_tokens": 152270283.0,
      "step": 126630
    },
    {
      "entropy": 1.881806494295597,
      "epoch": 0.3925729196293435,
      "grad_norm": 7.021206378936768,
      "learning_rate": 4.037701649477234e-06,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8548443630337715,
      "num_tokens": 152282446.0,
      "step": 126640
    },
    {
      "entropy": 1.9258292108774184,
      "epoch": 0.39260391875439316,
      "grad_norm": 8.01991081237793,
      "learning_rate": 4.037542241137839e-06,
      "loss": 0.489,
      "mean_token_accuracy": 0.8480772167444229,
      "num_tokens": 152294011.0,
      "step": 126650
    },
    {
      "entropy": 1.976582020521164,
      "epoch": 0.3926349178794429,
      "grad_norm": 8.293559074401855,
      "learning_rate": 4.0373828516772615e-06,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8434167116880417,
      "num_tokens": 152304686.0,
      "step": 126660
    },
    {
      "entropy": 1.8964557886123656,
      "epoch": 0.39266591700449255,
      "grad_norm": 8.829238891601562,
      "learning_rate": 4.037223481091777e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8579582333564758,
      "num_tokens": 152316928.0,
      "step": 126670
    },
    {
      "entropy": 1.889539574831724,
      "epoch": 0.3926969161295422,
      "grad_norm": 9.728771209716797,
      "learning_rate": 4.037064129377661e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8489784002304077,
      "num_tokens": 152329516.0,
      "step": 126680
    },
    {
      "entropy": 1.9084494978189468,
      "epoch": 0.39272791525459194,
      "grad_norm": 8.081507682800293,
      "learning_rate": 4.036904796531187e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8519722208380699,
      "num_tokens": 152340705.0,
      "step": 126690
    },
    {
      "entropy": 1.9536496102809906,
      "epoch": 0.3927589143796416,
      "grad_norm": 7.832414150238037,
      "learning_rate": 4.036745482548634e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8512647017836571,
      "num_tokens": 152351867.0,
      "step": 126700
    },
    {
      "entropy": 1.8545057475566864,
      "epoch": 0.39278991350469133,
      "grad_norm": 3.9451944828033447,
      "learning_rate": 4.03658618742628e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8511400848627091,
      "num_tokens": 152364222.0,
      "step": 126710
    },
    {
      "entropy": 1.8836131229996682,
      "epoch": 0.392820912629741,
      "grad_norm": 7.37644624710083,
      "learning_rate": 4.036426911160403e-06,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8465021848678589,
      "num_tokens": 152376071.0,
      "step": 126720
    },
    {
      "entropy": 1.908482810854912,
      "epoch": 0.39285191175479073,
      "grad_norm": 8.756072998046875,
      "learning_rate": 4.036267653747283e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8456471160054206,
      "num_tokens": 152387712.0,
      "step": 126730
    },
    {
      "entropy": 1.8821537524461747,
      "epoch": 0.3928829108798404,
      "grad_norm": 6.746029853820801,
      "learning_rate": 4.036108415183202e-06,
      "loss": 0.4408,
      "mean_token_accuracy": 0.853573402762413,
      "num_tokens": 152399464.0,
      "step": 126740
    },
    {
      "entropy": 1.7690861240029334,
      "epoch": 0.3929139100048901,
      "grad_norm": 8.50271224975586,
      "learning_rate": 4.035949195464444e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8643131896853447,
      "num_tokens": 152412703.0,
      "step": 126750
    },
    {
      "entropy": 1.9057681828737258,
      "epoch": 0.3929449091299398,
      "grad_norm": 7.735318183898926,
      "learning_rate": 4.035789994587288e-06,
      "loss": 0.4516,
      "mean_token_accuracy": 0.8501883447170258,
      "num_tokens": 152423734.0,
      "step": 126760
    },
    {
      "entropy": 1.980392262339592,
      "epoch": 0.3929759082549895,
      "grad_norm": 9.195259094238281,
      "learning_rate": 4.0356308125480205e-06,
      "loss": 0.57,
      "mean_token_accuracy": 0.830004945397377,
      "num_tokens": 152435210.0,
      "step": 126770
    },
    {
      "entropy": 1.9139831706881523,
      "epoch": 0.3930069073800392,
      "grad_norm": 3.758761167526245,
      "learning_rate": 4.035471649342926e-06,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8497570350766182,
      "num_tokens": 152447602.0,
      "step": 126780
    },
    {
      "entropy": 1.9336750611662865,
      "epoch": 0.3930379065050889,
      "grad_norm": 7.929681777954102,
      "learning_rate": 4.035312504968292e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.852272717654705,
      "num_tokens": 152458841.0,
      "step": 126790
    },
    {
      "entropy": 1.922458928823471,
      "epoch": 0.3930689056301386,
      "grad_norm": 7.431619644165039,
      "learning_rate": 4.035153379420406e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8492286145687103,
      "num_tokens": 152469765.0,
      "step": 126800
    },
    {
      "entropy": 1.9320431604981423,
      "epoch": 0.3930999047551883,
      "grad_norm": 7.810932636260986,
      "learning_rate": 4.034994272695553e-06,
      "loss": 0.5431,
      "mean_token_accuracy": 0.8349621132016182,
      "num_tokens": 152481239.0,
      "step": 126810
    },
    {
      "entropy": 1.911438837647438,
      "epoch": 0.39313090388023797,
      "grad_norm": 9.275655746459961,
      "learning_rate": 4.0348351847900256e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8501843497157097,
      "num_tokens": 152492347.0,
      "step": 126820
    },
    {
      "entropy": 1.8184763744473458,
      "epoch": 0.3931619030052877,
      "grad_norm": 8.230948448181152,
      "learning_rate": 4.034676115700114e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8377979636192322,
      "num_tokens": 152505653.0,
      "step": 126830
    },
    {
      "entropy": 1.9131136581301689,
      "epoch": 0.39319290213033736,
      "grad_norm": 7.071003437042236,
      "learning_rate": 4.0345170654221075e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8539568200707436,
      "num_tokens": 152517329.0,
      "step": 126840
    },
    {
      "entropy": 1.9276820585131644,
      "epoch": 0.3932239012553871,
      "grad_norm": 7.994453430175781,
      "learning_rate": 4.0343580339523e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8405784666538239,
      "num_tokens": 152528760.0,
      "step": 126850
    },
    {
      "entropy": 1.8666841223835946,
      "epoch": 0.39325490038043676,
      "grad_norm": 9.046262741088867,
      "learning_rate": 4.034199021286984e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8560965120792389,
      "num_tokens": 152540919.0,
      "step": 126860
    },
    {
      "entropy": 1.8318147659301758,
      "epoch": 0.3932858995054865,
      "grad_norm": 9.050399780273438,
      "learning_rate": 4.0340400274224554e-06,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8550884276628494,
      "num_tokens": 152553211.0,
      "step": 126870
    },
    {
      "entropy": 1.8964627265930176,
      "epoch": 0.39331689863053615,
      "grad_norm": 7.297853946685791,
      "learning_rate": 4.033881052355008e-06,
      "loss": 0.467,
      "mean_token_accuracy": 0.840998500585556,
      "num_tokens": 152565550.0,
      "step": 126880
    },
    {
      "entropy": 1.8291507720947267,
      "epoch": 0.3933478977555859,
      "grad_norm": 4.372984886169434,
      "learning_rate": 4.03372209608094e-06,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8626144364476204,
      "num_tokens": 152578617.0,
      "step": 126890
    },
    {
      "entropy": 1.9425703272223473,
      "epoch": 0.39337889688063554,
      "grad_norm": 9.826508522033691,
      "learning_rate": 4.033563158596547e-06,
      "loss": 0.5442,
      "mean_token_accuracy": 0.8346266880631447,
      "num_tokens": 152589941.0,
      "step": 126900
    },
    {
      "entropy": 1.8560305163264275,
      "epoch": 0.39340989600568527,
      "grad_norm": 9.724489212036133,
      "learning_rate": 4.0334042398981285e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8553323075175285,
      "num_tokens": 152602105.0,
      "step": 126910
    },
    {
      "entropy": 1.921321277320385,
      "epoch": 0.39344089513073494,
      "grad_norm": 8.991716384887695,
      "learning_rate": 4.033245339981984e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8386275783181191,
      "num_tokens": 152613975.0,
      "step": 126920
    },
    {
      "entropy": 1.9884714633226395,
      "epoch": 0.3934718942557846,
      "grad_norm": 7.993074893951416,
      "learning_rate": 4.033086458844415e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.8479173853993416,
      "num_tokens": 152624593.0,
      "step": 126930
    },
    {
      "entropy": 1.8521944627165794,
      "epoch": 0.39350289338083433,
      "grad_norm": 8.29841136932373,
      "learning_rate": 4.032927596481721e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8486550331115723,
      "num_tokens": 152636893.0,
      "step": 126940
    },
    {
      "entropy": 1.853082077205181,
      "epoch": 0.393533892505884,
      "grad_norm": 8.913239479064941,
      "learning_rate": 4.032768752890206e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8620326921343804,
      "num_tokens": 152648688.0,
      "step": 126950
    },
    {
      "entropy": 1.9352153725922108,
      "epoch": 0.3935648916309337,
      "grad_norm": 7.439796447753906,
      "learning_rate": 4.032609928066174e-06,
      "loss": 0.4972,
      "mean_token_accuracy": 0.8431309580802917,
      "num_tokens": 152660197.0,
      "step": 126960
    },
    {
      "entropy": 1.9533848583698272,
      "epoch": 0.3935958907559834,
      "grad_norm": 12.858529090881348,
      "learning_rate": 4.032451122005929e-06,
      "loss": 0.4903,
      "mean_token_accuracy": 0.8557668402791023,
      "num_tokens": 152671469.0,
      "step": 126970
    },
    {
      "entropy": 1.872800037264824,
      "epoch": 0.3936268898810331,
      "grad_norm": 7.507105350494385,
      "learning_rate": 4.032292334705776e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8509954452514649,
      "num_tokens": 152684255.0,
      "step": 126980
    },
    {
      "entropy": 1.8584757506847382,
      "epoch": 0.3936578890060828,
      "grad_norm": 3.439596176147461,
      "learning_rate": 4.032133566162022e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.853032597899437,
      "num_tokens": 152696156.0,
      "step": 126990
    },
    {
      "entropy": 1.841444942355156,
      "epoch": 0.3936888881311325,
      "grad_norm": 8.641845703125,
      "learning_rate": 4.031974816370976e-06,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8564640492200851,
      "num_tokens": 152708857.0,
      "step": 127000
    },
    {
      "entropy": 1.9723876774311067,
      "epoch": 0.3937198872561822,
      "grad_norm": 8.145781517028809,
      "learning_rate": 4.031816085328946e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8606492772698402,
      "num_tokens": 152719903.0,
      "step": 127010
    },
    {
      "entropy": 1.8345705628395081,
      "epoch": 0.3937508863812319,
      "grad_norm": 7.235647201538086,
      "learning_rate": 4.03165737303224e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8646158143877983,
      "num_tokens": 152732756.0,
      "step": 127020
    },
    {
      "entropy": 1.9376291260123253,
      "epoch": 0.39378188550628157,
      "grad_norm": 3.9151713848114014,
      "learning_rate": 4.031498679477171e-06,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8439146399497985,
      "num_tokens": 152744727.0,
      "step": 127030
    },
    {
      "entropy": 1.903841333091259,
      "epoch": 0.3938128846313313,
      "grad_norm": 8.165589332580566,
      "learning_rate": 4.03134000466005e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.840572564303875,
      "num_tokens": 152756881.0,
      "step": 127040
    },
    {
      "entropy": 1.8646742686629296,
      "epoch": 0.39384388375638096,
      "grad_norm": 5.022664546966553,
      "learning_rate": 4.0311813485771896e-06,
      "loss": 0.4021,
      "mean_token_accuracy": 0.8504794493317605,
      "num_tokens": 152769508.0,
      "step": 127050
    },
    {
      "entropy": 1.9327761620283126,
      "epoch": 0.3938748828814307,
      "grad_norm": 7.481938362121582,
      "learning_rate": 4.031022711224904e-06,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8473081424832344,
      "num_tokens": 152781394.0,
      "step": 127060
    },
    {
      "entropy": 1.9920929014682769,
      "epoch": 0.39390588200648036,
      "grad_norm": 7.522092819213867,
      "learning_rate": 4.030864092599508e-06,
      "loss": 0.5138,
      "mean_token_accuracy": 0.8395040348172188,
      "num_tokens": 152792361.0,
      "step": 127070
    },
    {
      "entropy": 1.8420628361403941,
      "epoch": 0.3939368811315301,
      "grad_norm": 7.69534158706665,
      "learning_rate": 4.030705492697317e-06,
      "loss": 0.4029,
      "mean_token_accuracy": 0.8510744541883468,
      "num_tokens": 152805416.0,
      "step": 127080
    },
    {
      "entropy": 1.9465318858623504,
      "epoch": 0.39396788025657975,
      "grad_norm": 7.537118911743164,
      "learning_rate": 4.030546911514647e-06,
      "loss": 0.4564,
      "mean_token_accuracy": 0.851013197004795,
      "num_tokens": 152817242.0,
      "step": 127090
    },
    {
      "entropy": 1.861098751425743,
      "epoch": 0.3939988793816295,
      "grad_norm": 8.616077423095703,
      "learning_rate": 4.030388349047818e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8491953507065773,
      "num_tokens": 152830104.0,
      "step": 127100
    },
    {
      "entropy": 1.9238109394907952,
      "epoch": 0.39402987850667914,
      "grad_norm": 8.807307243347168,
      "learning_rate": 4.0302298052931475e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8525387272238731,
      "num_tokens": 152841452.0,
      "step": 127110
    },
    {
      "entropy": 1.9179314807057382,
      "epoch": 0.39406087763172887,
      "grad_norm": 10.16199779510498,
      "learning_rate": 4.030071280246956e-06,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8477940663695336,
      "num_tokens": 152852810.0,
      "step": 127120
    },
    {
      "entropy": 1.913467188179493,
      "epoch": 0.39409187675677854,
      "grad_norm": 3.260228395462036,
      "learning_rate": 4.029912773905563e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.848815667629242,
      "num_tokens": 152864554.0,
      "step": 127130
    },
    {
      "entropy": 1.8686377540230752,
      "epoch": 0.39412287588182826,
      "grad_norm": 9.483077049255371,
      "learning_rate": 4.029754286265293e-06,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8370977133512497,
      "num_tokens": 152877426.0,
      "step": 127140
    },
    {
      "entropy": 1.929710279405117,
      "epoch": 0.39415387500687793,
      "grad_norm": 2.918980360031128,
      "learning_rate": 4.029595817322465e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.8524357438087463,
      "num_tokens": 152889339.0,
      "step": 127150
    },
    {
      "entropy": 1.879020507633686,
      "epoch": 0.39418487413192765,
      "grad_norm": 7.318241596221924,
      "learning_rate": 4.029437367073407e-06,
      "loss": 0.3923,
      "mean_token_accuracy": 0.868370558321476,
      "num_tokens": 152902178.0,
      "step": 127160
    },
    {
      "entropy": 1.8702649146318435,
      "epoch": 0.3942158732569773,
      "grad_norm": 8.086983680725098,
      "learning_rate": 4.029278935514442e-06,
      "loss": 0.4811,
      "mean_token_accuracy": 0.847602641582489,
      "num_tokens": 152914178.0,
      "step": 127170
    },
    {
      "entropy": 1.8651431202888489,
      "epoch": 0.394246872382027,
      "grad_norm": 8.561652183532715,
      "learning_rate": 4.029120522641896e-06,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8519322633743286,
      "num_tokens": 152926886.0,
      "step": 127180
    },
    {
      "entropy": 1.8370242148637772,
      "epoch": 0.3942778715070767,
      "grad_norm": 4.489937782287598,
      "learning_rate": 4.028962128452095e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.8447151646018028,
      "num_tokens": 152940001.0,
      "step": 127190
    },
    {
      "entropy": 1.806211344152689,
      "epoch": 0.3943088706321264,
      "grad_norm": 7.989223957061768,
      "learning_rate": 4.02880375294137e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8558181077241898,
      "num_tokens": 152953714.0,
      "step": 127200
    },
    {
      "entropy": 1.8846919029951095,
      "epoch": 0.3943398697571761,
      "grad_norm": 7.853222846984863,
      "learning_rate": 4.0286453961060475e-06,
      "loss": 0.4158,
      "mean_token_accuracy": 0.8626481354236603,
      "num_tokens": 152965285.0,
      "step": 127210
    },
    {
      "entropy": 1.8866841107606889,
      "epoch": 0.3943708688822258,
      "grad_norm": 3.7979695796966553,
      "learning_rate": 4.0284870579424576e-06,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8536775171756744,
      "num_tokens": 152978547.0,
      "step": 127220
    },
    {
      "entropy": 1.9279194965958595,
      "epoch": 0.3944018680072755,
      "grad_norm": 8.925294876098633,
      "learning_rate": 4.028328738446932e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8454255104064942,
      "num_tokens": 152990780.0,
      "step": 127230
    },
    {
      "entropy": 1.8884064748883247,
      "epoch": 0.39443286713232517,
      "grad_norm": 4.231663227081299,
      "learning_rate": 4.028170437615802e-06,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8594095915555954,
      "num_tokens": 153003179.0,
      "step": 127240
    },
    {
      "entropy": 1.9022858142852783,
      "epoch": 0.3944638662573749,
      "grad_norm": 4.5120038986206055,
      "learning_rate": 4.028012155445402e-06,
      "loss": 0.468,
      "mean_token_accuracy": 0.8501111850142479,
      "num_tokens": 153015801.0,
      "step": 127250
    },
    {
      "entropy": 1.9620060488581657,
      "epoch": 0.39449486538242456,
      "grad_norm": 7.684929370880127,
      "learning_rate": 4.027853891932065e-06,
      "loss": 0.4612,
      "mean_token_accuracy": 0.854415363073349,
      "num_tokens": 153027019.0,
      "step": 127260
    },
    {
      "entropy": 1.9046915173530579,
      "epoch": 0.3945258645074743,
      "grad_norm": 4.640239238739014,
      "learning_rate": 4.027695647072125e-06,
      "loss": 0.4763,
      "mean_token_accuracy": 0.8494089663028717,
      "num_tokens": 153039194.0,
      "step": 127270
    },
    {
      "entropy": 1.9410059854388237,
      "epoch": 0.39455686363252396,
      "grad_norm": 9.024331092834473,
      "learning_rate": 4.027537420861921e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.846622771024704,
      "num_tokens": 153050529.0,
      "step": 127280
    },
    {
      "entropy": 1.84144167304039,
      "epoch": 0.3945878627575737,
      "grad_norm": 10.240126609802246,
      "learning_rate": 4.027379213297787e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8550687476992607,
      "num_tokens": 153063542.0,
      "step": 127290
    },
    {
      "entropy": 1.8818937510251998,
      "epoch": 0.39461886188262335,
      "grad_norm": 4.128033638000488,
      "learning_rate": 4.027221024376063e-06,
      "loss": 0.4108,
      "mean_token_accuracy": 0.8581587508320808,
      "num_tokens": 153076371.0,
      "step": 127300
    },
    {
      "entropy": 1.9620690792798996,
      "epoch": 0.3946498610076731,
      "grad_norm": 8.45438289642334,
      "learning_rate": 4.027062854093087e-06,
      "loss": 0.5293,
      "mean_token_accuracy": 0.8367072865366936,
      "num_tokens": 153086758.0,
      "step": 127310
    },
    {
      "entropy": 1.879226452112198,
      "epoch": 0.39468086013272274,
      "grad_norm": 8.631587982177734,
      "learning_rate": 4.0269047024452e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8429103165864944,
      "num_tokens": 153099124.0,
      "step": 127320
    },
    {
      "entropy": 1.911029715836048,
      "epoch": 0.39471185925777247,
      "grad_norm": 6.511907577514648,
      "learning_rate": 4.0267465694287424e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8632428720593452,
      "num_tokens": 153110564.0,
      "step": 127330
    },
    {
      "entropy": 1.8814439699053764,
      "epoch": 0.39474285838282214,
      "grad_norm": 8.013952255249023,
      "learning_rate": 4.026588455040057e-06,
      "loss": 0.4122,
      "mean_token_accuracy": 0.8586022913455963,
      "num_tokens": 153123090.0,
      "step": 127340
    },
    {
      "entropy": 1.9142462641000748,
      "epoch": 0.39477385750787186,
      "grad_norm": 6.910181045532227,
      "learning_rate": 4.0264303592754846e-06,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8546132728457451,
      "num_tokens": 153134532.0,
      "step": 127350
    },
    {
      "entropy": 1.8884952366352081,
      "epoch": 0.39480485663292153,
      "grad_norm": 11.307275772094727,
      "learning_rate": 4.026272282131373e-06,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8564356371760369,
      "num_tokens": 153146234.0,
      "step": 127360
    },
    {
      "entropy": 1.845706556737423,
      "epoch": 0.39483585575797125,
      "grad_norm": 6.627220630645752,
      "learning_rate": 4.026114223604065e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8640777125954628,
      "num_tokens": 153159242.0,
      "step": 127370
    },
    {
      "entropy": 1.8887208893895149,
      "epoch": 0.3948668548830209,
      "grad_norm": 10.363401412963867,
      "learning_rate": 4.025956183689907e-06,
      "loss": 0.4095,
      "mean_token_accuracy": 0.8451639533042907,
      "num_tokens": 153172510.0,
      "step": 127380
    },
    {
      "entropy": 1.9080938518047332,
      "epoch": 0.39489785400807065,
      "grad_norm": 7.9390482902526855,
      "learning_rate": 4.025798162385246e-06,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8457752048969269,
      "num_tokens": 153184696.0,
      "step": 127390
    },
    {
      "entropy": 1.9023298189043998,
      "epoch": 0.3949288531331203,
      "grad_norm": 8.671669006347656,
      "learning_rate": 4.02564015968643e-06,
      "loss": 0.4492,
      "mean_token_accuracy": 0.851709508895874,
      "num_tokens": 153196489.0,
      "step": 127400
    },
    {
      "entropy": 1.8737532109022141,
      "epoch": 0.39495985225817,
      "grad_norm": 6.4061198234558105,
      "learning_rate": 4.025482175589809e-06,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8673104450106621,
      "num_tokens": 153207793.0,
      "step": 127410
    },
    {
      "entropy": 1.8804204553365707,
      "epoch": 0.3949908513832197,
      "grad_norm": 7.620257377624512,
      "learning_rate": 4.025324210091733e-06,
      "loss": 0.4195,
      "mean_token_accuracy": 0.8681916758418083,
      "num_tokens": 153219491.0,
      "step": 127420
    },
    {
      "entropy": 1.803827053308487,
      "epoch": 0.3950218505082694,
      "grad_norm": 4.123118877410889,
      "learning_rate": 4.0251662631885505e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8462093636393547,
      "num_tokens": 153232842.0,
      "step": 127430
    },
    {
      "entropy": 1.870395976305008,
      "epoch": 0.3950528496333191,
      "grad_norm": 8.609538078308105,
      "learning_rate": 4.025008334876618e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.8614588662981987,
      "num_tokens": 153244929.0,
      "step": 127440
    },
    {
      "entropy": 1.83759603202343,
      "epoch": 0.39508384875836877,
      "grad_norm": 8.134893417358398,
      "learning_rate": 4.024850425152286e-06,
      "loss": 0.3945,
      "mean_token_accuracy": 0.8604315787553787,
      "num_tokens": 153258122.0,
      "step": 127450
    },
    {
      "entropy": 1.919068345427513,
      "epoch": 0.3951148478834185,
      "grad_norm": 7.001434326171875,
      "learning_rate": 4.024692534011908e-06,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8476987138390542,
      "num_tokens": 153269204.0,
      "step": 127460
    },
    {
      "entropy": 1.839623036980629,
      "epoch": 0.39514584700846817,
      "grad_norm": 8.048707962036133,
      "learning_rate": 4.024534661451841e-06,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8495053514838219,
      "num_tokens": 153282189.0,
      "step": 127470
    },
    {
      "entropy": 1.9202652007341385,
      "epoch": 0.3951768461335179,
      "grad_norm": 8.451288223266602,
      "learning_rate": 4.02437680746844e-06,
      "loss": 0.4258,
      "mean_token_accuracy": 0.857101121544838,
      "num_tokens": 153293971.0,
      "step": 127480
    },
    {
      "entropy": 1.8971919193863869,
      "epoch": 0.39520784525856756,
      "grad_norm": 7.40570592880249,
      "learning_rate": 4.024218972058062e-06,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8695945486426353,
      "num_tokens": 153305994.0,
      "step": 127490
    },
    {
      "entropy": 1.8754101365804672,
      "epoch": 0.3952388443836173,
      "grad_norm": 7.888429641723633,
      "learning_rate": 4.0240611552170665e-06,
      "loss": 0.4087,
      "mean_token_accuracy": 0.8595315098762513,
      "num_tokens": 153319098.0,
      "step": 127500
    },
    {
      "entropy": 1.86890929043293,
      "epoch": 0.39526984350866695,
      "grad_norm": 8.941306114196777,
      "learning_rate": 4.023903356941811e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8576035022735595,
      "num_tokens": 153331242.0,
      "step": 127510
    },
    {
      "entropy": 1.8997353851795196,
      "epoch": 0.3953008426337167,
      "grad_norm": 7.998995780944824,
      "learning_rate": 4.023745577228658e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8570963263511657,
      "num_tokens": 153344304.0,
      "step": 127520
    },
    {
      "entropy": 1.820690706372261,
      "epoch": 0.39533184175876634,
      "grad_norm": 7.785648822784424,
      "learning_rate": 4.023587816073965e-06,
      "loss": 0.4075,
      "mean_token_accuracy": 0.8614529862999916,
      "num_tokens": 153358390.0,
      "step": 127530
    },
    {
      "entropy": 1.8702727839350701,
      "epoch": 0.39536284088381607,
      "grad_norm": 4.13304328918457,
      "learning_rate": 4.023430073474097e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8575648859143257,
      "num_tokens": 153370126.0,
      "step": 127540
    },
    {
      "entropy": 1.81645250543952,
      "epoch": 0.39539384000886574,
      "grad_norm": 2.846233606338501,
      "learning_rate": 4.023272349425415e-06,
      "loss": 0.3549,
      "mean_token_accuracy": 0.8668262645602226,
      "num_tokens": 153383650.0,
      "step": 127550
    },
    {
      "entropy": 1.899166676402092,
      "epoch": 0.39542483913391546,
      "grad_norm": 8.356348037719727,
      "learning_rate": 4.023114643924286e-06,
      "loss": 0.4868,
      "mean_token_accuracy": 0.8431504383683205,
      "num_tokens": 153395416.0,
      "step": 127560
    },
    {
      "entropy": 1.8529985576868058,
      "epoch": 0.39545583825896513,
      "grad_norm": 2.3607380390167236,
      "learning_rate": 4.022956956967072e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8648950770497322,
      "num_tokens": 153407938.0,
      "step": 127570
    },
    {
      "entropy": 1.8317057311534881,
      "epoch": 0.39548683738401486,
      "grad_norm": 8.973584175109863,
      "learning_rate": 4.022799288550142e-06,
      "loss": 0.4097,
      "mean_token_accuracy": 0.8503635957837105,
      "num_tokens": 153420425.0,
      "step": 127580
    },
    {
      "entropy": 1.967195200920105,
      "epoch": 0.3955178365090645,
      "grad_norm": 8.250533103942871,
      "learning_rate": 4.022641638669861e-06,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8539433136582375,
      "num_tokens": 153431948.0,
      "step": 127590
    },
    {
      "entropy": 1.8443527117371559,
      "epoch": 0.39554883563411425,
      "grad_norm": 8.231595039367676,
      "learning_rate": 4.022484007322598e-06,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8479686647653579,
      "num_tokens": 153445161.0,
      "step": 127600
    },
    {
      "entropy": 1.9092737153172492,
      "epoch": 0.3955798347591639,
      "grad_norm": 8.519765853881836,
      "learning_rate": 4.022326394504722e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8497352659702301,
      "num_tokens": 153456782.0,
      "step": 127610
    },
    {
      "entropy": 1.9170891061425208,
      "epoch": 0.39561083388421364,
      "grad_norm": 7.752858638763428,
      "learning_rate": 4.0221688002126016e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.8510538384318351,
      "num_tokens": 153468381.0,
      "step": 127620
    },
    {
      "entropy": 1.8833612218499183,
      "epoch": 0.3956418330092633,
      "grad_norm": 4.76165246963501,
      "learning_rate": 4.022011224442611e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8437580853700638,
      "num_tokens": 153481424.0,
      "step": 127630
    },
    {
      "entropy": 1.9320109218358994,
      "epoch": 0.39567283213431304,
      "grad_norm": 4.280218601226807,
      "learning_rate": 4.02185366719112e-06,
      "loss": 0.4584,
      "mean_token_accuracy": 0.848100745677948,
      "num_tokens": 153493443.0,
      "step": 127640
    },
    {
      "entropy": 1.960513624548912,
      "epoch": 0.3957038312593627,
      "grad_norm": 8.801411628723145,
      "learning_rate": 4.021696128454502e-06,
      "loss": 0.503,
      "mean_token_accuracy": 0.8425059333443642,
      "num_tokens": 153504656.0,
      "step": 127650
    },
    {
      "entropy": 1.9096018448472023,
      "epoch": 0.3957348303844124,
      "grad_norm": 3.8925299644470215,
      "learning_rate": 4.0215386082291315e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8456997722387314,
      "num_tokens": 153516555.0,
      "step": 127660
    },
    {
      "entropy": 1.9139437288045884,
      "epoch": 0.3957658295094621,
      "grad_norm": 7.8598127365112305,
      "learning_rate": 4.021381106511384e-06,
      "loss": 0.5097,
      "mean_token_accuracy": 0.8359185487031937,
      "num_tokens": 153528681.0,
      "step": 127670
    },
    {
      "entropy": 1.925663386285305,
      "epoch": 0.39579682863451177,
      "grad_norm": 7.7925944328308105,
      "learning_rate": 4.021223623297635e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8525381490588189,
      "num_tokens": 153540256.0,
      "step": 127680
    },
    {
      "entropy": 1.9570807069540024,
      "epoch": 0.3958278277595615,
      "grad_norm": 7.370650291442871,
      "learning_rate": 4.021066158584261e-06,
      "loss": 0.5033,
      "mean_token_accuracy": 0.848370935022831,
      "num_tokens": 153551273.0,
      "step": 127690
    },
    {
      "entropy": 1.8661099091172217,
      "epoch": 0.39585882688461116,
      "grad_norm": 4.928553581237793,
      "learning_rate": 4.020908712367641e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8484266191720963,
      "num_tokens": 153563180.0,
      "step": 127700
    },
    {
      "entropy": 1.9060293585062027,
      "epoch": 0.3958898260096609,
      "grad_norm": 9.447659492492676,
      "learning_rate": 4.020751284644153e-06,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8432995095849037,
      "num_tokens": 153575032.0,
      "step": 127710
    },
    {
      "entropy": 1.953291055560112,
      "epoch": 0.39592082513471055,
      "grad_norm": 7.042132377624512,
      "learning_rate": 4.020593875410178e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8566155821084976,
      "num_tokens": 153586028.0,
      "step": 127720
    },
    {
      "entropy": 1.9117076322436333,
      "epoch": 0.3959518242597603,
      "grad_norm": 8.872218132019043,
      "learning_rate": 4.020436484662098e-06,
      "loss": 0.4485,
      "mean_token_accuracy": 0.8476521626114846,
      "num_tokens": 153599040.0,
      "step": 127730
    },
    {
      "entropy": 1.8983178213238716,
      "epoch": 0.39598282338480995,
      "grad_norm": 8.394247055053711,
      "learning_rate": 4.020279112396293e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8589667662978172,
      "num_tokens": 153610777.0,
      "step": 127740
    },
    {
      "entropy": 1.9231909066438675,
      "epoch": 0.39601382250985967,
      "grad_norm": 10.013471603393555,
      "learning_rate": 4.0201217586091465e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8497250765562058,
      "num_tokens": 153621814.0,
      "step": 127750
    },
    {
      "entropy": 1.900232744216919,
      "epoch": 0.39604482163490934,
      "grad_norm": 9.529150009155273,
      "learning_rate": 4.019964423297043e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.842905105650425,
      "num_tokens": 153634211.0,
      "step": 127760
    },
    {
      "entropy": 1.8112396225333214,
      "epoch": 0.39607582075995906,
      "grad_norm": 9.09139347076416,
      "learning_rate": 4.019807106456367e-06,
      "loss": 0.3551,
      "mean_token_accuracy": 0.8626585349440574,
      "num_tokens": 153647738.0,
      "step": 127770
    },
    {
      "entropy": 1.8331888422369957,
      "epoch": 0.39610681988500873,
      "grad_norm": 6.125622272491455,
      "learning_rate": 4.0196498080835054e-06,
      "loss": 0.3867,
      "mean_token_accuracy": 0.8652116924524307,
      "num_tokens": 153661042.0,
      "step": 127780
    },
    {
      "entropy": 1.9142784118652343,
      "epoch": 0.39613781901005846,
      "grad_norm": 4.094144821166992,
      "learning_rate": 4.019492528174845e-06,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8501773819327354,
      "num_tokens": 153673112.0,
      "step": 127790
    },
    {
      "entropy": 1.8980029091238975,
      "epoch": 0.3961688181351081,
      "grad_norm": 3.3827157020568848,
      "learning_rate": 4.0193352667267725e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8512454509735108,
      "num_tokens": 153685077.0,
      "step": 127800
    },
    {
      "entropy": 1.716671919822693,
      "epoch": 0.39619981726015785,
      "grad_norm": 3.326303482055664,
      "learning_rate": 4.019178023735678e-06,
      "loss": 0.2917,
      "mean_token_accuracy": 0.8814573958516121,
      "num_tokens": 153699159.0,
      "step": 127810
    },
    {
      "entropy": 1.7483276754617691,
      "epoch": 0.3962308163852075,
      "grad_norm": 8.083634376525879,
      "learning_rate": 4.019020799197951e-06,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8582610175013542,
      "num_tokens": 153713542.0,
      "step": 127820
    },
    {
      "entropy": 1.915240579843521,
      "epoch": 0.39626181551025724,
      "grad_norm": 10.262462615966797,
      "learning_rate": 4.018863593109982e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8524910137057304,
      "num_tokens": 153725218.0,
      "step": 127830
    },
    {
      "entropy": 1.8652495056390763,
      "epoch": 0.3962928146353069,
      "grad_norm": 9.742287635803223,
      "learning_rate": 4.018706405468165e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8553525045514107,
      "num_tokens": 153737029.0,
      "step": 127840
    },
    {
      "entropy": 1.9075027361512185,
      "epoch": 0.39632381376035664,
      "grad_norm": 9.754807472229004,
      "learning_rate": 4.018549236268891e-06,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8644733369350434,
      "num_tokens": 153748555.0,
      "step": 127850
    },
    {
      "entropy": 1.8555178031325341,
      "epoch": 0.3963548128854063,
      "grad_norm": 10.997365951538086,
      "learning_rate": 4.018392085508554e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8505488753318786,
      "num_tokens": 153761293.0,
      "step": 127860
    },
    {
      "entropy": 1.8854364529252052,
      "epoch": 0.39638581201045603,
      "grad_norm": 10.64345932006836,
      "learning_rate": 4.018234953183549e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8452537760138512,
      "num_tokens": 153773673.0,
      "step": 127870
    },
    {
      "entropy": 1.9308582305908204,
      "epoch": 0.3964168111355057,
      "grad_norm": 8.477107048034668,
      "learning_rate": 4.018077839290272e-06,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8420649126172066,
      "num_tokens": 153785000.0,
      "step": 127880
    },
    {
      "entropy": 1.8462091460824013,
      "epoch": 0.3964478102605554,
      "grad_norm": 4.125033855438232,
      "learning_rate": 4.01792074382512e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8481978312134743,
      "num_tokens": 153797926.0,
      "step": 127890
    },
    {
      "entropy": 1.9186622604727746,
      "epoch": 0.3964788093856051,
      "grad_norm": 8.226593971252441,
      "learning_rate": 4.0177636667844914e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8365380227565765,
      "num_tokens": 153809721.0,
      "step": 127900
    },
    {
      "entropy": 1.8989548355340957,
      "epoch": 0.39650980851065476,
      "grad_norm": 7.229928970336914,
      "learning_rate": 4.017606608164784e-06,
      "loss": 0.4022,
      "mean_token_accuracy": 0.8661728858947754,
      "num_tokens": 153821479.0,
      "step": 127910
    },
    {
      "entropy": 1.875709056854248,
      "epoch": 0.3965408076357045,
      "grad_norm": 5.093418598175049,
      "learning_rate": 4.017449567962398e-06,
      "loss": 0.4293,
      "mean_token_accuracy": 0.8482154250144959,
      "num_tokens": 153833724.0,
      "step": 127920
    },
    {
      "entropy": 1.9880733832716941,
      "epoch": 0.39657180676075415,
      "grad_norm": 8.915475845336914,
      "learning_rate": 4.0172925461737336e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8514209687709808,
      "num_tokens": 153844594.0,
      "step": 127930
    },
    {
      "entropy": 1.8570735067129136,
      "epoch": 0.3966028058858039,
      "grad_norm": 4.229099273681641,
      "learning_rate": 4.017135542795195e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8634174600243568,
      "num_tokens": 153856932.0,
      "step": 127940
    },
    {
      "entropy": 1.9012201085686684,
      "epoch": 0.39663380501085355,
      "grad_norm": 11.16518783569336,
      "learning_rate": 4.016978557823181e-06,
      "loss": 0.4162,
      "mean_token_accuracy": 0.8594008833169937,
      "num_tokens": 153868678.0,
      "step": 127950
    },
    {
      "entropy": 1.7977351397275925,
      "epoch": 0.39666480413590327,
      "grad_norm": 7.758661270141602,
      "learning_rate": 4.016821591254099e-06,
      "loss": 0.3605,
      "mean_token_accuracy": 0.8647834539413453,
      "num_tokens": 153882380.0,
      "step": 127960
    },
    {
      "entropy": 1.86228808760643,
      "epoch": 0.39669580326095294,
      "grad_norm": 4.080726623535156,
      "learning_rate": 4.016664643084352e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8644084423780442,
      "num_tokens": 153894569.0,
      "step": 127970
    },
    {
      "entropy": 1.9319358631968497,
      "epoch": 0.39672680238600266,
      "grad_norm": 7.336208820343018,
      "learning_rate": 4.016507713310346e-06,
      "loss": 0.474,
      "mean_token_accuracy": 0.8546565786004067,
      "num_tokens": 153905410.0,
      "step": 127980
    },
    {
      "entropy": 1.8738821625709534,
      "epoch": 0.39675780151105233,
      "grad_norm": 8.000831604003906,
      "learning_rate": 4.016350801928489e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8442562118172645,
      "num_tokens": 153918966.0,
      "step": 127990
    },
    {
      "entropy": 1.8228215545415878,
      "epoch": 0.39678880063610206,
      "grad_norm": 6.075873374938965,
      "learning_rate": 4.0161939089351855e-06,
      "loss": 0.4462,
      "mean_token_accuracy": 0.852639339864254,
      "num_tokens": 153931674.0,
      "step": 128000
    },
    {
      "entropy": 1.8811346724629403,
      "epoch": 0.3968197997611517,
      "grad_norm": 8.297804832458496,
      "learning_rate": 4.016037034326846e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8590738877654076,
      "num_tokens": 153943910.0,
      "step": 128010
    },
    {
      "entropy": 1.9253056347370148,
      "epoch": 0.39685079888620145,
      "grad_norm": 9.22485637664795,
      "learning_rate": 4.015880178099881e-06,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8511346384882927,
      "num_tokens": 153955547.0,
      "step": 128020
    },
    {
      "entropy": 1.814525055885315,
      "epoch": 0.3968817980112511,
      "grad_norm": 7.030633449554443,
      "learning_rate": 4.0157233402507e-06,
      "loss": 0.3723,
      "mean_token_accuracy": 0.8703080207109452,
      "num_tokens": 153968920.0,
      "step": 128030
    },
    {
      "entropy": 1.9247617200016975,
      "epoch": 0.39691279713630084,
      "grad_norm": 9.265406608581543,
      "learning_rate": 4.015566520775715e-06,
      "loss": 0.4815,
      "mean_token_accuracy": 0.849250017106533,
      "num_tokens": 153980346.0,
      "step": 128040
    },
    {
      "entropy": 1.9104125529527665,
      "epoch": 0.3969437962613505,
      "grad_norm": 3.5930426120758057,
      "learning_rate": 4.015409719671339e-06,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8543394804000854,
      "num_tokens": 153992553.0,
      "step": 128050
    },
    {
      "entropy": 1.866149678826332,
      "epoch": 0.39697479538640024,
      "grad_norm": 6.8628926277160645,
      "learning_rate": 4.015252936933985e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8579761996865273,
      "num_tokens": 154004883.0,
      "step": 128060
    },
    {
      "entropy": 1.9460046872496606,
      "epoch": 0.3970057945114499,
      "grad_norm": 9.495827674865723,
      "learning_rate": 4.015096172560067e-06,
      "loss": 0.4824,
      "mean_token_accuracy": 0.8454651072621345,
      "num_tokens": 154016816.0,
      "step": 128070
    },
    {
      "entropy": 1.8353903412818908,
      "epoch": 0.39703679363649963,
      "grad_norm": 8.298847198486328,
      "learning_rate": 4.014939426546002e-06,
      "loss": 0.4392,
      "mean_token_accuracy": 0.8565979763865471,
      "num_tokens": 154029332.0,
      "step": 128080
    },
    {
      "entropy": 1.8864996194839478,
      "epoch": 0.3970677927615493,
      "grad_norm": 8.607029914855957,
      "learning_rate": 4.014782698888205e-06,
      "loss": 0.4839,
      "mean_token_accuracy": 0.842142577469349,
      "num_tokens": 154040881.0,
      "step": 128090
    },
    {
      "entropy": 1.8832897543907166,
      "epoch": 0.397098791886599,
      "grad_norm": 3.863781452178955,
      "learning_rate": 4.014625989583094e-06,
      "loss": 0.4027,
      "mean_token_accuracy": 0.8512258976697922,
      "num_tokens": 154052885.0,
      "step": 128100
    },
    {
      "entropy": 1.8848641395568848,
      "epoch": 0.3971297910116487,
      "grad_norm": 7.123293876647949,
      "learning_rate": 4.014469298627088e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.856060591340065,
      "num_tokens": 154064134.0,
      "step": 128110
    },
    {
      "entropy": 1.8236127287149428,
      "epoch": 0.3971607901366984,
      "grad_norm": 5.063018321990967,
      "learning_rate": 4.0143126260166075e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8463837578892708,
      "num_tokens": 154077307.0,
      "step": 128120
    },
    {
      "entropy": 1.9382410168647766,
      "epoch": 0.3971917892617481,
      "grad_norm": 8.761957168579102,
      "learning_rate": 4.014155971748069e-06,
      "loss": 0.5266,
      "mean_token_accuracy": 0.8408651709556579,
      "num_tokens": 154088700.0,
      "step": 128130
    },
    {
      "entropy": 1.888756561279297,
      "epoch": 0.3972227883867978,
      "grad_norm": 9.47224235534668,
      "learning_rate": 4.013999335817898e-06,
      "loss": 0.4318,
      "mean_token_accuracy": 0.8587385430932045,
      "num_tokens": 154100614.0,
      "step": 128140
    },
    {
      "entropy": 1.9622124269604684,
      "epoch": 0.3972537875118475,
      "grad_norm": 8.621251106262207,
      "learning_rate": 4.013842718222516e-06,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8376434296369553,
      "num_tokens": 154112125.0,
      "step": 128150
    },
    {
      "entropy": 1.9215806141495704,
      "epoch": 0.39728478663689715,
      "grad_norm": 8.473328590393066,
      "learning_rate": 4.013686118958345e-06,
      "loss": 0.4631,
      "mean_token_accuracy": 0.852269695699215,
      "num_tokens": 154123987.0,
      "step": 128160
    },
    {
      "entropy": 1.8634486734867095,
      "epoch": 0.39731578576194687,
      "grad_norm": 8.564555168151855,
      "learning_rate": 4.013529538021809e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8536643877625465,
      "num_tokens": 154137001.0,
      "step": 128170
    },
    {
      "entropy": 1.9504847824573517,
      "epoch": 0.39734678488699654,
      "grad_norm": 8.242449760437012,
      "learning_rate": 4.013372975409336e-06,
      "loss": 0.4803,
      "mean_token_accuracy": 0.8405697852373123,
      "num_tokens": 154148533.0,
      "step": 128180
    },
    {
      "entropy": 1.8996037617325783,
      "epoch": 0.39737778401204626,
      "grad_norm": 10.100019454956055,
      "learning_rate": 4.01321643111735e-06,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8533552616834641,
      "num_tokens": 154160553.0,
      "step": 128190
    },
    {
      "entropy": 1.8098711207509042,
      "epoch": 0.39740878313709593,
      "grad_norm": 9.241110801696777,
      "learning_rate": 4.013059905142279e-06,
      "loss": 0.3669,
      "mean_token_accuracy": 0.871603924036026,
      "num_tokens": 154174245.0,
      "step": 128200
    },
    {
      "entropy": 1.9533662751317025,
      "epoch": 0.39743978226214566,
      "grad_norm": 9.006794929504395,
      "learning_rate": 4.012903397480549e-06,
      "loss": 0.4952,
      "mean_token_accuracy": 0.8515314489603043,
      "num_tokens": 154185508.0,
      "step": 128210
    },
    {
      "entropy": 1.847785583138466,
      "epoch": 0.3974707813871953,
      "grad_norm": 4.733590126037598,
      "learning_rate": 4.012746908128595e-06,
      "loss": 0.4237,
      "mean_token_accuracy": 0.8513537049293518,
      "num_tokens": 154198114.0,
      "step": 128220
    },
    {
      "entropy": 1.908326794207096,
      "epoch": 0.39750178051224505,
      "grad_norm": 7.348505973815918,
      "learning_rate": 4.012590437082841e-06,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8644715324044228,
      "num_tokens": 154210418.0,
      "step": 128230
    },
    {
      "entropy": 1.9537509649991989,
      "epoch": 0.3975327796372947,
      "grad_norm": 7.655646800994873,
      "learning_rate": 4.0124339843397216e-06,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8567704558372498,
      "num_tokens": 154221854.0,
      "step": 128240
    },
    {
      "entropy": 1.8390052139759063,
      "epoch": 0.39756377876234444,
      "grad_norm": 8.591985702514648,
      "learning_rate": 4.012277549895668e-06,
      "loss": 0.4074,
      "mean_token_accuracy": 0.859342285990715,
      "num_tokens": 154234898.0,
      "step": 128250
    },
    {
      "entropy": 1.8981177046895028,
      "epoch": 0.3975947778873941,
      "grad_norm": 3.5473618507385254,
      "learning_rate": 4.012121133747113e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8572326749563217,
      "num_tokens": 154247303.0,
      "step": 128260
    },
    {
      "entropy": 1.9318851605057716,
      "epoch": 0.39762577701244384,
      "grad_norm": 6.904049873352051,
      "learning_rate": 4.011964735890492e-06,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8485365882515907,
      "num_tokens": 154259127.0,
      "step": 128270
    },
    {
      "entropy": 1.856573697924614,
      "epoch": 0.3976567761374935,
      "grad_norm": 7.510590553283691,
      "learning_rate": 4.011808356322238e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8440538689494133,
      "num_tokens": 154271893.0,
      "step": 128280
    },
    {
      "entropy": 1.9165368214249612,
      "epoch": 0.39768777526254323,
      "grad_norm": 3.662731885910034,
      "learning_rate": 4.011651995038789e-06,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8502262338995934,
      "num_tokens": 154283424.0,
      "step": 128290
    },
    {
      "entropy": 1.9400472521781922,
      "epoch": 0.3977187743875929,
      "grad_norm": 8.449536323547363,
      "learning_rate": 4.011495652036581e-06,
      "loss": 0.4438,
      "mean_token_accuracy": 0.864606074988842,
      "num_tokens": 154294507.0,
      "step": 128300
    },
    {
      "entropy": 1.9318364679813385,
      "epoch": 0.3977497735126426,
      "grad_norm": 7.544651031494141,
      "learning_rate": 4.011339327312052e-06,
      "loss": 0.4951,
      "mean_token_accuracy": 0.851364016532898,
      "num_tokens": 154306187.0,
      "step": 128310
    },
    {
      "entropy": 1.859896543622017,
      "epoch": 0.3977807726376923,
      "grad_norm": 7.740284442901611,
      "learning_rate": 4.0111830208616405e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8656293138861656,
      "num_tokens": 154318477.0,
      "step": 128320
    },
    {
      "entropy": 1.9752064496278763,
      "epoch": 0.397811771762742,
      "grad_norm": 7.771737575531006,
      "learning_rate": 4.011026732681787e-06,
      "loss": 0.47,
      "mean_token_accuracy": 0.8587437257170677,
      "num_tokens": 154329265.0,
      "step": 128330
    },
    {
      "entropy": 1.826411110162735,
      "epoch": 0.3978427708877917,
      "grad_norm": 9.697999000549316,
      "learning_rate": 4.010870462768933e-06,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8450366973876953,
      "num_tokens": 154343157.0,
      "step": 128340
    },
    {
      "entropy": 1.8539129607379436,
      "epoch": 0.3978737700128414,
      "grad_norm": 9.17566967010498,
      "learning_rate": 4.010714211119519e-06,
      "loss": 0.4242,
      "mean_token_accuracy": 0.8553266122937202,
      "num_tokens": 154354274.0,
      "step": 128350
    },
    {
      "entropy": 1.8710831806063652,
      "epoch": 0.3979047691378911,
      "grad_norm": 6.939380168914795,
      "learning_rate": 4.010557977729989e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8506567060947419,
      "num_tokens": 154367463.0,
      "step": 128360
    },
    {
      "entropy": 1.9934969753026963,
      "epoch": 0.3979357682629408,
      "grad_norm": 9.315759658813477,
      "learning_rate": 4.010401762596786e-06,
      "loss": 0.5463,
      "mean_token_accuracy": 0.8378830254077911,
      "num_tokens": 154378868.0,
      "step": 128370
    },
    {
      "entropy": 1.9435987308621407,
      "epoch": 0.39796676738799047,
      "grad_norm": 7.836657524108887,
      "learning_rate": 4.010245565716356e-06,
      "loss": 0.443,
      "mean_token_accuracy": 0.857276414334774,
      "num_tokens": 154390849.0,
      "step": 128380
    },
    {
      "entropy": 1.8966143906116486,
      "epoch": 0.3979977665130402,
      "grad_norm": 8.332956314086914,
      "learning_rate": 4.010089387085143e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8508517384529114,
      "num_tokens": 154402741.0,
      "step": 128390
    },
    {
      "entropy": 1.9793959528207778,
      "epoch": 0.39802876563808987,
      "grad_norm": 8.801210403442383,
      "learning_rate": 4.009933226699596e-06,
      "loss": 0.5463,
      "mean_token_accuracy": 0.8385202527046204,
      "num_tokens": 154413301.0,
      "step": 128400
    },
    {
      "entropy": 1.9293900340795518,
      "epoch": 0.39805976476313953,
      "grad_norm": 10.354657173156738,
      "learning_rate": 4.00977708455616e-06,
      "loss": 0.5402,
      "mean_token_accuracy": 0.8320423439145088,
      "num_tokens": 154424899.0,
      "step": 128410
    },
    {
      "entropy": 1.7377336770296097,
      "epoch": 0.39809076388818926,
      "grad_norm": 7.612285137176514,
      "learning_rate": 4.009620960651285e-06,
      "loss": 0.3372,
      "mean_token_accuracy": 0.8693819224834443,
      "num_tokens": 154439450.0,
      "step": 128420
    },
    {
      "entropy": 1.9407954409718513,
      "epoch": 0.3981217630132389,
      "grad_norm": 7.916438579559326,
      "learning_rate": 4.00946485498142e-06,
      "loss": 0.4653,
      "mean_token_accuracy": 0.8480086639523506,
      "num_tokens": 154451790.0,
      "step": 128430
    },
    {
      "entropy": 1.9476406499743462,
      "epoch": 0.39815276213828865,
      "grad_norm": 4.425556659698486,
      "learning_rate": 4.009308767543017e-06,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8459201797842979,
      "num_tokens": 154463897.0,
      "step": 128440
    },
    {
      "entropy": 2.0008150786161423,
      "epoch": 0.3981837612633383,
      "grad_norm": 9.704833030700684,
      "learning_rate": 4.009152698332526e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8437413021922111,
      "num_tokens": 154475527.0,
      "step": 128450
    },
    {
      "entropy": 1.7799909293651581,
      "epoch": 0.39821476038838804,
      "grad_norm": 3.9981141090393066,
      "learning_rate": 4.0089966473464e-06,
      "loss": 0.3358,
      "mean_token_accuracy": 0.8721995532512665,
      "num_tokens": 154489848.0,
      "step": 128460
    },
    {
      "entropy": 1.9085913822054863,
      "epoch": 0.3982457595134377,
      "grad_norm": 5.100532054901123,
      "learning_rate": 4.008840614581093e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.856728957593441,
      "num_tokens": 154502460.0,
      "step": 128470
    },
    {
      "entropy": 1.9310724124312402,
      "epoch": 0.39827675863848744,
      "grad_norm": 7.409860610961914,
      "learning_rate": 4.008684600033059e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8475088179111481,
      "num_tokens": 154515158.0,
      "step": 128480
    },
    {
      "entropy": 1.9532846316695214,
      "epoch": 0.3983077577635371,
      "grad_norm": 6.9244561195373535,
      "learning_rate": 4.008528603698753e-06,
      "loss": 0.4868,
      "mean_token_accuracy": 0.8392842382192611,
      "num_tokens": 154526053.0,
      "step": 128490
    },
    {
      "entropy": 1.864257997274399,
      "epoch": 0.39833875688858683,
      "grad_norm": 8.621221542358398,
      "learning_rate": 4.008372625574633e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8733592018485069,
      "num_tokens": 154537962.0,
      "step": 128500
    },
    {
      "entropy": 1.7741185277700424,
      "epoch": 0.3983697560136365,
      "grad_norm": 12.221007347106934,
      "learning_rate": 4.008216665657155e-06,
      "loss": 0.36,
      "mean_token_accuracy": 0.8721621558070183,
      "num_tokens": 154551264.0,
      "step": 128510
    },
    {
      "entropy": 1.8366203412413598,
      "epoch": 0.3984007551386862,
      "grad_norm": 8.838532447814941,
      "learning_rate": 4.008060723942776e-06,
      "loss": 0.3882,
      "mean_token_accuracy": 0.8588550105690956,
      "num_tokens": 154563900.0,
      "step": 128520
    },
    {
      "entropy": 1.9761778950691222,
      "epoch": 0.3984317542637359,
      "grad_norm": 8.271575927734375,
      "learning_rate": 4.007904800427958e-06,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8369174391031265,
      "num_tokens": 154574439.0,
      "step": 128530
    },
    {
      "entropy": 1.871373300254345,
      "epoch": 0.3984627533887856,
      "grad_norm": 3.379486322402954,
      "learning_rate": 4.0077488951091595e-06,
      "loss": 0.4258,
      "mean_token_accuracy": 0.852562639117241,
      "num_tokens": 154586791.0,
      "step": 128540
    },
    {
      "entropy": 1.8863118886947632,
      "epoch": 0.3984937525138353,
      "grad_norm": 8.740182876586914,
      "learning_rate": 4.007593007982842e-06,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8477497771382332,
      "num_tokens": 154599491.0,
      "step": 128550
    },
    {
      "entropy": 1.8409544050693512,
      "epoch": 0.398524751638885,
      "grad_norm": 4.045310020446777,
      "learning_rate": 4.007437139045469e-06,
      "loss": 0.3935,
      "mean_token_accuracy": 0.8576883286237716,
      "num_tokens": 154612812.0,
      "step": 128560
    },
    {
      "entropy": 1.8716674730181695,
      "epoch": 0.3985557507639347,
      "grad_norm": 4.682071208953857,
      "learning_rate": 4.007281288293502e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8520221546292305,
      "num_tokens": 154625583.0,
      "step": 128570
    },
    {
      "entropy": 1.9001308396458625,
      "epoch": 0.3985867498889844,
      "grad_norm": 7.292641639709473,
      "learning_rate": 4.007125455723405e-06,
      "loss": 0.4524,
      "mean_token_accuracy": 0.8661713659763336,
      "num_tokens": 154637640.0,
      "step": 128580
    },
    {
      "entropy": 1.9417346104979516,
      "epoch": 0.3986177490140341,
      "grad_norm": 8.141678810119629,
      "learning_rate": 4.006969641331644e-06,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8521144181489945,
      "num_tokens": 154649418.0,
      "step": 128590
    },
    {
      "entropy": 1.9820822283625603,
      "epoch": 0.3986487481390838,
      "grad_norm": 10.370221138000488,
      "learning_rate": 4.0068138451146845e-06,
      "loss": 0.5383,
      "mean_token_accuracy": 0.8394621968269348,
      "num_tokens": 154660964.0,
      "step": 128600
    },
    {
      "entropy": 1.836016722023487,
      "epoch": 0.39867974726413347,
      "grad_norm": 3.998152017593384,
      "learning_rate": 4.006658067068994e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8594549700617791,
      "num_tokens": 154674134.0,
      "step": 128610
    },
    {
      "entropy": 1.9683793038129807,
      "epoch": 0.3987107463891832,
      "grad_norm": 8.68581771850586,
      "learning_rate": 4.0065023071910395e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.866059435904026,
      "num_tokens": 154685620.0,
      "step": 128620
    },
    {
      "entropy": 1.9310366541147232,
      "epoch": 0.39874174551423286,
      "grad_norm": 8.650096893310547,
      "learning_rate": 4.006346565477289e-06,
      "loss": 0.4354,
      "mean_token_accuracy": 0.8481193721294403,
      "num_tokens": 154697293.0,
      "step": 128630
    },
    {
      "entropy": 1.9691759124398232,
      "epoch": 0.3987727446392826,
      "grad_norm": 7.362710952758789,
      "learning_rate": 4.006190841924217e-06,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8400867089629174,
      "num_tokens": 154708876.0,
      "step": 128640
    },
    {
      "entropy": 1.8157849997282027,
      "epoch": 0.39880374376433225,
      "grad_norm": 3.911681652069092,
      "learning_rate": 4.006035136528288e-06,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8522502645850182,
      "num_tokens": 154721940.0,
      "step": 128650
    },
    {
      "entropy": 1.9943493276834487,
      "epoch": 0.3988347428893819,
      "grad_norm": 9.267487525939941,
      "learning_rate": 4.005879449285979e-06,
      "loss": 0.5359,
      "mean_token_accuracy": 0.8361817836761475,
      "num_tokens": 154732230.0,
      "step": 128660
    },
    {
      "entropy": 1.9468561723828315,
      "epoch": 0.39886574201443165,
      "grad_norm": 8.407017707824707,
      "learning_rate": 4.00572378019376e-06,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8474451348185539,
      "num_tokens": 154743873.0,
      "step": 128670
    },
    {
      "entropy": 1.883972604572773,
      "epoch": 0.3988967411394813,
      "grad_norm": 10.778926849365234,
      "learning_rate": 4.005568129248105e-06,
      "loss": 0.4577,
      "mean_token_accuracy": 0.8490250527858734,
      "num_tokens": 154756322.0,
      "step": 128680
    },
    {
      "entropy": 1.9082299306988717,
      "epoch": 0.39892774026453104,
      "grad_norm": 8.559072494506836,
      "learning_rate": 4.0054124964454904e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8479592993855476,
      "num_tokens": 154768822.0,
      "step": 128690
    },
    {
      "entropy": 1.915241825580597,
      "epoch": 0.3989587393895807,
      "grad_norm": 8.535916328430176,
      "learning_rate": 4.005256881782389e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.848818339407444,
      "num_tokens": 154780787.0,
      "step": 128700
    },
    {
      "entropy": 1.85912893563509,
      "epoch": 0.39898973851463043,
      "grad_norm": 8.055673599243164,
      "learning_rate": 4.005101285255279e-06,
      "loss": 0.4004,
      "mean_token_accuracy": 0.8605345159769058,
      "num_tokens": 154793208.0,
      "step": 128710
    },
    {
      "entropy": 1.9649493724107743,
      "epoch": 0.3990207376396801,
      "grad_norm": 8.454475402832031,
      "learning_rate": 4.004945706860638e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8544456735253334,
      "num_tokens": 154804930.0,
      "step": 128720
    },
    {
      "entropy": 1.923764231801033,
      "epoch": 0.3990517367647298,
      "grad_norm": 7.50667142868042,
      "learning_rate": 4.004790146594944e-06,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8489472419023514,
      "num_tokens": 154817425.0,
      "step": 128730
    },
    {
      "entropy": 1.8540160700678825,
      "epoch": 0.3990827358897795,
      "grad_norm": 4.772271633148193,
      "learning_rate": 4.004634604454677e-06,
      "loss": 0.4086,
      "mean_token_accuracy": 0.8635622560977936,
      "num_tokens": 154829558.0,
      "step": 128740
    },
    {
      "entropy": 1.8815680295228958,
      "epoch": 0.3991137350148292,
      "grad_norm": 8.889459609985352,
      "learning_rate": 4.004479080436317e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8565109595656395,
      "num_tokens": 154841865.0,
      "step": 128750
    },
    {
      "entropy": 1.8703197434544563,
      "epoch": 0.3991447341398789,
      "grad_norm": 3.177785634994507,
      "learning_rate": 4.004323574536345e-06,
      "loss": 0.4666,
      "mean_token_accuracy": 0.851426774263382,
      "num_tokens": 154854216.0,
      "step": 128760
    },
    {
      "entropy": 1.8539531543850898,
      "epoch": 0.3991757332649286,
      "grad_norm": 7.972646236419678,
      "learning_rate": 4.004168086751243e-06,
      "loss": 0.3709,
      "mean_token_accuracy": 0.8601129725575447,
      "num_tokens": 154867350.0,
      "step": 128770
    },
    {
      "entropy": 1.9226906821131706,
      "epoch": 0.3992067323899783,
      "grad_norm": 3.4731192588806152,
      "learning_rate": 4.0040126170774955e-06,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8491318255662919,
      "num_tokens": 154879654.0,
      "step": 128780
    },
    {
      "entropy": 1.8054320514202118,
      "epoch": 0.399237731515028,
      "grad_norm": 9.286853790283203,
      "learning_rate": 4.003857165511587e-06,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8595024034380913,
      "num_tokens": 154893043.0,
      "step": 128790
    },
    {
      "entropy": 1.8495088413357734,
      "epoch": 0.3992687306400777,
      "grad_norm": 8.713004112243652,
      "learning_rate": 4.003701732050002e-06,
      "loss": 0.423,
      "mean_token_accuracy": 0.8574372127652168,
      "num_tokens": 154905390.0,
      "step": 128800
    },
    {
      "entropy": 1.87299737483263,
      "epoch": 0.3992997297651274,
      "grad_norm": 9.513683319091797,
      "learning_rate": 4.003546316689225e-06,
      "loss": 0.4342,
      "mean_token_accuracy": 0.8459521770477295,
      "num_tokens": 154918334.0,
      "step": 128810
    },
    {
      "entropy": 1.9595311015844346,
      "epoch": 0.39933072889017707,
      "grad_norm": 9.396753311157227,
      "learning_rate": 4.003390919425746e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8480179443955421,
      "num_tokens": 154929275.0,
      "step": 128820
    },
    {
      "entropy": 1.8720010727643968,
      "epoch": 0.3993617280152268,
      "grad_norm": 8.757028579711914,
      "learning_rate": 4.003235540256052e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8585854455828666,
      "num_tokens": 154941320.0,
      "step": 128830
    },
    {
      "entropy": 1.9553624257445335,
      "epoch": 0.39939272714027646,
      "grad_norm": 10.621780395507812,
      "learning_rate": 4.003080179176631e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8457315772771835,
      "num_tokens": 154952903.0,
      "step": 128840
    },
    {
      "entropy": 1.7949346914887427,
      "epoch": 0.3994237262653262,
      "grad_norm": 2.807647466659546,
      "learning_rate": 4.002924836183973e-06,
      "loss": 0.3984,
      "mean_token_accuracy": 0.8518297106027604,
      "num_tokens": 154967282.0,
      "step": 128850
    },
    {
      "entropy": 1.8987135276198388,
      "epoch": 0.39945472539037585,
      "grad_norm": 8.776113510131836,
      "learning_rate": 4.002769511274571e-06,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8481128692626954,
      "num_tokens": 154979746.0,
      "step": 128860
    },
    {
      "entropy": 1.9235922917723656,
      "epoch": 0.3994857245154256,
      "grad_norm": 9.092692375183105,
      "learning_rate": 4.002614204444915e-06,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8417194813489914,
      "num_tokens": 154991799.0,
      "step": 128870
    },
    {
      "entropy": 1.8608113884925843,
      "epoch": 0.39951672364047525,
      "grad_norm": 7.770845413208008,
      "learning_rate": 4.002458915691497e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8441184401512146,
      "num_tokens": 155005153.0,
      "step": 128880
    },
    {
      "entropy": 1.9966261342167855,
      "epoch": 0.39954772276552497,
      "grad_norm": 7.349844932556152,
      "learning_rate": 4.002303645010813e-06,
      "loss": 0.496,
      "mean_token_accuracy": 0.8491346299648285,
      "num_tokens": 155016749.0,
      "step": 128890
    },
    {
      "entropy": 1.8176558762788773,
      "epoch": 0.39957872189057464,
      "grad_norm": 7.31537389755249,
      "learning_rate": 4.002148392399357e-06,
      "loss": 0.386,
      "mean_token_accuracy": 0.8675624996423721,
      "num_tokens": 155030413.0,
      "step": 128900
    },
    {
      "entropy": 1.7823689445853232,
      "epoch": 0.3996097210156243,
      "grad_norm": 8.566244125366211,
      "learning_rate": 4.001993157853624e-06,
      "loss": 0.4131,
      "mean_token_accuracy": 0.8506886452436447,
      "num_tokens": 155044038.0,
      "step": 128910
    },
    {
      "entropy": 1.8545478105545044,
      "epoch": 0.39964072014067403,
      "grad_norm": 7.566455364227295,
      "learning_rate": 4.001837941370112e-06,
      "loss": 0.3984,
      "mean_token_accuracy": 0.8633383393287659,
      "num_tokens": 155056896.0,
      "step": 128920
    },
    {
      "entropy": 1.94671870470047,
      "epoch": 0.3996717192657237,
      "grad_norm": 9.15937328338623,
      "learning_rate": 4.0016827429453155e-06,
      "loss": 0.4891,
      "mean_token_accuracy": 0.8502067849040031,
      "num_tokens": 155067894.0,
      "step": 128930
    },
    {
      "entropy": 1.8951860249042511,
      "epoch": 0.3997027183907734,
      "grad_norm": 9.120491027832031,
      "learning_rate": 4.001527562575736e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8581095486879349,
      "num_tokens": 155079974.0,
      "step": 128940
    },
    {
      "entropy": 1.973567920923233,
      "epoch": 0.3997337175158231,
      "grad_norm": 8.294892311096191,
      "learning_rate": 4.001372400257873e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8433159068226814,
      "num_tokens": 155091036.0,
      "step": 128950
    },
    {
      "entropy": 1.85403670668602,
      "epoch": 0.3997647166408728,
      "grad_norm": 4.179879188537598,
      "learning_rate": 4.001217255988225e-06,
      "loss": 0.4339,
      "mean_token_accuracy": 0.8468100443482399,
      "num_tokens": 155103812.0,
      "step": 128960
    },
    {
      "entropy": 1.8100439786911011,
      "epoch": 0.3997957157659225,
      "grad_norm": 3.8184187412261963,
      "learning_rate": 4.001062129763296e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.870311813056469,
      "num_tokens": 155117031.0,
      "step": 128970
    },
    {
      "entropy": 1.9028596684336663,
      "epoch": 0.3998267148909722,
      "grad_norm": 7.593201637268066,
      "learning_rate": 4.000907021579585e-06,
      "loss": 0.4921,
      "mean_token_accuracy": 0.8530150771141052,
      "num_tokens": 155128674.0,
      "step": 128980
    },
    {
      "entropy": 1.9456078946590423,
      "epoch": 0.3998577140160219,
      "grad_norm": 8.108650207519531,
      "learning_rate": 4.000751931433597e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8464592576026917,
      "num_tokens": 155139476.0,
      "step": 128990
    },
    {
      "entropy": 1.9551035940647126,
      "epoch": 0.3998887131410716,
      "grad_norm": 9.936694145202637,
      "learning_rate": 4.000596859321837e-06,
      "loss": 0.5173,
      "mean_token_accuracy": 0.8364872843027115,
      "num_tokens": 155151556.0,
      "step": 129000
    },
    {
      "entropy": 1.7921033650636673,
      "epoch": 0.3999197122661213,
      "grad_norm": 6.708187103271484,
      "learning_rate": 4.000441805240809e-06,
      "loss": 0.4141,
      "mean_token_accuracy": 0.8557077631354332,
      "num_tokens": 155165232.0,
      "step": 129010
    },
    {
      "entropy": 1.9106485813856124,
      "epoch": 0.399950711391171,
      "grad_norm": 3.6080920696258545,
      "learning_rate": 4.00028676918702e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8508982062339783,
      "num_tokens": 155176886.0,
      "step": 129020
    },
    {
      "entropy": 1.9070033580064774,
      "epoch": 0.39998171051622067,
      "grad_norm": 7.818281650543213,
      "learning_rate": 4.000131751156976e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8354123339056969,
      "num_tokens": 155189493.0,
      "step": 129030
    },
    {
      "entropy": 1.9348053887486458,
      "epoch": 0.4000127096412704,
      "grad_norm": 9.461575508117676,
      "learning_rate": 3.999976751147185e-06,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8445191279053688,
      "num_tokens": 155200851.0,
      "step": 129040
    },
    {
      "entropy": 1.958994448184967,
      "epoch": 0.40004370876632006,
      "grad_norm": 10.124030113220215,
      "learning_rate": 3.999821769154158e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8466064438223839,
      "num_tokens": 155212316.0,
      "step": 129050
    },
    {
      "entropy": 1.9629956305027008,
      "epoch": 0.4000747078913698,
      "grad_norm": 7.477722644805908,
      "learning_rate": 3.999666805174402e-06,
      "loss": 0.5206,
      "mean_token_accuracy": 0.8446162462234497,
      "num_tokens": 155223329.0,
      "step": 129060
    },
    {
      "entropy": 1.8508127138018609,
      "epoch": 0.40010570701641945,
      "grad_norm": 9.45833969116211,
      "learning_rate": 3.999511859204431e-06,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8511440798640251,
      "num_tokens": 155235927.0,
      "step": 129070
    },
    {
      "entropy": 1.8051378756761551,
      "epoch": 0.4001367061414692,
      "grad_norm": 8.635483741760254,
      "learning_rate": 3.999356931240755e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.846815425157547,
      "num_tokens": 155248596.0,
      "step": 129080
    },
    {
      "entropy": 1.7926154106855392,
      "epoch": 0.40016770526651885,
      "grad_norm": 8.050457954406738,
      "learning_rate": 3.999202021279885e-06,
      "loss": 0.3689,
      "mean_token_accuracy": 0.8725738286972046,
      "num_tokens": 155261547.0,
      "step": 129090
    },
    {
      "entropy": 1.9420858532190324,
      "epoch": 0.40019870439156857,
      "grad_norm": 9.84321403503418,
      "learning_rate": 3.999047129318338e-06,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8433678567409515,
      "num_tokens": 155272699.0,
      "step": 129100
    },
    {
      "entropy": 1.7918237030506134,
      "epoch": 0.40022970351661824,
      "grad_norm": 4.470568656921387,
      "learning_rate": 3.998892255352627e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8606519907712936,
      "num_tokens": 155286705.0,
      "step": 129110
    },
    {
      "entropy": 1.8863151326775551,
      "epoch": 0.40026070264166796,
      "grad_norm": 7.721844673156738,
      "learning_rate": 3.998737399379268e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8649403393268585,
      "num_tokens": 155298688.0,
      "step": 129120
    },
    {
      "entropy": 1.9350427404046058,
      "epoch": 0.40029170176671763,
      "grad_norm": 7.882344722747803,
      "learning_rate": 3.998582561394776e-06,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8500554978847503,
      "num_tokens": 155310013.0,
      "step": 129130
    },
    {
      "entropy": 1.8537417277693748,
      "epoch": 0.4003227008917673,
      "grad_norm": 8.492751121520996,
      "learning_rate": 3.998427741395671e-06,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8571956619620323,
      "num_tokens": 155322433.0,
      "step": 129140
    },
    {
      "entropy": 1.9245151728391647,
      "epoch": 0.400353700016817,
      "grad_norm": 2.7976245880126953,
      "learning_rate": 3.99827293937847e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8548303097486496,
      "num_tokens": 155334298.0,
      "step": 129150
    },
    {
      "entropy": 1.9198249489068986,
      "epoch": 0.4003846991418667,
      "grad_norm": 3.582043170928955,
      "learning_rate": 3.998118155339692e-06,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8364661797881127,
      "num_tokens": 155345756.0,
      "step": 129160
    },
    {
      "entropy": 1.9318518668413163,
      "epoch": 0.4004156982669164,
      "grad_norm": 8.037443161010742,
      "learning_rate": 3.997963389275859e-06,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8507610768079757,
      "num_tokens": 155356700.0,
      "step": 129170
    },
    {
      "entropy": 1.9149801477789878,
      "epoch": 0.4004466973919661,
      "grad_norm": 8.608445167541504,
      "learning_rate": 3.99780864118349e-06,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8519381731748581,
      "num_tokens": 155367938.0,
      "step": 129180
    },
    {
      "entropy": 1.8103131994605064,
      "epoch": 0.4004776965170158,
      "grad_norm": 5.219323635101318,
      "learning_rate": 3.9976539110591086e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8530828103423118,
      "num_tokens": 155381084.0,
      "step": 129190
    },
    {
      "entropy": 1.7915343165397644,
      "epoch": 0.4005086956420655,
      "grad_norm": 8.658716201782227,
      "learning_rate": 3.997499198899237e-06,
      "loss": 0.3531,
      "mean_token_accuracy": 0.8671585947275162,
      "num_tokens": 155394355.0,
      "step": 129200
    },
    {
      "entropy": 1.791802540421486,
      "epoch": 0.4005396947671152,
      "grad_norm": 7.633795261383057,
      "learning_rate": 3.997344504700401e-06,
      "loss": 0.3853,
      "mean_token_accuracy": 0.8690174669027328,
      "num_tokens": 155407374.0,
      "step": 129210
    },
    {
      "entropy": 1.8313867419958114,
      "epoch": 0.4005706938921649,
      "grad_norm": 7.055886745452881,
      "learning_rate": 3.997189828459124e-06,
      "loss": 0.3852,
      "mean_token_accuracy": 0.8657983660697937,
      "num_tokens": 155420166.0,
      "step": 129220
    },
    {
      "entropy": 1.8343457579612732,
      "epoch": 0.4006016930172146,
      "grad_norm": 8.553956031799316,
      "learning_rate": 3.997035170171932e-06,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8479964166879654,
      "num_tokens": 155432912.0,
      "step": 129230
    },
    {
      "entropy": 1.8380364283919335,
      "epoch": 0.40063269214226427,
      "grad_norm": 4.127384185791016,
      "learning_rate": 3.996880529835352e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8608715251088143,
      "num_tokens": 155445327.0,
      "step": 129240
    },
    {
      "entropy": 1.8258098438382149,
      "epoch": 0.400663691267314,
      "grad_norm": 3.8394930362701416,
      "learning_rate": 3.996725907445914e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8447212398052215,
      "num_tokens": 155458107.0,
      "step": 129250
    },
    {
      "entropy": 1.9279157117009162,
      "epoch": 0.40069469039236366,
      "grad_norm": 8.191624641418457,
      "learning_rate": 3.996571303000143e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8399505227804184,
      "num_tokens": 155469474.0,
      "step": 129260
    },
    {
      "entropy": 1.8246611893177032,
      "epoch": 0.4007256895174134,
      "grad_norm": 4.23314905166626,
      "learning_rate": 3.996416716494572e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8508252546191215,
      "num_tokens": 155482390.0,
      "step": 129270
    },
    {
      "entropy": 1.796310657262802,
      "epoch": 0.40075668864246305,
      "grad_norm": 3.9624927043914795,
      "learning_rate": 3.996262147925729e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.857945391535759,
      "num_tokens": 155494865.0,
      "step": 129280
    },
    {
      "entropy": 1.8756929695606233,
      "epoch": 0.4007876877675128,
      "grad_norm": 10.10248851776123,
      "learning_rate": 3.996107597290148e-06,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8458318576216698,
      "num_tokens": 155507263.0,
      "step": 129290
    },
    {
      "entropy": 1.853571632504463,
      "epoch": 0.40081868689256245,
      "grad_norm": 8.234441757202148,
      "learning_rate": 3.99595306458436e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8422810658812523,
      "num_tokens": 155519150.0,
      "step": 129300
    },
    {
      "entropy": 1.8298091277480126,
      "epoch": 0.4008496860176122,
      "grad_norm": 4.043398380279541,
      "learning_rate": 3.995798549804898e-06,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8495631635189056,
      "num_tokens": 155531954.0,
      "step": 129310
    },
    {
      "entropy": 1.8689694941043853,
      "epoch": 0.40088068514266184,
      "grad_norm": 8.660704612731934,
      "learning_rate": 3.995644052948298e-06,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8553383216261864,
      "num_tokens": 155543288.0,
      "step": 129320
    },
    {
      "entropy": 1.8053039491176606,
      "epoch": 0.40091168426771157,
      "grad_norm": 8.172130584716797,
      "learning_rate": 3.995489574011096e-06,
      "loss": 0.4413,
      "mean_token_accuracy": 0.8528603136539459,
      "num_tokens": 155555748.0,
      "step": 129330
    },
    {
      "entropy": 1.9306787729263306,
      "epoch": 0.40094268339276123,
      "grad_norm": 9.331960678100586,
      "learning_rate": 3.995335112989825e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8444873213768005,
      "num_tokens": 155566893.0,
      "step": 129340
    },
    {
      "entropy": 1.8586668640375137,
      "epoch": 0.40097368251781096,
      "grad_norm": 9.09153938293457,
      "learning_rate": 3.995180669881025e-06,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8564663842320442,
      "num_tokens": 155578818.0,
      "step": 129350
    },
    {
      "entropy": 1.8796383678913116,
      "epoch": 0.4010046816428606,
      "grad_norm": 8.167010307312012,
      "learning_rate": 3.995026244681234e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.842990031838417,
      "num_tokens": 155590442.0,
      "step": 129360
    },
    {
      "entropy": 1.9271098881959916,
      "epoch": 0.40103568076791035,
      "grad_norm": 9.545135498046875,
      "learning_rate": 3.994871837386989e-06,
      "loss": 0.515,
      "mean_token_accuracy": 0.8358207404613495,
      "num_tokens": 155601889.0,
      "step": 129370
    },
    {
      "entropy": 1.9072769358754158,
      "epoch": 0.40106667989296,
      "grad_norm": 7.5232038497924805,
      "learning_rate": 3.994717447994832e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8516602620482445,
      "num_tokens": 155614028.0,
      "step": 129380
    },
    {
      "entropy": 1.82378898113966,
      "epoch": 0.4010976790180097,
      "grad_norm": 5.845879554748535,
      "learning_rate": 3.994563076501303e-06,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8528670743107796,
      "num_tokens": 155626507.0,
      "step": 129390
    },
    {
      "entropy": 1.8010734841227531,
      "epoch": 0.4011286781430594,
      "grad_norm": 3.6198856830596924,
      "learning_rate": 3.994408722902945e-06,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8558384582400322,
      "num_tokens": 155639677.0,
      "step": 129400
    },
    {
      "entropy": 1.8200062766671181,
      "epoch": 0.4011596772681091,
      "grad_norm": 3.8507888317108154,
      "learning_rate": 3.9942543871963006e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8555910781025886,
      "num_tokens": 155652366.0,
      "step": 129410
    },
    {
      "entropy": 1.866781549155712,
      "epoch": 0.4011906763931588,
      "grad_norm": 6.747231483459473,
      "learning_rate": 3.994100069377912e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8502749800682068,
      "num_tokens": 155664907.0,
      "step": 129420
    },
    {
      "entropy": 1.8888089403510093,
      "epoch": 0.4012216755182085,
      "grad_norm": 8.488142967224121,
      "learning_rate": 3.993945769444325e-06,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8422931343317032,
      "num_tokens": 155676728.0,
      "step": 129430
    },
    {
      "entropy": 1.827822931110859,
      "epoch": 0.4012526746432582,
      "grad_norm": 4.2824883460998535,
      "learning_rate": 3.9937914873920855e-06,
      "loss": 0.4095,
      "mean_token_accuracy": 0.855113011598587,
      "num_tokens": 155689256.0,
      "step": 129440
    },
    {
      "entropy": 1.82640281021595,
      "epoch": 0.40128367376830787,
      "grad_norm": 9.845888137817383,
      "learning_rate": 3.99363722321774e-06,
      "loss": 0.3959,
      "mean_token_accuracy": 0.8682078808546067,
      "num_tokens": 155701978.0,
      "step": 129450
    },
    {
      "entropy": 1.8602233231067657,
      "epoch": 0.4013146728933576,
      "grad_norm": 9.442498207092285,
      "learning_rate": 3.9934829769178365e-06,
      "loss": 0.5007,
      "mean_token_accuracy": 0.8478624925017357,
      "num_tokens": 155714775.0,
      "step": 129460
    },
    {
      "entropy": 1.9146529287099838,
      "epoch": 0.40134567201840726,
      "grad_norm": 7.782778263092041,
      "learning_rate": 3.993328748488922e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8408509835600853,
      "num_tokens": 155726524.0,
      "step": 129470
    },
    {
      "entropy": 1.8323335126042366,
      "epoch": 0.401376671143457,
      "grad_norm": 4.129303455352783,
      "learning_rate": 3.993174537927546e-06,
      "loss": 0.3861,
      "mean_token_accuracy": 0.8619013637304306,
      "num_tokens": 155739316.0,
      "step": 129480
    },
    {
      "entropy": 1.8000301405787469,
      "epoch": 0.40140767026850666,
      "grad_norm": 4.743396282196045,
      "learning_rate": 3.993020345230262e-06,
      "loss": 0.4435,
      "mean_token_accuracy": 0.8503546267747879,
      "num_tokens": 155752513.0,
      "step": 129490
    },
    {
      "entropy": 1.846390649676323,
      "epoch": 0.4014386693935564,
      "grad_norm": 3.5405542850494385,
      "learning_rate": 3.9928661703936164e-06,
      "loss": 0.5197,
      "mean_token_accuracy": 0.8494727715849877,
      "num_tokens": 155764952.0,
      "step": 129500
    },
    {
      "entropy": 1.8727680340409278,
      "epoch": 0.40146966851860605,
      "grad_norm": 7.60482931137085,
      "learning_rate": 3.992712013414165e-06,
      "loss": 0.4306,
      "mean_token_accuracy": 0.8475085526704789,
      "num_tokens": 155777679.0,
      "step": 129510
    },
    {
      "entropy": 1.9042856127023697,
      "epoch": 0.4015006676436558,
      "grad_norm": 7.735726833343506,
      "learning_rate": 3.992557874288459e-06,
      "loss": 0.49,
      "mean_token_accuracy": 0.8530709683895111,
      "num_tokens": 155789778.0,
      "step": 129520
    },
    {
      "entropy": 1.8619258537888528,
      "epoch": 0.40153166676870544,
      "grad_norm": 8.040087699890137,
      "learning_rate": 3.992403753013052e-06,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8604286208748817,
      "num_tokens": 155801838.0,
      "step": 129530
    },
    {
      "entropy": 1.9772648215293884,
      "epoch": 0.40156266589375517,
      "grad_norm": 10.059983253479004,
      "learning_rate": 3.9922496495845015e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8421792671084404,
      "num_tokens": 155812867.0,
      "step": 129540
    },
    {
      "entropy": 1.9216817393898964,
      "epoch": 0.40159366501880484,
      "grad_norm": 7.800021648406982,
      "learning_rate": 3.992095563999361e-06,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8464254021644593,
      "num_tokens": 155823811.0,
      "step": 129550
    },
    {
      "entropy": 1.828902542591095,
      "epoch": 0.40162466414385456,
      "grad_norm": 10.652735710144043,
      "learning_rate": 3.991941496254188e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8681714370846748,
      "num_tokens": 155836495.0,
      "step": 129560
    },
    {
      "entropy": 1.7818224623799324,
      "epoch": 0.40165566326890423,
      "grad_norm": 8.231464385986328,
      "learning_rate": 3.991787446345542e-06,
      "loss": 0.431,
      "mean_token_accuracy": 0.8553294375538826,
      "num_tokens": 155849733.0,
      "step": 129570
    },
    {
      "entropy": 1.933050660789013,
      "epoch": 0.40168666239395395,
      "grad_norm": 9.135824203491211,
      "learning_rate": 3.991633414269979e-06,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8501686662435531,
      "num_tokens": 155861168.0,
      "step": 129580
    },
    {
      "entropy": 1.9456676498055459,
      "epoch": 0.4017176615190036,
      "grad_norm": 7.7797393798828125,
      "learning_rate": 3.9914794000240604e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8487606376409531,
      "num_tokens": 155872569.0,
      "step": 129590
    },
    {
      "entropy": 1.9395177766680718,
      "epoch": 0.40174866064405335,
      "grad_norm": 4.4937334060668945,
      "learning_rate": 3.9913254036043455e-06,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8409850835800171,
      "num_tokens": 155884343.0,
      "step": 129600
    },
    {
      "entropy": 1.96222285926342,
      "epoch": 0.401779659769103,
      "grad_norm": 8.145906448364258,
      "learning_rate": 3.991171425007396e-06,
      "loss": 0.5238,
      "mean_token_accuracy": 0.832419815659523,
      "num_tokens": 155895310.0,
      "step": 129610
    },
    {
      "entropy": 1.9075540453195572,
      "epoch": 0.40181065889415274,
      "grad_norm": 6.813295364379883,
      "learning_rate": 3.991017464229776e-06,
      "loss": 0.4826,
      "mean_token_accuracy": 0.8529313713312149,
      "num_tokens": 155907503.0,
      "step": 129620
    },
    {
      "entropy": 1.8687444925308228,
      "epoch": 0.4018416580192024,
      "grad_norm": 8.512529373168945,
      "learning_rate": 3.990863521268047e-06,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8458928510546684,
      "num_tokens": 155919609.0,
      "step": 129630
    },
    {
      "entropy": 1.937071642279625,
      "epoch": 0.4018726571442521,
      "grad_norm": 8.758728981018066,
      "learning_rate": 3.990709596118774e-06,
      "loss": 0.486,
      "mean_token_accuracy": 0.8510117352008819,
      "num_tokens": 155931150.0,
      "step": 129640
    },
    {
      "entropy": 1.9854602545499802,
      "epoch": 0.4019036562693018,
      "grad_norm": 7.997261047363281,
      "learning_rate": 3.990555688778521e-06,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8390512794256211,
      "num_tokens": 155943070.0,
      "step": 129650
    },
    {
      "entropy": 1.8987844973802566,
      "epoch": 0.40193465539435147,
      "grad_norm": 3.581861734390259,
      "learning_rate": 3.990401799243856e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8575407922267914,
      "num_tokens": 155955435.0,
      "step": 129660
    },
    {
      "entropy": 1.9078197583556176,
      "epoch": 0.4019656545194012,
      "grad_norm": 8.173437118530273,
      "learning_rate": 3.990247927511345e-06,
      "loss": 0.4263,
      "mean_token_accuracy": 0.8537064164876937,
      "num_tokens": 155967188.0,
      "step": 129670
    },
    {
      "entropy": 1.9190190985798836,
      "epoch": 0.40199665364445086,
      "grad_norm": 9.123544692993164,
      "learning_rate": 3.990094073577556e-06,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8527463868260383,
      "num_tokens": 155978679.0,
      "step": 129680
    },
    {
      "entropy": 1.9677730560302735,
      "epoch": 0.4020276527695006,
      "grad_norm": 8.531991004943848,
      "learning_rate": 3.9899402374390585e-06,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8477371469140053,
      "num_tokens": 155989818.0,
      "step": 129690
    },
    {
      "entropy": 1.8901991337537765,
      "epoch": 0.40205865189455026,
      "grad_norm": 4.090048789978027,
      "learning_rate": 3.989786419092422e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8558493494987488,
      "num_tokens": 156002187.0,
      "step": 129700
    },
    {
      "entropy": 1.90907621383667,
      "epoch": 0.4020896510196,
      "grad_norm": 5.129952907562256,
      "learning_rate": 3.989632618534216e-06,
      "loss": 0.4733,
      "mean_token_accuracy": 0.8454459354281425,
      "num_tokens": 156014603.0,
      "step": 129710
    },
    {
      "entropy": 1.821659305691719,
      "epoch": 0.40212065014464965,
      "grad_norm": 7.665596961975098,
      "learning_rate": 3.9894788357610134e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8629834577441216,
      "num_tokens": 156028439.0,
      "step": 129720
    },
    {
      "entropy": 1.8766173496842384,
      "epoch": 0.4021516492696994,
      "grad_norm": 3.5944769382476807,
      "learning_rate": 3.989325070769388e-06,
      "loss": 0.4096,
      "mean_token_accuracy": 0.8606648370623589,
      "num_tokens": 156040404.0,
      "step": 129730
    },
    {
      "entropy": 1.8934729993343353,
      "epoch": 0.40218264839474904,
      "grad_norm": 3.758676052093506,
      "learning_rate": 3.98917132355591e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8599832877516747,
      "num_tokens": 156052297.0,
      "step": 129740
    },
    {
      "entropy": 1.9285873532295228,
      "epoch": 0.40221364751979877,
      "grad_norm": 7.010026454925537,
      "learning_rate": 3.989017594117158e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8508692249655724,
      "num_tokens": 156063027.0,
      "step": 129750
    },
    {
      "entropy": 1.7936880350112916,
      "epoch": 0.40224464664484844,
      "grad_norm": 9.10521125793457,
      "learning_rate": 3.9888638824497034e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8516425848007202,
      "num_tokens": 156075651.0,
      "step": 129760
    },
    {
      "entropy": 1.8739636823534966,
      "epoch": 0.40227564576989816,
      "grad_norm": 8.865586280822754,
      "learning_rate": 3.988710188550125e-06,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8452788576483726,
      "num_tokens": 156087306.0,
      "step": 129770
    },
    {
      "entropy": 1.9241146951913835,
      "epoch": 0.40230664489494783,
      "grad_norm": 8.022590637207031,
      "learning_rate": 3.988556512415e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8422558963298797,
      "num_tokens": 156098765.0,
      "step": 129780
    },
    {
      "entropy": 1.8548906803131104,
      "epoch": 0.40233764401999755,
      "grad_norm": 6.8973493576049805,
      "learning_rate": 3.988402854040903e-06,
      "loss": 0.414,
      "mean_token_accuracy": 0.8576776921749115,
      "num_tokens": 156110825.0,
      "step": 129790
    },
    {
      "entropy": 1.7588834911584854,
      "epoch": 0.4023686431450472,
      "grad_norm": 8.585075378417969,
      "learning_rate": 3.988249213424419e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8594303146004677,
      "num_tokens": 156124449.0,
      "step": 129800
    },
    {
      "entropy": 1.8105765163898468,
      "epoch": 0.40239964227009695,
      "grad_norm": 8.616347312927246,
      "learning_rate": 3.9880955905621235e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.850447241961956,
      "num_tokens": 156137506.0,
      "step": 129810
    },
    {
      "entropy": 1.8526711270213128,
      "epoch": 0.4024306413951466,
      "grad_norm": 8.065713882446289,
      "learning_rate": 3.9879419854506e-06,
      "loss": 0.4683,
      "mean_token_accuracy": 0.8551761761307717,
      "num_tokens": 156149785.0,
      "step": 129820
    },
    {
      "entropy": 1.8890022948384284,
      "epoch": 0.40246164052019634,
      "grad_norm": 7.917675495147705,
      "learning_rate": 3.987788398086428e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8588599741458893,
      "num_tokens": 156160850.0,
      "step": 129830
    },
    {
      "entropy": 1.784026588499546,
      "epoch": 0.402492639645246,
      "grad_norm": 9.626664161682129,
      "learning_rate": 3.987634828466191e-06,
      "loss": 0.3686,
      "mean_token_accuracy": 0.8714671805500984,
      "num_tokens": 156173409.0,
      "step": 129840
    },
    {
      "entropy": 1.8477544769644738,
      "epoch": 0.40252363877029573,
      "grad_norm": 8.537195205688477,
      "learning_rate": 3.987481276586474e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8426776066422462,
      "num_tokens": 156186477.0,
      "step": 129850
    },
    {
      "entropy": 1.736336489021778,
      "epoch": 0.4025546378953454,
      "grad_norm": 2.7146239280700684,
      "learning_rate": 3.98732774244386e-06,
      "loss": 0.4036,
      "mean_token_accuracy": 0.862640056014061,
      "num_tokens": 156200279.0,
      "step": 129860
    },
    {
      "entropy": 1.8483411461114883,
      "epoch": 0.4025856370203951,
      "grad_norm": 3.8119378089904785,
      "learning_rate": 3.987174226034936e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8419306710362434,
      "num_tokens": 156212592.0,
      "step": 129870
    },
    {
      "entropy": 1.9207856342196465,
      "epoch": 0.4026166361454448,
      "grad_norm": 3.5462663173675537,
      "learning_rate": 3.987020727356287e-06,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8504367902874946,
      "num_tokens": 156223684.0,
      "step": 129880
    },
    {
      "entropy": 1.8283285796642303,
      "epoch": 0.40264763527049446,
      "grad_norm": 8.442768096923828,
      "learning_rate": 3.9868672464045005e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8510240688920021,
      "num_tokens": 156236409.0,
      "step": 129890
    },
    {
      "entropy": 1.9546339631080627,
      "epoch": 0.4026786343955442,
      "grad_norm": 7.817316055297852,
      "learning_rate": 3.986713783176166e-06,
      "loss": 0.47,
      "mean_token_accuracy": 0.8490804046392441,
      "num_tokens": 156247811.0,
      "step": 129900
    },
    {
      "entropy": 1.9404812157154083,
      "epoch": 0.40270963352059386,
      "grad_norm": 8.437518119812012,
      "learning_rate": 3.986560337667872e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8467909589409828,
      "num_tokens": 156259779.0,
      "step": 129910
    },
    {
      "entropy": 1.9066607609391213,
      "epoch": 0.4027406326456436,
      "grad_norm": 7.6331305503845215,
      "learning_rate": 3.986406909876207e-06,
      "loss": 0.4798,
      "mean_token_accuracy": 0.8462503507733345,
      "num_tokens": 156271585.0,
      "step": 129920
    },
    {
      "entropy": 1.8766513273119927,
      "epoch": 0.40277163177069325,
      "grad_norm": 3.2706258296966553,
      "learning_rate": 3.986253499797765e-06,
      "loss": 0.4622,
      "mean_token_accuracy": 0.8481630995869637,
      "num_tokens": 156284035.0,
      "step": 129930
    },
    {
      "entropy": 1.8868813574314118,
      "epoch": 0.402802630895743,
      "grad_norm": 8.877392768859863,
      "learning_rate": 3.986100107429135e-06,
      "loss": 0.484,
      "mean_token_accuracy": 0.8435383632779121,
      "num_tokens": 156296009.0,
      "step": 129940
    },
    {
      "entropy": 1.9167783632874489,
      "epoch": 0.40283363002079264,
      "grad_norm": 8.852008819580078,
      "learning_rate": 3.985946732766913e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8485684707760811,
      "num_tokens": 156307643.0,
      "step": 129950
    },
    {
      "entropy": 1.9228922203183174,
      "epoch": 0.40286462914584237,
      "grad_norm": 7.26152229309082,
      "learning_rate": 3.98579337580769e-06,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8554493680596351,
      "num_tokens": 156319030.0,
      "step": 129960
    },
    {
      "entropy": 1.874145193397999,
      "epoch": 0.40289562827089204,
      "grad_norm": 8.481700897216797,
      "learning_rate": 3.985640036548062e-06,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8435692340135574,
      "num_tokens": 156331237.0,
      "step": 129970
    },
    {
      "entropy": 1.8589432962238788,
      "epoch": 0.40292662739594176,
      "grad_norm": 7.1247687339782715,
      "learning_rate": 3.985486714984625e-06,
      "loss": 0.4164,
      "mean_token_accuracy": 0.8517856851220131,
      "num_tokens": 156344722.0,
      "step": 129980
    },
    {
      "entropy": 1.9143261551856994,
      "epoch": 0.40295762652099143,
      "grad_norm": 9.23549747467041,
      "learning_rate": 3.985333411113975e-06,
      "loss": 0.509,
      "mean_token_accuracy": 0.8484362363815308,
      "num_tokens": 156356528.0,
      "step": 129990
    },
    {
      "entropy": 1.9343707114458084,
      "epoch": 0.40298862564604115,
      "grad_norm": 8.890726089477539,
      "learning_rate": 3.985180124932709e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8550547063350677,
      "num_tokens": 156367827.0,
      "step": 130000
    },
    {
      "entropy": 1.8906396552920341,
      "epoch": 0.4030196247710908,
      "grad_norm": 3.963322639465332,
      "learning_rate": 3.9850268564374256e-06,
      "loss": 0.4382,
      "mean_token_accuracy": 0.8531676039099694,
      "num_tokens": 156379328.0,
      "step": 130010
    },
    {
      "entropy": 1.76583993434906,
      "epoch": 0.40305062389614055,
      "grad_norm": 9.098811149597168,
      "learning_rate": 3.9848736056247245e-06,
      "loss": 0.3939,
      "mean_token_accuracy": 0.8658165216445923,
      "num_tokens": 156392630.0,
      "step": 130020
    },
    {
      "entropy": 1.838332974910736,
      "epoch": 0.4030816230211902,
      "grad_norm": 8.506731033325195,
      "learning_rate": 3.984720372491206e-06,
      "loss": 0.435,
      "mean_token_accuracy": 0.8585373058915138,
      "num_tokens": 156405536.0,
      "step": 130030
    },
    {
      "entropy": 1.806590475142002,
      "epoch": 0.40311262214623994,
      "grad_norm": 5.69447660446167,
      "learning_rate": 3.984567157033471e-06,
      "loss": 0.449,
      "mean_token_accuracy": 0.8464230895042419,
      "num_tokens": 156418437.0,
      "step": 130040
    },
    {
      "entropy": 1.9229248493909836,
      "epoch": 0.4031436212712896,
      "grad_norm": 8.785764694213867,
      "learning_rate": 3.9844139592481195e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8359622538089753,
      "num_tokens": 156430560.0,
      "step": 130050
    },
    {
      "entropy": 1.9032806217670442,
      "epoch": 0.40317462039633933,
      "grad_norm": 8.77900218963623,
      "learning_rate": 3.984260779131759e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8543096274137497,
      "num_tokens": 156442730.0,
      "step": 130060
    },
    {
      "entropy": 1.9101463302969932,
      "epoch": 0.403205619521389,
      "grad_norm": 7.89210319519043,
      "learning_rate": 3.984107616680989e-06,
      "loss": 0.4598,
      "mean_token_accuracy": 0.8459172546863556,
      "num_tokens": 156454070.0,
      "step": 130070
    },
    {
      "entropy": 1.9074820682406426,
      "epoch": 0.4032366186464387,
      "grad_norm": 9.048776626586914,
      "learning_rate": 3.983954471892417e-06,
      "loss": 0.4616,
      "mean_token_accuracy": 0.8488796040415764,
      "num_tokens": 156465623.0,
      "step": 130080
    },
    {
      "entropy": 1.8907459661364556,
      "epoch": 0.4032676177714884,
      "grad_norm": 9.203822135925293,
      "learning_rate": 3.983801344762646e-06,
      "loss": 0.439,
      "mean_token_accuracy": 0.8531972080469131,
      "num_tokens": 156477633.0,
      "step": 130090
    },
    {
      "entropy": 1.9408058658242227,
      "epoch": 0.4032986168965381,
      "grad_norm": 8.398517608642578,
      "learning_rate": 3.983648235288285e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8516205415129662,
      "num_tokens": 156488438.0,
      "step": 130100
    },
    {
      "entropy": 1.8947420373558999,
      "epoch": 0.4033296160215878,
      "grad_norm": 7.668551445007324,
      "learning_rate": 3.983495143465942e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8498603880405426,
      "num_tokens": 156499812.0,
      "step": 130110
    },
    {
      "entropy": 1.7990976139903068,
      "epoch": 0.4033606151466375,
      "grad_norm": 3.560288190841675,
      "learning_rate": 3.983342069292223e-06,
      "loss": 0.3632,
      "mean_token_accuracy": 0.8639690637588501,
      "num_tokens": 156512790.0,
      "step": 130120
    },
    {
      "entropy": 1.8366239294409752,
      "epoch": 0.4033916142716872,
      "grad_norm": 9.26343059539795,
      "learning_rate": 3.983189012763739e-06,
      "loss": 0.3931,
      "mean_token_accuracy": 0.8648645102977752,
      "num_tokens": 156525094.0,
      "step": 130130
    },
    {
      "entropy": 1.8781366422772408,
      "epoch": 0.40342261339673685,
      "grad_norm": 7.926459789276123,
      "learning_rate": 3.983035973877099e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8611466780304908,
      "num_tokens": 156537483.0,
      "step": 130140
    },
    {
      "entropy": 1.8221728757023812,
      "epoch": 0.4034536125217866,
      "grad_norm": 3.698610782623291,
      "learning_rate": 3.982882952628916e-06,
      "loss": 0.3992,
      "mean_token_accuracy": 0.8600765854120255,
      "num_tokens": 156549954.0,
      "step": 130150
    },
    {
      "entropy": 1.9444189876317979,
      "epoch": 0.40348461164683624,
      "grad_norm": 8.673882484436035,
      "learning_rate": 3.9827299490158e-06,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8447524651885032,
      "num_tokens": 156561547.0,
      "step": 130160
    },
    {
      "entropy": 1.836028276383877,
      "epoch": 0.40351561077188597,
      "grad_norm": 9.08564567565918,
      "learning_rate": 3.982576963034364e-06,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8561201602220535,
      "num_tokens": 156574641.0,
      "step": 130170
    },
    {
      "entropy": 1.918661078810692,
      "epoch": 0.40354660989693564,
      "grad_norm": 7.405217170715332,
      "learning_rate": 3.982423994681225e-06,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8417747050523758,
      "num_tokens": 156586063.0,
      "step": 130180
    },
    {
      "entropy": 1.8970381796360016,
      "epoch": 0.40357760902198536,
      "grad_norm": 8.172847747802734,
      "learning_rate": 3.982271043952995e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8463637545704842,
      "num_tokens": 156598100.0,
      "step": 130190
    },
    {
      "entropy": 1.896133790910244,
      "epoch": 0.40360860814703503,
      "grad_norm": 2.4702107906341553,
      "learning_rate": 3.9821181108462895e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8457187637686729,
      "num_tokens": 156609713.0,
      "step": 130200
    },
    {
      "entropy": 1.8511537820100785,
      "epoch": 0.40363960727208475,
      "grad_norm": 3.5449514389038086,
      "learning_rate": 3.981965195357727e-06,
      "loss": 0.4154,
      "mean_token_accuracy": 0.8625165119767189,
      "num_tokens": 156622527.0,
      "step": 130210
    },
    {
      "entropy": 1.8973491430282592,
      "epoch": 0.4036706063971344,
      "grad_norm": 7.712673664093018,
      "learning_rate": 3.981812297483923e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8516885727643967,
      "num_tokens": 156634254.0,
      "step": 130220
    },
    {
      "entropy": 1.8966005221009254,
      "epoch": 0.40370160552218415,
      "grad_norm": 8.126500129699707,
      "learning_rate": 3.981659417221498e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8538167625665665,
      "num_tokens": 156646618.0,
      "step": 130230
    },
    {
      "entropy": 1.85729219019413,
      "epoch": 0.4037326046472338,
      "grad_norm": 9.697176933288574,
      "learning_rate": 3.98150655456707e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8491919696331024,
      "num_tokens": 156659338.0,
      "step": 130240
    },
    {
      "entropy": 1.8698885142803192,
      "epoch": 0.40376360377228354,
      "grad_norm": 7.321453094482422,
      "learning_rate": 3.981353709517259e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.850645835697651,
      "num_tokens": 156671170.0,
      "step": 130250
    },
    {
      "entropy": 1.8850924864411354,
      "epoch": 0.4037946028973332,
      "grad_norm": 8.737915992736816,
      "learning_rate": 3.9812008820686864e-06,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8499901428818702,
      "num_tokens": 156683391.0,
      "step": 130260
    },
    {
      "entropy": 1.9188417434692382,
      "epoch": 0.40382560202238293,
      "grad_norm": 8.441498756408691,
      "learning_rate": 3.981048072217976e-06,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8469459965825081,
      "num_tokens": 156695474.0,
      "step": 130270
    },
    {
      "entropy": 1.832290355861187,
      "epoch": 0.4038566011474326,
      "grad_norm": 5.164080619812012,
      "learning_rate": 3.980895279961748e-06,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8454208686947823,
      "num_tokens": 156708620.0,
      "step": 130280
    },
    {
      "entropy": 1.9142166703939438,
      "epoch": 0.40388760027248233,
      "grad_norm": 7.258204936981201,
      "learning_rate": 3.980742505296629e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8631983742117881,
      "num_tokens": 156720342.0,
      "step": 130290
    },
    {
      "entropy": 1.9650160878896714,
      "epoch": 0.403918599397532,
      "grad_norm": 7.455199241638184,
      "learning_rate": 3.980589748219241e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8472874984145164,
      "num_tokens": 156732490.0,
      "step": 130300
    },
    {
      "entropy": 1.816853478550911,
      "epoch": 0.4039495985225817,
      "grad_norm": 3.841028928756714,
      "learning_rate": 3.980437008726212e-06,
      "loss": 0.3849,
      "mean_token_accuracy": 0.8634612441062928,
      "num_tokens": 156745943.0,
      "step": 130310
    },
    {
      "entropy": 1.9299751341342926,
      "epoch": 0.4039805976476314,
      "grad_norm": 4.5318474769592285,
      "learning_rate": 3.980284286814167e-06,
      "loss": 0.4939,
      "mean_token_accuracy": 0.8458565920591354,
      "num_tokens": 156757351.0,
      "step": 130320
    },
    {
      "entropy": 1.8615489616990089,
      "epoch": 0.4040115967726811,
      "grad_norm": 8.26394271850586,
      "learning_rate": 3.980131582479735e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8518011167645454,
      "num_tokens": 156768452.0,
      "step": 130330
    },
    {
      "entropy": 1.9241826206445694,
      "epoch": 0.4040425958977308,
      "grad_norm": 9.016687393188477,
      "learning_rate": 3.979978895719543e-06,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8521223112940788,
      "num_tokens": 156779453.0,
      "step": 130340
    },
    {
      "entropy": 2.0048281461000443,
      "epoch": 0.4040735950227805,
      "grad_norm": 8.775766372680664,
      "learning_rate": 3.979826226530221e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8402179971337318,
      "num_tokens": 156789976.0,
      "step": 130350
    },
    {
      "entropy": 1.8103154242038726,
      "epoch": 0.4041045941478302,
      "grad_norm": 3.5552315711975098,
      "learning_rate": 3.9796735749084e-06,
      "loss": 0.3948,
      "mean_token_accuracy": 0.8577532470226288,
      "num_tokens": 156803981.0,
      "step": 130360
    },
    {
      "entropy": 1.9318371683359146,
      "epoch": 0.4041355932728799,
      "grad_norm": 8.001313209533691,
      "learning_rate": 3.97952094085071e-06,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8499713644385338,
      "num_tokens": 156816659.0,
      "step": 130370
    },
    {
      "entropy": 1.9495342776179314,
      "epoch": 0.40416659239792957,
      "grad_norm": 4.196635723114014,
      "learning_rate": 3.979368324353783e-06,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8435593828558922,
      "num_tokens": 156828530.0,
      "step": 130380
    },
    {
      "entropy": 1.8581201523542403,
      "epoch": 0.40419759152297924,
      "grad_norm": 7.378039836883545,
      "learning_rate": 3.979215725414253e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.8562660500407219,
      "num_tokens": 156841203.0,
      "step": 130390
    },
    {
      "entropy": 1.8645180001854897,
      "epoch": 0.40422859064802896,
      "grad_norm": 7.564363956451416,
      "learning_rate": 3.9790631440287516e-06,
      "loss": 0.3761,
      "mean_token_accuracy": 0.8670828878879547,
      "num_tokens": 156853244.0,
      "step": 130400
    },
    {
      "entropy": 1.9289032772183419,
      "epoch": 0.40425958977307863,
      "grad_norm": 4.7226481437683105,
      "learning_rate": 3.978910580193916e-06,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8462576329708099,
      "num_tokens": 156864953.0,
      "step": 130410
    },
    {
      "entropy": 1.9896679311990737,
      "epoch": 0.40429058889812836,
      "grad_norm": 9.488637924194336,
      "learning_rate": 3.978758033906382e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8406891971826553,
      "num_tokens": 156875482.0,
      "step": 130420
    },
    {
      "entropy": 1.859781025350094,
      "epoch": 0.404321588023178,
      "grad_norm": 3.904017210006714,
      "learning_rate": 3.978605505162784e-06,
      "loss": 0.3909,
      "mean_token_accuracy": 0.8608370333909988,
      "num_tokens": 156887299.0,
      "step": 130430
    },
    {
      "entropy": 1.8310018733143807,
      "epoch": 0.40435258714822775,
      "grad_norm": 8.23388671875,
      "learning_rate": 3.978452993959761e-06,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8504728376865387,
      "num_tokens": 156900461.0,
      "step": 130440
    },
    {
      "entropy": 1.877791903913021,
      "epoch": 0.4043835862732774,
      "grad_norm": 7.384108066558838,
      "learning_rate": 3.978300500293951e-06,
      "loss": 0.4635,
      "mean_token_accuracy": 0.8471254542469978,
      "num_tokens": 156913284.0,
      "step": 130450
    },
    {
      "entropy": 1.6766640424728394,
      "epoch": 0.40441458539832714,
      "grad_norm": 3.489778995513916,
      "learning_rate": 3.978148024161993e-06,
      "loss": 0.3129,
      "mean_token_accuracy": 0.873348993062973,
      "num_tokens": 156928424.0,
      "step": 130460
    },
    {
      "entropy": 1.8710533007979393,
      "epoch": 0.4044455845233768,
      "grad_norm": 6.751518726348877,
      "learning_rate": 3.977995565560528e-06,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8581561148166656,
      "num_tokens": 156940418.0,
      "step": 130470
    },
    {
      "entropy": 1.8982125908136367,
      "epoch": 0.40447658364842654,
      "grad_norm": 7.623547554016113,
      "learning_rate": 3.977843124486196e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.859375411272049,
      "num_tokens": 156952679.0,
      "step": 130480
    },
    {
      "entropy": 1.9297777190804482,
      "epoch": 0.4045075827734762,
      "grad_norm": 8.541840553283691,
      "learning_rate": 3.977690700935639e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8506446555256844,
      "num_tokens": 156964447.0,
      "step": 130490
    },
    {
      "entropy": 1.8197549387812615,
      "epoch": 0.40453858189852593,
      "grad_norm": 4.050614833831787,
      "learning_rate": 3.9775382949055e-06,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8605873674154282,
      "num_tokens": 156977447.0,
      "step": 130500
    },
    {
      "entropy": 1.7729148700833322,
      "epoch": 0.4045695810235756,
      "grad_norm": 4.528966903686523,
      "learning_rate": 3.977385906392423e-06,
      "loss": 0.3785,
      "mean_token_accuracy": 0.8583493396639824,
      "num_tokens": 156991476.0,
      "step": 130510
    },
    {
      "entropy": 1.9727448597550392,
      "epoch": 0.4046005801486253,
      "grad_norm": 9.29497241973877,
      "learning_rate": 3.977233535393054e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.843626768887043,
      "num_tokens": 157002963.0,
      "step": 130520
    },
    {
      "entropy": 1.9293553322553634,
      "epoch": 0.404631579273675,
      "grad_norm": 9.896842956542969,
      "learning_rate": 3.977081181904036e-06,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8478871420025825,
      "num_tokens": 157014701.0,
      "step": 130530
    },
    {
      "entropy": 1.977966983616352,
      "epoch": 0.4046625783987247,
      "grad_norm": 8.197452545166016,
      "learning_rate": 3.976928845922018e-06,
      "loss": 0.5224,
      "mean_token_accuracy": 0.8346485048532486,
      "num_tokens": 157026114.0,
      "step": 130540
    },
    {
      "entropy": 1.9490982070565224,
      "epoch": 0.4046935775237744,
      "grad_norm": 3.955064058303833,
      "learning_rate": 3.976776527443644e-06,
      "loss": 0.4488,
      "mean_token_accuracy": 0.8499052435159683,
      "num_tokens": 157038105.0,
      "step": 130550
    },
    {
      "entropy": 1.921256160736084,
      "epoch": 0.4047245766488241,
      "grad_norm": 7.925784111022949,
      "learning_rate": 3.9766242264655655e-06,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8429896846413613,
      "num_tokens": 157050666.0,
      "step": 130560
    },
    {
      "entropy": 1.9208467230200768,
      "epoch": 0.4047555757738738,
      "grad_norm": 4.279351711273193,
      "learning_rate": 3.976471942984431e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8471018105745316,
      "num_tokens": 157062556.0,
      "step": 130570
    },
    {
      "entropy": 1.8374285951256752,
      "epoch": 0.4047865748989235,
      "grad_norm": 8.640376091003418,
      "learning_rate": 3.976319676996889e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8620464310050011,
      "num_tokens": 157075344.0,
      "step": 130580
    },
    {
      "entropy": 1.9202581122517586,
      "epoch": 0.40481757402397317,
      "grad_norm": 7.99689245223999,
      "learning_rate": 3.976167428499592e-06,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8536014124751091,
      "num_tokens": 157086341.0,
      "step": 130590
    },
    {
      "entropy": 1.9324933484196662,
      "epoch": 0.4048485731490229,
      "grad_norm": 6.284863471984863,
      "learning_rate": 3.976015197489192e-06,
      "loss": 0.5003,
      "mean_token_accuracy": 0.8359458863735199,
      "num_tokens": 157098510.0,
      "step": 130600
    },
    {
      "entropy": 1.811049999296665,
      "epoch": 0.40487957227407256,
      "grad_norm": 8.513614654541016,
      "learning_rate": 3.97586298396234e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8566889658570289,
      "num_tokens": 157111635.0,
      "step": 130610
    },
    {
      "entropy": 1.895810031890869,
      "epoch": 0.40491057139912223,
      "grad_norm": 8.150436401367188,
      "learning_rate": 3.975710787915691e-06,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8592909917235374,
      "num_tokens": 157123381.0,
      "step": 130620
    },
    {
      "entropy": 1.8806810095906257,
      "epoch": 0.40494157052417196,
      "grad_norm": 9.090899467468262,
      "learning_rate": 3.9755586093459e-06,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8572476267814636,
      "num_tokens": 157135740.0,
      "step": 130630
    },
    {
      "entropy": 1.7848129168152809,
      "epoch": 0.4049725696492216,
      "grad_norm": 4.283092975616455,
      "learning_rate": 3.97540644824962e-06,
      "loss": 0.3555,
      "mean_token_accuracy": 0.8615420237183571,
      "num_tokens": 157148676.0,
      "step": 130640
    },
    {
      "entropy": 1.8715970084071158,
      "epoch": 0.40500356877427135,
      "grad_norm": 9.602254867553711,
      "learning_rate": 3.975254304623512e-06,
      "loss": 0.4169,
      "mean_token_accuracy": 0.8585036590695381,
      "num_tokens": 157161564.0,
      "step": 130650
    },
    {
      "entropy": 1.8966206587851047,
      "epoch": 0.405034567899321,
      "grad_norm": 6.511444091796875,
      "learning_rate": 3.975102178464229e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8608822152018547,
      "num_tokens": 157174560.0,
      "step": 130660
    },
    {
      "entropy": 1.9423101127147675,
      "epoch": 0.40506556702437074,
      "grad_norm": 8.167618751525879,
      "learning_rate": 3.974950069768429e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8432308033108711,
      "num_tokens": 157186728.0,
      "step": 130670
    },
    {
      "entropy": 1.9459773391485213,
      "epoch": 0.4050965661494204,
      "grad_norm": 8.282353401184082,
      "learning_rate": 3.974797978532774e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.8470408290624618,
      "num_tokens": 157198743.0,
      "step": 130680
    },
    {
      "entropy": 1.8234005078673363,
      "epoch": 0.40512756527447014,
      "grad_norm": 3.401383399963379,
      "learning_rate": 3.974645904753922e-06,
      "loss": 0.3927,
      "mean_token_accuracy": 0.8553168892860412,
      "num_tokens": 157211958.0,
      "step": 130690
    },
    {
      "entropy": 1.9370823100209236,
      "epoch": 0.4051585643995198,
      "grad_norm": 6.692420959472656,
      "learning_rate": 3.974493848428535e-06,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8524964049458503,
      "num_tokens": 157223127.0,
      "step": 130700
    },
    {
      "entropy": 1.9691085010766982,
      "epoch": 0.40518956352456953,
      "grad_norm": 8.417125701904297,
      "learning_rate": 3.974341809553272e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8546971321105957,
      "num_tokens": 157234327.0,
      "step": 130710
    },
    {
      "entropy": 1.8974324196577073,
      "epoch": 0.4052205626496192,
      "grad_norm": 9.326712608337402,
      "learning_rate": 3.974189788124799e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8539463758468628,
      "num_tokens": 157246787.0,
      "step": 130720
    },
    {
      "entropy": 1.9472648695111274,
      "epoch": 0.4052515617746689,
      "grad_norm": 8.63284683227539,
      "learning_rate": 3.974037784139778e-06,
      "loss": 0.4983,
      "mean_token_accuracy": 0.8510000556707382,
      "num_tokens": 157258105.0,
      "step": 130730
    },
    {
      "entropy": 1.9133790254592895,
      "epoch": 0.4052825608997186,
      "grad_norm": 7.009464263916016,
      "learning_rate": 3.973885797594873e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.8659856930375099,
      "num_tokens": 157269330.0,
      "step": 130740
    },
    {
      "entropy": 1.9145527601242065,
      "epoch": 0.4053135600247683,
      "grad_norm": 6.98771333694458,
      "learning_rate": 3.973733828486749e-06,
      "loss": 0.4235,
      "mean_token_accuracy": 0.8578495368361473,
      "num_tokens": 157282000.0,
      "step": 130750
    },
    {
      "entropy": 1.9058576181530953,
      "epoch": 0.405344559149818,
      "grad_norm": 9.95231819152832,
      "learning_rate": 3.9735818768120745e-06,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8475971177220345,
      "num_tokens": 157293637.0,
      "step": 130760
    },
    {
      "entropy": 1.789780667424202,
      "epoch": 0.4053755582748677,
      "grad_norm": 7.046390533447266,
      "learning_rate": 3.973429942567513e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8669392168521881,
      "num_tokens": 157306793.0,
      "step": 130770
    },
    {
      "entropy": 1.988757422566414,
      "epoch": 0.4054065573999174,
      "grad_norm": 9.767016410827637,
      "learning_rate": 3.973278025749736e-06,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8483888059854507,
      "num_tokens": 157318126.0,
      "step": 130780
    },
    {
      "entropy": 1.9013223245739936,
      "epoch": 0.4054375565249671,
      "grad_norm": 4.095386028289795,
      "learning_rate": 3.97312612635541e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8486203476786613,
      "num_tokens": 157329583.0,
      "step": 130790
    },
    {
      "entropy": 1.9669806063175201,
      "epoch": 0.40546855565001677,
      "grad_norm": 7.4656596183776855,
      "learning_rate": 3.9729742443812056e-06,
      "loss": 0.5092,
      "mean_token_accuracy": 0.8445022612810135,
      "num_tokens": 157340387.0,
      "step": 130800
    },
    {
      "entropy": 1.8516760841012,
      "epoch": 0.4054995547750665,
      "grad_norm": Infinity,
      "learning_rate": 3.972822379823793e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.8624450042843819,
      "num_tokens": 157353043.0,
      "step": 130810
    },
    {
      "entropy": 1.915142248570919,
      "epoch": 0.40553055390011616,
      "grad_norm": 7.543905258178711,
      "learning_rate": 3.972670532679844e-06,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8492273300886154,
      "num_tokens": 157365310.0,
      "step": 130820
    },
    {
      "entropy": 1.8361800596117974,
      "epoch": 0.4055615530251659,
      "grad_norm": 4.767913341522217,
      "learning_rate": 3.9725187029460316e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8620738372206688,
      "num_tokens": 157377616.0,
      "step": 130830
    },
    {
      "entropy": 1.9048438847064972,
      "epoch": 0.40559255215021556,
      "grad_norm": 8.4035005569458,
      "learning_rate": 3.972366890619029e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8537156358361244,
      "num_tokens": 157389121.0,
      "step": 130840
    },
    {
      "entropy": 1.9738389521837234,
      "epoch": 0.4056235512752653,
      "grad_norm": 8.513267517089844,
      "learning_rate": 3.972215095695508e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8514445766806602,
      "num_tokens": 157400361.0,
      "step": 130850
    },
    {
      "entropy": 1.9617274329066277,
      "epoch": 0.40565455040031495,
      "grad_norm": 4.258702278137207,
      "learning_rate": 3.972063318172147e-06,
      "loss": 0.4877,
      "mean_token_accuracy": 0.847972746193409,
      "num_tokens": 157411570.0,
      "step": 130860
    },
    {
      "entropy": 1.9130688726902008,
      "epoch": 0.4056855495253646,
      "grad_norm": 7.041831970214844,
      "learning_rate": 3.97191155804562e-06,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8498567163944244,
      "num_tokens": 157423298.0,
      "step": 130870
    },
    {
      "entropy": 1.8860970079898833,
      "epoch": 0.40571654865041434,
      "grad_norm": 7.607229709625244,
      "learning_rate": 3.971759815312605e-06,
      "loss": 0.4647,
      "mean_token_accuracy": 0.8536626756191253,
      "num_tokens": 157434605.0,
      "step": 130880
    },
    {
      "entropy": 1.9061809971928596,
      "epoch": 0.405747547775464,
      "grad_norm": 7.09719705581665,
      "learning_rate": 3.971608089969779e-06,
      "loss": 0.4901,
      "mean_token_accuracy": 0.848254905641079,
      "num_tokens": 157446268.0,
      "step": 130890
    },
    {
      "entropy": 1.7818841926753521,
      "epoch": 0.40577854690051374,
      "grad_norm": 4.486595630645752,
      "learning_rate": 3.971456382013821e-06,
      "loss": 0.3597,
      "mean_token_accuracy": 0.8559565782546997,
      "num_tokens": 157460438.0,
      "step": 130900
    },
    {
      "entropy": 1.763669066131115,
      "epoch": 0.4058095460255634,
      "grad_norm": 3.9831771850585938,
      "learning_rate": 3.97130469144141e-06,
      "loss": 0.3303,
      "mean_token_accuracy": 0.8683875605463982,
      "num_tokens": 157475220.0,
      "step": 130910
    },
    {
      "entropy": 1.8374385461211205,
      "epoch": 0.40584054515061313,
      "grad_norm": 8.03730583190918,
      "learning_rate": 3.9711530182492266e-06,
      "loss": 0.4052,
      "mean_token_accuracy": 0.858574740588665,
      "num_tokens": 157487520.0,
      "step": 130920
    },
    {
      "entropy": 1.9221992582082748,
      "epoch": 0.4058715442756628,
      "grad_norm": 9.426704406738281,
      "learning_rate": 3.971001362433953e-06,
      "loss": 0.484,
      "mean_token_accuracy": 0.8404927179217339,
      "num_tokens": 157499623.0,
      "step": 130930
    },
    {
      "entropy": 1.8548153042793274,
      "epoch": 0.4059025434007125,
      "grad_norm": 4.058870792388916,
      "learning_rate": 3.97084972399227e-06,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8509564101696014,
      "num_tokens": 157512269.0,
      "step": 130940
    },
    {
      "entropy": 1.898500706255436,
      "epoch": 0.4059335425257622,
      "grad_norm": 7.100831985473633,
      "learning_rate": 3.970698102920861e-06,
      "loss": 0.4517,
      "mean_token_accuracy": 0.8517476499080658,
      "num_tokens": 157524534.0,
      "step": 130950
    },
    {
      "entropy": 1.938453209400177,
      "epoch": 0.4059645416508119,
      "grad_norm": 7.448749542236328,
      "learning_rate": 3.970546499216411e-06,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8483160063624382,
      "num_tokens": 157535978.0,
      "step": 130960
    },
    {
      "entropy": 1.8814398035407067,
      "epoch": 0.4059955407758616,
      "grad_norm": 8.682611465454102,
      "learning_rate": 3.970394912875604e-06,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8522828727960586,
      "num_tokens": 157548563.0,
      "step": 130970
    },
    {
      "entropy": 1.9689362928271295,
      "epoch": 0.4060265399009113,
      "grad_norm": 8.23831558227539,
      "learning_rate": 3.970243343895126e-06,
      "loss": 0.4459,
      "mean_token_accuracy": 0.8545093983411789,
      "num_tokens": 157559741.0,
      "step": 130980
    },
    {
      "entropy": 1.8692365869879723,
      "epoch": 0.406057539025961,
      "grad_norm": 8.109051704406738,
      "learning_rate": 3.970091792271663e-06,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8523834586143494,
      "num_tokens": 157571939.0,
      "step": 130990
    },
    {
      "entropy": 1.9243786290287972,
      "epoch": 0.4060885381510107,
      "grad_norm": 10.623024940490723,
      "learning_rate": 3.969940258001903e-06,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8421697750687599,
      "num_tokens": 157583545.0,
      "step": 131000
    },
    {
      "entropy": 1.817786581814289,
      "epoch": 0.40611953727606037,
      "grad_norm": 9.550358772277832,
      "learning_rate": 3.969788741082535e-06,
      "loss": 0.3799,
      "mean_token_accuracy": 0.8594713494181633,
      "num_tokens": 157596216.0,
      "step": 131010
    },
    {
      "entropy": 1.9418366000056266,
      "epoch": 0.4061505364011101,
      "grad_norm": 7.840337753295898,
      "learning_rate": 3.969637241510247e-06,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8537692606449128,
      "num_tokens": 157607196.0,
      "step": 131020
    },
    {
      "entropy": 1.8996048077940941,
      "epoch": 0.40618153552615976,
      "grad_norm": 8.591992378234863,
      "learning_rate": 3.9694857592817295e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8468469053506851,
      "num_tokens": 157619202.0,
      "step": 131030
    },
    {
      "entropy": 1.8347910821437836,
      "epoch": 0.4062125346512095,
      "grad_norm": 3.285649299621582,
      "learning_rate": 3.969334294393675e-06,
      "loss": 0.448,
      "mean_token_accuracy": 0.8462332352995873,
      "num_tokens": 157632132.0,
      "step": 131040
    },
    {
      "entropy": 1.9242204323410987,
      "epoch": 0.40624353377625916,
      "grad_norm": 4.810262680053711,
      "learning_rate": 3.969182846842773e-06,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8468155965209008,
      "num_tokens": 157643575.0,
      "step": 131050
    },
    {
      "entropy": 1.9612080991268157,
      "epoch": 0.4062745329013089,
      "grad_norm": 9.51435375213623,
      "learning_rate": 3.9690314166257186e-06,
      "loss": 0.5294,
      "mean_token_accuracy": 0.8334100112318993,
      "num_tokens": 157654699.0,
      "step": 131060
    },
    {
      "entropy": 1.9367137864232062,
      "epoch": 0.40630553202635855,
      "grad_norm": 9.918806076049805,
      "learning_rate": 3.968880003739205e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.851118703186512,
      "num_tokens": 157665786.0,
      "step": 131070
    },
    {
      "entropy": 1.8767028152942657,
      "epoch": 0.4063365311514083,
      "grad_norm": 10.18786334991455,
      "learning_rate": 3.9687286081799244e-06,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8559330344200134,
      "num_tokens": 157678031.0,
      "step": 131080
    },
    {
      "entropy": 1.9314553529024123,
      "epoch": 0.40636753027645794,
      "grad_norm": 6.053582191467285,
      "learning_rate": 3.9685772299445754e-06,
      "loss": 0.4421,
      "mean_token_accuracy": 0.8622308388352394,
      "num_tokens": 157689998.0,
      "step": 131090
    },
    {
      "entropy": 1.6999683201313018,
      "epoch": 0.40639852940150767,
      "grad_norm": 5.942327499389648,
      "learning_rate": 3.9684258690298525e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8629785016179085,
      "num_tokens": 157704433.0,
      "step": 131100
    },
    {
      "entropy": 1.864445061981678,
      "epoch": 0.40642952852655734,
      "grad_norm": 7.0887346267700195,
      "learning_rate": 3.968274525432454e-06,
      "loss": 0.4122,
      "mean_token_accuracy": 0.8601906552910805,
      "num_tokens": 157717013.0,
      "step": 131110
    },
    {
      "entropy": 1.8925062775611878,
      "epoch": 0.406460527651607,
      "grad_norm": 10.952113151550293,
      "learning_rate": 3.968123199149077e-06,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8586850896477699,
      "num_tokens": 157729145.0,
      "step": 131120
    },
    {
      "entropy": 1.9502949953079223,
      "epoch": 0.40649152677665673,
      "grad_norm": 10.193359375,
      "learning_rate": 3.967971890176421e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8508687302470207,
      "num_tokens": 157740267.0,
      "step": 131130
    },
    {
      "entropy": 1.9118008241057396,
      "epoch": 0.4065225259017064,
      "grad_norm": 3.9366962909698486,
      "learning_rate": 3.967820598511186e-06,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8517689287662507,
      "num_tokens": 157751681.0,
      "step": 131140
    },
    {
      "entropy": 1.9390445798635483,
      "epoch": 0.4065535250267561,
      "grad_norm": 9.202500343322754,
      "learning_rate": 3.9676693241500725e-06,
      "loss": 0.4959,
      "mean_token_accuracy": 0.8437588036060333,
      "num_tokens": 157762806.0,
      "step": 131150
    },
    {
      "entropy": 1.873625774681568,
      "epoch": 0.4065845241518058,
      "grad_norm": 9.123173713684082,
      "learning_rate": 3.967518067089782e-06,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8525558322668075,
      "num_tokens": 157775344.0,
      "step": 131160
    },
    {
      "entropy": 1.8773087307810783,
      "epoch": 0.4066155232768555,
      "grad_norm": 7.2432684898376465,
      "learning_rate": 3.967366827327019e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8548357993364334,
      "num_tokens": 157788421.0,
      "step": 131170
    },
    {
      "entropy": 1.9489660397171975,
      "epoch": 0.4066465224019052,
      "grad_norm": 8.408989906311035,
      "learning_rate": 3.9672156048584825e-06,
      "loss": 0.5063,
      "mean_token_accuracy": 0.8430541291832924,
      "num_tokens": 157799818.0,
      "step": 131180
    },
    {
      "entropy": 1.8156598702073097,
      "epoch": 0.4066775215269549,
      "grad_norm": 3.8361010551452637,
      "learning_rate": 3.9670643996808805e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.8656172275543212,
      "num_tokens": 157812552.0,
      "step": 131190
    },
    {
      "entropy": 1.876960425078869,
      "epoch": 0.4067085206520046,
      "grad_norm": 4.139932632446289,
      "learning_rate": 3.966913211790917e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8435041248798371,
      "num_tokens": 157824792.0,
      "step": 131200
    },
    {
      "entropy": 1.8027958139777183,
      "epoch": 0.4067395197770543,
      "grad_norm": 8.876256942749023,
      "learning_rate": 3.966762041185298e-06,
      "loss": 0.4064,
      "mean_token_accuracy": 0.8562049359083176,
      "num_tokens": 157837760.0,
      "step": 131210
    },
    {
      "entropy": 1.845954880863428,
      "epoch": 0.406770518902104,
      "grad_norm": 5.402946949005127,
      "learning_rate": 3.966610887860731e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8410774037241936,
      "num_tokens": 157850821.0,
      "step": 131220
    },
    {
      "entropy": 1.9590726420283318,
      "epoch": 0.4068015180271537,
      "grad_norm": 9.471705436706543,
      "learning_rate": 3.966459751813921e-06,
      "loss": 0.4975,
      "mean_token_accuracy": 0.8433647140860557,
      "num_tokens": 157862272.0,
      "step": 131230
    },
    {
      "entropy": 1.8670239843428136,
      "epoch": 0.40683251715220337,
      "grad_norm": 8.589696884155273,
      "learning_rate": 3.966308633041582e-06,
      "loss": 0.432,
      "mean_token_accuracy": 0.8549284100532532,
      "num_tokens": 157874516.0,
      "step": 131240
    },
    {
      "entropy": 1.8836659103631974,
      "epoch": 0.4068635162772531,
      "grad_norm": 7.1528449058532715,
      "learning_rate": 3.9661575315404185e-06,
      "loss": 0.536,
      "mean_token_accuracy": 0.8425373405218124,
      "num_tokens": 157887098.0,
      "step": 131250
    },
    {
      "entropy": 1.9439944818615913,
      "epoch": 0.40689451540230276,
      "grad_norm": 4.756436824798584,
      "learning_rate": 3.966006447307143e-06,
      "loss": 0.5211,
      "mean_token_accuracy": 0.8354384452104568,
      "num_tokens": 157898487.0,
      "step": 131260
    },
    {
      "entropy": 1.9490654364228248,
      "epoch": 0.4069255145273525,
      "grad_norm": 8.603204727172852,
      "learning_rate": 3.965855380338467e-06,
      "loss": 0.4939,
      "mean_token_accuracy": 0.841651976108551,
      "num_tokens": 157909630.0,
      "step": 131270
    },
    {
      "entropy": 1.831677147746086,
      "epoch": 0.40695651365240215,
      "grad_norm": 4.0721611976623535,
      "learning_rate": 3.965704330631102e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8579563215374947,
      "num_tokens": 157922831.0,
      "step": 131280
    },
    {
      "entropy": 1.8643750965595245,
      "epoch": 0.4069875127774519,
      "grad_norm": 9.7418851852417,
      "learning_rate": 3.965553298181761e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.848840269446373,
      "num_tokens": 157934557.0,
      "step": 131290
    },
    {
      "entropy": 1.9244591280817986,
      "epoch": 0.40701851190250155,
      "grad_norm": 7.745121955871582,
      "learning_rate": 3.965402282987159e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8503209039568901,
      "num_tokens": 157945931.0,
      "step": 131300
    },
    {
      "entropy": 1.866335666179657,
      "epoch": 0.40704951102755127,
      "grad_norm": 8.668558120727539,
      "learning_rate": 3.96525128504401e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.849652573466301,
      "num_tokens": 157958476.0,
      "step": 131310
    },
    {
      "entropy": 1.8344866186380386,
      "epoch": 0.40708051015260094,
      "grad_norm": 7.905651569366455,
      "learning_rate": 3.965100304349029e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.859797616302967,
      "num_tokens": 157971189.0,
      "step": 131320
    },
    {
      "entropy": 1.9577922523021698,
      "epoch": 0.40711150927765066,
      "grad_norm": 7.413038730621338,
      "learning_rate": 3.964949340898934e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8549491494894028,
      "num_tokens": 157981983.0,
      "step": 131330
    },
    {
      "entropy": 1.9203011736273765,
      "epoch": 0.40714250840270033,
      "grad_norm": 9.896458625793457,
      "learning_rate": 3.9647983946904416e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.8488733306527138,
      "num_tokens": 157993609.0,
      "step": 131340
    },
    {
      "entropy": 1.9292757645249368,
      "epoch": 0.40717350752775006,
      "grad_norm": 9.621637344360352,
      "learning_rate": 3.964647465720271e-06,
      "loss": 0.5658,
      "mean_token_accuracy": 0.8377192333340645,
      "num_tokens": 158005340.0,
      "step": 131350
    },
    {
      "entropy": 1.9438342347741127,
      "epoch": 0.4072045066527997,
      "grad_norm": 7.706085681915283,
      "learning_rate": 3.964496553985139e-06,
      "loss": 0.4935,
      "mean_token_accuracy": 0.8447881907224655,
      "num_tokens": 158016548.0,
      "step": 131360
    },
    {
      "entropy": 1.825490552186966,
      "epoch": 0.4072355057778494,
      "grad_norm": 4.330419063568115,
      "learning_rate": 3.964345659481768e-06,
      "loss": 0.4223,
      "mean_token_accuracy": 0.8532146289944649,
      "num_tokens": 158029148.0,
      "step": 131370
    },
    {
      "entropy": 1.8673056378960609,
      "epoch": 0.4072665049028991,
      "grad_norm": 7.6978230476379395,
      "learning_rate": 3.964194782206878e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8550721898674964,
      "num_tokens": 158041404.0,
      "step": 131380
    },
    {
      "entropy": 1.8040344282984733,
      "epoch": 0.4072975040279488,
      "grad_norm": 5.431396007537842,
      "learning_rate": 3.964043922157191e-06,
      "loss": 0.3902,
      "mean_token_accuracy": 0.8702194571495057,
      "num_tokens": 158054518.0,
      "step": 131390
    },
    {
      "entropy": 1.8987352877855301,
      "epoch": 0.4073285031529985,
      "grad_norm": 7.753914833068848,
      "learning_rate": 3.963893079329429e-06,
      "loss": 0.4926,
      "mean_token_accuracy": 0.8489699825644493,
      "num_tokens": 158065523.0,
      "step": 131400
    },
    {
      "entropy": 1.918555736541748,
      "epoch": 0.4073595022780482,
      "grad_norm": 8.709218978881836,
      "learning_rate": 3.9637422537203165e-06,
      "loss": 0.4729,
      "mean_token_accuracy": 0.8495317488908768,
      "num_tokens": 158077808.0,
      "step": 131410
    },
    {
      "entropy": 1.9698147997260094,
      "epoch": 0.4073905014030979,
      "grad_norm": 8.470252990722656,
      "learning_rate": 3.963591445326578e-06,
      "loss": 0.5221,
      "mean_token_accuracy": 0.8412455469369888,
      "num_tokens": 158089356.0,
      "step": 131420
    },
    {
      "entropy": 1.7146087184548378,
      "epoch": 0.4074215005281476,
      "grad_norm": 3.0368010997772217,
      "learning_rate": 3.963440654144938e-06,
      "loss": 0.3414,
      "mean_token_accuracy": 0.8623756617307663,
      "num_tokens": 158104069.0,
      "step": 131430
    },
    {
      "entropy": 1.9334975138306618,
      "epoch": 0.4074524996531973,
      "grad_norm": 7.6624884605407715,
      "learning_rate": 3.963289880172123e-06,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8561296299099922,
      "num_tokens": 158115199.0,
      "step": 131440
    },
    {
      "entropy": 1.9267579466104507,
      "epoch": 0.40748349877824697,
      "grad_norm": 7.29675817489624,
      "learning_rate": 3.96313912340486e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8549336135387421,
      "num_tokens": 158127014.0,
      "step": 131450
    },
    {
      "entropy": 1.8517323270440103,
      "epoch": 0.4075144979032967,
      "grad_norm": Infinity,
      "learning_rate": 3.962988383839877e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8570884570479393,
      "num_tokens": 158139840.0,
      "step": 131460
    },
    {
      "entropy": 1.9198008626699448,
      "epoch": 0.40754549702834636,
      "grad_norm": 9.076776504516602,
      "learning_rate": 3.962837661473903e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8622155413031578,
      "num_tokens": 158150997.0,
      "step": 131470
    },
    {
      "entropy": 1.9445372179150582,
      "epoch": 0.4075764961533961,
      "grad_norm": 9.01667308807373,
      "learning_rate": 3.962686956303667e-06,
      "loss": 0.5047,
      "mean_token_accuracy": 0.8443640530109405,
      "num_tokens": 158162446.0,
      "step": 131480
    },
    {
      "entropy": 1.8989515826106071,
      "epoch": 0.40760749527844575,
      "grad_norm": 4.06826639175415,
      "learning_rate": 3.9625362683259e-06,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8507105484604836,
      "num_tokens": 158174917.0,
      "step": 131490
    },
    {
      "entropy": 1.8795487105846405,
      "epoch": 0.4076384944034955,
      "grad_norm": 4.378306865692139,
      "learning_rate": 3.962385597537333e-06,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8498224586248397,
      "num_tokens": 158187150.0,
      "step": 131500
    },
    {
      "entropy": 1.709404329955578,
      "epoch": 0.40766949352854515,
      "grad_norm": 7.759102821350098,
      "learning_rate": 3.9622349439346985e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8657671257853508,
      "num_tokens": 158201223.0,
      "step": 131510
    },
    {
      "entropy": 1.9634048074483872,
      "epoch": 0.40770049265359487,
      "grad_norm": 9.048603057861328,
      "learning_rate": 3.962084307514729e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8406730964779854,
      "num_tokens": 158212564.0,
      "step": 131520
    },
    {
      "entropy": 1.9156591862440109,
      "epoch": 0.40773149177864454,
      "grad_norm": 8.055933952331543,
      "learning_rate": 3.9619336882741595e-06,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8517060205340385,
      "num_tokens": 158224398.0,
      "step": 131530
    },
    {
      "entropy": 1.851639135181904,
      "epoch": 0.40776249090369426,
      "grad_norm": 4.349390029907227,
      "learning_rate": 3.961783086209726e-06,
      "loss": 0.4029,
      "mean_token_accuracy": 0.8623872861266136,
      "num_tokens": 158237555.0,
      "step": 131540
    },
    {
      "entropy": 1.8899920910596848,
      "epoch": 0.40779349002874393,
      "grad_norm": 6.555295467376709,
      "learning_rate": 3.96163250131816e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8546891719102859,
      "num_tokens": 158249575.0,
      "step": 131550
    },
    {
      "entropy": 1.8582070901989938,
      "epoch": 0.40782448915379366,
      "grad_norm": 8.720921516418457,
      "learning_rate": 3.961481933596203e-06,
      "loss": 0.4457,
      "mean_token_accuracy": 0.8513376504182816,
      "num_tokens": 158262561.0,
      "step": 131560
    },
    {
      "entropy": 1.9029300913214684,
      "epoch": 0.4078554882788433,
      "grad_norm": 7.594869136810303,
      "learning_rate": 3.9613313830405895e-06,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8554488077759743,
      "num_tokens": 158274086.0,
      "step": 131570
    },
    {
      "entropy": 1.8785912677645684,
      "epoch": 0.40788648740389305,
      "grad_norm": 4.259986400604248,
      "learning_rate": 3.961180849648059e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8466733306646347,
      "num_tokens": 158286621.0,
      "step": 131580
    },
    {
      "entropy": 1.8832870185375215,
      "epoch": 0.4079174865289427,
      "grad_norm": 3.7143633365631104,
      "learning_rate": 3.961030333415349e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8520120859146119,
      "num_tokens": 158298950.0,
      "step": 131590
    },
    {
      "entropy": 1.8503879860043526,
      "epoch": 0.40794848565399244,
      "grad_norm": 6.5647478103637695,
      "learning_rate": 3.960879834339202e-06,
      "loss": 0.3874,
      "mean_token_accuracy": 0.8644073367118835,
      "num_tokens": 158311634.0,
      "step": 131600
    },
    {
      "entropy": 1.785201308131218,
      "epoch": 0.4079794847790421,
      "grad_norm": 8.026589393615723,
      "learning_rate": 3.960729352416358e-06,
      "loss": 0.4128,
      "mean_token_accuracy": 0.849951197206974,
      "num_tokens": 158325117.0,
      "step": 131610
    },
    {
      "entropy": 1.9552603572607041,
      "epoch": 0.4080104839040918,
      "grad_norm": 8.554306983947754,
      "learning_rate": 3.960578887643557e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.846280574798584,
      "num_tokens": 158336274.0,
      "step": 131620
    },
    {
      "entropy": 1.8943400636315346,
      "epoch": 0.4080414830291415,
      "grad_norm": 9.436079978942871,
      "learning_rate": 3.960428440017544e-06,
      "loss": 0.4796,
      "mean_token_accuracy": 0.8490863978862763,
      "num_tokens": 158347481.0,
      "step": 131630
    },
    {
      "entropy": 1.8362084731459618,
      "epoch": 0.4080724821541912,
      "grad_norm": 4.264366626739502,
      "learning_rate": 3.960278009535063e-06,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8494585782289505,
      "num_tokens": 158360076.0,
      "step": 131640
    },
    {
      "entropy": 1.9207329094409942,
      "epoch": 0.4081034812792409,
      "grad_norm": 8.190292358398438,
      "learning_rate": 3.960127596192855e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8595984593033791,
      "num_tokens": 158371437.0,
      "step": 131650
    },
    {
      "entropy": 1.9442640289664268,
      "epoch": 0.40813448040429057,
      "grad_norm": 8.884259223937988,
      "learning_rate": 3.959977199987669e-06,
      "loss": 0.4553,
      "mean_token_accuracy": 0.8574367508292198,
      "num_tokens": 158382706.0,
      "step": 131660
    },
    {
      "entropy": 1.8724360197782517,
      "epoch": 0.4081654795293403,
      "grad_norm": 4.065852642059326,
      "learning_rate": 3.959826820916251e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8485496342182159,
      "num_tokens": 158394833.0,
      "step": 131670
    },
    {
      "entropy": 1.9313586950302124,
      "epoch": 0.40819647865438996,
      "grad_norm": 4.444271087646484,
      "learning_rate": 3.9596764589753435e-06,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8469718486070633,
      "num_tokens": 158406591.0,
      "step": 131680
    },
    {
      "entropy": 1.9100508004426957,
      "epoch": 0.4082274777794397,
      "grad_norm": 4.082368850708008,
      "learning_rate": 3.959526114161699e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8586499407887459,
      "num_tokens": 158418411.0,
      "step": 131690
    },
    {
      "entropy": 1.8683311700820924,
      "epoch": 0.40825847690448935,
      "grad_norm": 3.7615411281585693,
      "learning_rate": 3.959375786472065e-06,
      "loss": 0.3954,
      "mean_token_accuracy": 0.8578609451651573,
      "num_tokens": 158430806.0,
      "step": 131700
    },
    {
      "entropy": 1.902983796596527,
      "epoch": 0.4082894760295391,
      "grad_norm": 6.945703983306885,
      "learning_rate": 3.95922547590319e-06,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8584466874599457,
      "num_tokens": 158442686.0,
      "step": 131710
    },
    {
      "entropy": 1.8431276768445968,
      "epoch": 0.40832047515458875,
      "grad_norm": 8.36220932006836,
      "learning_rate": 3.959075182451826e-06,
      "loss": 0.4039,
      "mean_token_accuracy": 0.8638790413737297,
      "num_tokens": 158455043.0,
      "step": 131720
    },
    {
      "entropy": 1.8704046532511711,
      "epoch": 0.40835147427963847,
      "grad_norm": 8.115221977233887,
      "learning_rate": 3.958924906114722e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8541171133518219,
      "num_tokens": 158467164.0,
      "step": 131730
    },
    {
      "entropy": 1.89015693962574,
      "epoch": 0.40838247340468814,
      "grad_norm": 8.641586303710938,
      "learning_rate": 3.958774646888633e-06,
      "loss": 0.5154,
      "mean_token_accuracy": 0.8427566275000572,
      "num_tokens": 158479113.0,
      "step": 131740
    },
    {
      "entropy": 1.917370368540287,
      "epoch": 0.40841347252973786,
      "grad_norm": 8.092529296875,
      "learning_rate": 3.958624404770311e-06,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8460395202040673,
      "num_tokens": 158490501.0,
      "step": 131750
    },
    {
      "entropy": 1.8727830082178116,
      "epoch": 0.40844447165478753,
      "grad_norm": 4.2208404541015625,
      "learning_rate": 3.95847417975651e-06,
      "loss": 0.4159,
      "mean_token_accuracy": 0.854328741133213,
      "num_tokens": 158502344.0,
      "step": 131760
    },
    {
      "entropy": 1.8874020084738732,
      "epoch": 0.40847547077983726,
      "grad_norm": 8.900503158569336,
      "learning_rate": 3.958323971843983e-06,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8399735167622566,
      "num_tokens": 158514713.0,
      "step": 131770
    },
    {
      "entropy": 1.9528047412633895,
      "epoch": 0.4085064699048869,
      "grad_norm": 6.548738956451416,
      "learning_rate": 3.958173781029487e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8562960177659988,
      "num_tokens": 158526164.0,
      "step": 131780
    },
    {
      "entropy": 1.8597876712679864,
      "epoch": 0.40853746902993665,
      "grad_norm": 4.265164375305176,
      "learning_rate": 3.95802360730978e-06,
      "loss": 0.3839,
      "mean_token_accuracy": 0.861919678747654,
      "num_tokens": 158538663.0,
      "step": 131790
    },
    {
      "entropy": 1.874061642587185,
      "epoch": 0.4085684681549863,
      "grad_norm": 3.2929224967956543,
      "learning_rate": 3.957873450681617e-06,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8565281003713607,
      "num_tokens": 158550773.0,
      "step": 131800
    },
    {
      "entropy": 1.943674847483635,
      "epoch": 0.40859946728003604,
      "grad_norm": 9.038628578186035,
      "learning_rate": 3.9577233111417575e-06,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8586441352963448,
      "num_tokens": 158562087.0,
      "step": 131810
    },
    {
      "entropy": 1.88618975430727,
      "epoch": 0.4086304664050857,
      "grad_norm": 8.164410591125488,
      "learning_rate": 3.95757318868696e-06,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8438228338956832,
      "num_tokens": 158573774.0,
      "step": 131820
    },
    {
      "entropy": 1.9306119337677956,
      "epoch": 0.40866146553013544,
      "grad_norm": 9.125570297241211,
      "learning_rate": 3.957423083313984e-06,
      "loss": 0.4908,
      "mean_token_accuracy": 0.839359101653099,
      "num_tokens": 158585857.0,
      "step": 131830
    },
    {
      "entropy": 1.9744738072156907,
      "epoch": 0.4086924646551851,
      "grad_norm": 8.555429458618164,
      "learning_rate": 3.957272995019592e-06,
      "loss": 0.5472,
      "mean_token_accuracy": 0.8404243916273118,
      "num_tokens": 158597937.0,
      "step": 131840
    },
    {
      "entropy": 1.9038583174347878,
      "epoch": 0.40872346378023483,
      "grad_norm": 2.9242498874664307,
      "learning_rate": 3.9571229238005436e-06,
      "loss": 0.4247,
      "mean_token_accuracy": 0.8589347749948502,
      "num_tokens": 158610210.0,
      "step": 131850
    },
    {
      "entropy": 1.9494069427251817,
      "epoch": 0.4087544629052845,
      "grad_norm": 3.915905237197876,
      "learning_rate": 3.956972869653602e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8493471398949624,
      "num_tokens": 158621519.0,
      "step": 131860
    },
    {
      "entropy": 1.8807632595300674,
      "epoch": 0.40878546203033417,
      "grad_norm": 7.95693302154541,
      "learning_rate": 3.956822832575532e-06,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8504085868597031,
      "num_tokens": 158633877.0,
      "step": 131870
    },
    {
      "entropy": 1.971236227452755,
      "epoch": 0.4088164611553839,
      "grad_norm": 7.221638202667236,
      "learning_rate": 3.956672812563096e-06,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8400038599967956,
      "num_tokens": 158645327.0,
      "step": 131880
    },
    {
      "entropy": 1.9444902956485748,
      "epoch": 0.40884746028043356,
      "grad_norm": 8.52728271484375,
      "learning_rate": 3.956522809613061e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8471096143126488,
      "num_tokens": 158657499.0,
      "step": 131890
    },
    {
      "entropy": 1.8855510473251342,
      "epoch": 0.4088784594054833,
      "grad_norm": 6.880878448486328,
      "learning_rate": 3.95637282372219e-06,
      "loss": 0.4621,
      "mean_token_accuracy": 0.856865806877613,
      "num_tokens": 158669407.0,
      "step": 131900
    },
    {
      "entropy": 1.9185251086950301,
      "epoch": 0.40890945853053295,
      "grad_norm": 8.515263557434082,
      "learning_rate": 3.956222854887252e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8487884759902954,
      "num_tokens": 158681430.0,
      "step": 131910
    },
    {
      "entropy": 1.8058866739273072,
      "epoch": 0.4089404576555827,
      "grad_norm": 3.930187463760376,
      "learning_rate": 3.956072903105014e-06,
      "loss": 0.3995,
      "mean_token_accuracy": 0.8495476976037025,
      "num_tokens": 158695298.0,
      "step": 131920
    },
    {
      "entropy": 1.9411882251501082,
      "epoch": 0.40897145678063235,
      "grad_norm": 6.896012306213379,
      "learning_rate": 3.955922968372246e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8471732020378113,
      "num_tokens": 158706166.0,
      "step": 131930
    },
    {
      "entropy": 1.9487490490078927,
      "epoch": 0.40900245590568207,
      "grad_norm": 7.6823015213012695,
      "learning_rate": 3.955773050685715e-06,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8591760620474815,
      "num_tokens": 158717557.0,
      "step": 131940
    },
    {
      "entropy": 1.855629739165306,
      "epoch": 0.40903345503073174,
      "grad_norm": 7.680805683135986,
      "learning_rate": 3.955623150042193e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8519311249256134,
      "num_tokens": 158730328.0,
      "step": 131950
    },
    {
      "entropy": 1.9119280502200127,
      "epoch": 0.40906445415578147,
      "grad_norm": 7.97050666809082,
      "learning_rate": 3.9554732664384495e-06,
      "loss": 0.4739,
      "mean_token_accuracy": 0.8517009600996971,
      "num_tokens": 158742681.0,
      "step": 131960
    },
    {
      "entropy": 1.8887418761849404,
      "epoch": 0.40909545328083113,
      "grad_norm": 4.2701802253723145,
      "learning_rate": 3.955323399871258e-06,
      "loss": 0.4534,
      "mean_token_accuracy": 0.8548057928681374,
      "num_tokens": 158754792.0,
      "step": 131970
    },
    {
      "entropy": 1.955525654554367,
      "epoch": 0.40912645240588086,
      "grad_norm": 8.67664623260498,
      "learning_rate": 3.955173550337391e-06,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8389379262924195,
      "num_tokens": 158766595.0,
      "step": 131980
    },
    {
      "entropy": 1.9655676484107971,
      "epoch": 0.4091574515309305,
      "grad_norm": 7.483652591705322,
      "learning_rate": 3.955023717833621e-06,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8467064067721367,
      "num_tokens": 158777624.0,
      "step": 131990
    },
    {
      "entropy": 1.8954958975315095,
      "epoch": 0.40918845065598025,
      "grad_norm": 8.60129165649414,
      "learning_rate": 3.954873902356724e-06,
      "loss": 0.4291,
      "mean_token_accuracy": 0.8588397979736329,
      "num_tokens": 158789808.0,
      "step": 132000
    },
    {
      "entropy": 1.9237147703766824,
      "epoch": 0.4092194497810299,
      "grad_norm": 7.778072357177734,
      "learning_rate": 3.9547241039034745e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8559079930186272,
      "num_tokens": 158801021.0,
      "step": 132010
    },
    {
      "entropy": 1.9441052988171577,
      "epoch": 0.40925044890607964,
      "grad_norm": 8.119583129882812,
      "learning_rate": 3.954574322470649e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8437977254390716,
      "num_tokens": 158812670.0,
      "step": 132020
    },
    {
      "entropy": 1.9320714622735977,
      "epoch": 0.4092814480311293,
      "grad_norm": 7.461652755737305,
      "learning_rate": 3.954424558055025e-06,
      "loss": 0.568,
      "mean_token_accuracy": 0.842699658870697,
      "num_tokens": 158823864.0,
      "step": 132030
    },
    {
      "entropy": 1.9182468011975289,
      "epoch": 0.40931244715617904,
      "grad_norm": 7.672018051147461,
      "learning_rate": 3.954274810653379e-06,
      "loss": 0.4464,
      "mean_token_accuracy": 0.8570101261138916,
      "num_tokens": 158835728.0,
      "step": 132040
    },
    {
      "entropy": 1.9373611778020858,
      "epoch": 0.4093434462812287,
      "grad_norm": 8.900008201599121,
      "learning_rate": 3.954125080262492e-06,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8497121125459671,
      "num_tokens": 158846813.0,
      "step": 132050
    },
    {
      "entropy": 1.8692992970347404,
      "epoch": 0.40937444540627843,
      "grad_norm": 7.517188549041748,
      "learning_rate": 3.953975366879141e-06,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8590711817145348,
      "num_tokens": 158858240.0,
      "step": 132060
    },
    {
      "entropy": 1.8589076712727546,
      "epoch": 0.4094054445313281,
      "grad_norm": 9.378162384033203,
      "learning_rate": 3.953825670500109e-06,
      "loss": 0.4061,
      "mean_token_accuracy": 0.861855249106884,
      "num_tokens": 158870488.0,
      "step": 132070
    },
    {
      "entropy": 1.9007334470748902,
      "epoch": 0.4094364436563778,
      "grad_norm": 8.594630241394043,
      "learning_rate": 3.953675991122176e-06,
      "loss": 0.5,
      "mean_token_accuracy": 0.8480967715382576,
      "num_tokens": 158882023.0,
      "step": 132080
    },
    {
      "entropy": 1.9315433949232101,
      "epoch": 0.4094674427814275,
      "grad_norm": 3.873143196105957,
      "learning_rate": 3.953526328742123e-06,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8507410883903503,
      "num_tokens": 158893185.0,
      "step": 132090
    },
    {
      "entropy": 1.9150833919644357,
      "epoch": 0.4094984419064772,
      "grad_norm": 3.768460273742676,
      "learning_rate": 3.953376683356738e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8455236718058586,
      "num_tokens": 158904520.0,
      "step": 132100
    },
    {
      "entropy": 1.8527635991573335,
      "epoch": 0.4095294410315269,
      "grad_norm": 8.320259094238281,
      "learning_rate": 3.953227054962798e-06,
      "loss": 0.3745,
      "mean_token_accuracy": 0.8663365572690964,
      "num_tokens": 158917450.0,
      "step": 132110
    },
    {
      "entropy": 1.8949743598699569,
      "epoch": 0.40956044015657656,
      "grad_norm": 3.7683377265930176,
      "learning_rate": 3.953077443557093e-06,
      "loss": 0.5338,
      "mean_token_accuracy": 0.8330464109778404,
      "num_tokens": 158929795.0,
      "step": 132120
    },
    {
      "entropy": 1.9238438218832017,
      "epoch": 0.4095914392816263,
      "grad_norm": 8.981724739074707,
      "learning_rate": 3.952927849136406e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8446515381336213,
      "num_tokens": 158941331.0,
      "step": 132130
    },
    {
      "entropy": 1.9227786138653755,
      "epoch": 0.40962243840667595,
      "grad_norm": 7.715068817138672,
      "learning_rate": 3.952778271697524e-06,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8444668635725975,
      "num_tokens": 158953028.0,
      "step": 132140
    },
    {
      "entropy": 1.9023156434297561,
      "epoch": 0.4096534375317257,
      "grad_norm": 8.244248390197754,
      "learning_rate": 3.952628711237235e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.850898091495037,
      "num_tokens": 158964269.0,
      "step": 132150
    },
    {
      "entropy": 1.8805926099419594,
      "epoch": 0.40968443665677534,
      "grad_norm": 7.6930999755859375,
      "learning_rate": 3.952479167752328e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.863546060025692,
      "num_tokens": 158976310.0,
      "step": 132160
    },
    {
      "entropy": 1.7959831669926642,
      "epoch": 0.40971543578182507,
      "grad_norm": 5.07602596282959,
      "learning_rate": 3.952329641239589e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8648563221096992,
      "num_tokens": 158989932.0,
      "step": 132170
    },
    {
      "entropy": 1.9297270089387895,
      "epoch": 0.40974643490687473,
      "grad_norm": 7.6279096603393555,
      "learning_rate": 3.9521801316958105e-06,
      "loss": 0.4611,
      "mean_token_accuracy": 0.8553159907460213,
      "num_tokens": 159001128.0,
      "step": 132180
    },
    {
      "entropy": 1.8945249140262603,
      "epoch": 0.40977743403192446,
      "grad_norm": 7.788156032562256,
      "learning_rate": 3.952030639117782e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8605994209647179,
      "num_tokens": 159012619.0,
      "step": 132190
    },
    {
      "entropy": 1.9229301661252975,
      "epoch": 0.40980843315697413,
      "grad_norm": 9.636686325073242,
      "learning_rate": 3.951881163502295e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8494541838765144,
      "num_tokens": 159023540.0,
      "step": 132200
    },
    {
      "entropy": 1.8150003015995027,
      "epoch": 0.40983943228202385,
      "grad_norm": 8.98440933227539,
      "learning_rate": 3.951731704846143e-06,
      "loss": 0.3751,
      "mean_token_accuracy": 0.8665079712867737,
      "num_tokens": 159036264.0,
      "step": 132210
    },
    {
      "entropy": 1.9748639404773711,
      "epoch": 0.4098704314070735,
      "grad_norm": 10.900146484375,
      "learning_rate": 3.951582263146119e-06,
      "loss": 0.5727,
      "mean_token_accuracy": 0.8254036799073219,
      "num_tokens": 159046762.0,
      "step": 132220
    },
    {
      "entropy": 1.8873142629861832,
      "epoch": 0.40990143053212325,
      "grad_norm": 9.89717960357666,
      "learning_rate": 3.951432838399017e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8460609570145607,
      "num_tokens": 159058574.0,
      "step": 132230
    },
    {
      "entropy": 1.8014112293720246,
      "epoch": 0.4099324296571729,
      "grad_norm": 8.128996849060059,
      "learning_rate": 3.95128343060163e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8637454330921173,
      "num_tokens": 159072392.0,
      "step": 132240
    },
    {
      "entropy": 1.8816646337509155,
      "epoch": 0.40996342878222264,
      "grad_norm": 2.441240072250366,
      "learning_rate": 3.9511340397507555e-06,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8551323056221009,
      "num_tokens": 159084913.0,
      "step": 132250
    },
    {
      "entropy": 1.8489706605672835,
      "epoch": 0.4099944279072723,
      "grad_norm": 4.065392971038818,
      "learning_rate": 3.950984665843191e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8606104895472526,
      "num_tokens": 159097664.0,
      "step": 132260
    },
    {
      "entropy": 1.8708397269248962,
      "epoch": 0.41002542703232203,
      "grad_norm": 4.442748069763184,
      "learning_rate": 3.950835308875733e-06,
      "loss": 0.4184,
      "mean_token_accuracy": 0.8600472211837769,
      "num_tokens": 159109240.0,
      "step": 132270
    },
    {
      "entropy": 1.8376126080751418,
      "epoch": 0.4100564261573717,
      "grad_norm": 7.40627908706665,
      "learning_rate": 3.9506859688451805e-06,
      "loss": 0.4016,
      "mean_token_accuracy": 0.8633526057004929,
      "num_tokens": 159121773.0,
      "step": 132280
    },
    {
      "entropy": 1.9250367239117623,
      "epoch": 0.4100874252824214,
      "grad_norm": 7.536020755767822,
      "learning_rate": 3.950536645748332e-06,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8445648729801178,
      "num_tokens": 159133238.0,
      "step": 132290
    },
    {
      "entropy": 1.8859627723693848,
      "epoch": 0.4101184244074711,
      "grad_norm": 10.275099754333496,
      "learning_rate": 3.950387339581987e-06,
      "loss": 0.4357,
      "mean_token_accuracy": 0.8571148261427879,
      "num_tokens": 159144881.0,
      "step": 132300
    },
    {
      "entropy": 1.8043443158268928,
      "epoch": 0.4101494235325208,
      "grad_norm": 6.609226703643799,
      "learning_rate": 3.950238050342948e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8644479662179947,
      "num_tokens": 159157919.0,
      "step": 132310
    },
    {
      "entropy": 1.9193567991256715,
      "epoch": 0.4101804226575705,
      "grad_norm": 8.553771018981934,
      "learning_rate": 3.950088778028016e-06,
      "loss": 0.5014,
      "mean_token_accuracy": 0.8534865245223046,
      "num_tokens": 159169473.0,
      "step": 132320
    },
    {
      "entropy": 1.9148662850260734,
      "epoch": 0.4102114217826202,
      "grad_norm": 5.254544258117676,
      "learning_rate": 3.949939522633992e-06,
      "loss": 0.49,
      "mean_token_accuracy": 0.8425288826227189,
      "num_tokens": 159181088.0,
      "step": 132330
    },
    {
      "entropy": 1.8751920118927956,
      "epoch": 0.4102424209076699,
      "grad_norm": 8.47884750366211,
      "learning_rate": 3.9497902841576826e-06,
      "loss": 0.4212,
      "mean_token_accuracy": 0.8611443415284157,
      "num_tokens": 159192624.0,
      "step": 132340
    },
    {
      "entropy": 1.9114878982305528,
      "epoch": 0.41027342003271955,
      "grad_norm": 8.459266662597656,
      "learning_rate": 3.949641062595889e-06,
      "loss": 0.4893,
      "mean_token_accuracy": 0.8466789424419403,
      "num_tokens": 159203363.0,
      "step": 132350
    },
    {
      "entropy": 1.9088921546936035,
      "epoch": 0.4103044191577693,
      "grad_norm": 7.341428756713867,
      "learning_rate": 3.949491857945419e-06,
      "loss": 0.4507,
      "mean_token_accuracy": 0.8548379436135292,
      "num_tokens": 159215263.0,
      "step": 132360
    },
    {
      "entropy": 1.8097696974873543,
      "epoch": 0.41033541828281894,
      "grad_norm": 8.788670539855957,
      "learning_rate": 3.949342670203077e-06,
      "loss": 0.4085,
      "mean_token_accuracy": 0.8639630541205406,
      "num_tokens": 159227890.0,
      "step": 132370
    },
    {
      "entropy": 1.921763353049755,
      "epoch": 0.41036641740786867,
      "grad_norm": 8.037771224975586,
      "learning_rate": 3.94919349936567e-06,
      "loss": 0.5278,
      "mean_token_accuracy": 0.8411315530538559,
      "num_tokens": 159239371.0,
      "step": 132380
    },
    {
      "entropy": 1.9325927823781968,
      "epoch": 0.41039741653291834,
      "grad_norm": 9.67178726196289,
      "learning_rate": 3.949044345430004e-06,
      "loss": 0.4994,
      "mean_token_accuracy": 0.8497194886207581,
      "num_tokens": 159249796.0,
      "step": 132390
    },
    {
      "entropy": 1.8051844477653503,
      "epoch": 0.41042841565796806,
      "grad_norm": 5.3989691734313965,
      "learning_rate": 3.94889520839289e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8578594639897347,
      "num_tokens": 159262693.0,
      "step": 132400
    },
    {
      "entropy": 1.9161362245678901,
      "epoch": 0.41045941478301773,
      "grad_norm": 8.300895690917969,
      "learning_rate": 3.948746088251138e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.853403514623642,
      "num_tokens": 159274712.0,
      "step": 132410
    },
    {
      "entropy": 1.8416450381278993,
      "epoch": 0.41049041390806745,
      "grad_norm": 6.615303039550781,
      "learning_rate": 3.948596985001556e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8720286816358567,
      "num_tokens": 159287211.0,
      "step": 132420
    },
    {
      "entropy": 1.935446311533451,
      "epoch": 0.4105214130331171,
      "grad_norm": 8.131443977355957,
      "learning_rate": 3.948447898640955e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8410095199942589,
      "num_tokens": 159298375.0,
      "step": 132430
    },
    {
      "entropy": 1.8042462676763535,
      "epoch": 0.41055241215816685,
      "grad_norm": 6.6756911277771,
      "learning_rate": 3.948298829166149e-06,
      "loss": 0.4102,
      "mean_token_accuracy": 0.8577197238802909,
      "num_tokens": 159311942.0,
      "step": 132440
    },
    {
      "entropy": 1.8535395577549933,
      "epoch": 0.4105834112832165,
      "grad_norm": 8.319269180297852,
      "learning_rate": 3.948149776573948e-06,
      "loss": 0.5231,
      "mean_token_accuracy": 0.841600701212883,
      "num_tokens": 159325565.0,
      "step": 132450
    },
    {
      "entropy": 1.853706520795822,
      "epoch": 0.41061441040826624,
      "grad_norm": 7.631629467010498,
      "learning_rate": 3.948000740861168e-06,
      "loss": 0.4186,
      "mean_token_accuracy": 0.8647748947143554,
      "num_tokens": 159337661.0,
      "step": 132460
    },
    {
      "entropy": 1.8631751976907254,
      "epoch": 0.4106454095333159,
      "grad_norm": 3.5813517570495605,
      "learning_rate": 3.947851722024622e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8632732674479484,
      "num_tokens": 159349920.0,
      "step": 132470
    },
    {
      "entropy": 1.9363053843379021,
      "epoch": 0.41067640865836563,
      "grad_norm": 7.310238361358643,
      "learning_rate": 3.947702720061125e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8465033307671547,
      "num_tokens": 159361594.0,
      "step": 132480
    },
    {
      "entropy": 1.8684279143810272,
      "epoch": 0.4107074077834153,
      "grad_norm": 9.23641300201416,
      "learning_rate": 3.947553734967494e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8531864553689956,
      "num_tokens": 159373909.0,
      "step": 132490
    },
    {
      "entropy": 1.9086994245648383,
      "epoch": 0.410738406908465,
      "grad_norm": 9.247127532958984,
      "learning_rate": 3.947404766740546e-06,
      "loss": 0.4944,
      "mean_token_accuracy": 0.8483163312077522,
      "num_tokens": 159386422.0,
      "step": 132500
    },
    {
      "entropy": 1.9499953478574752,
      "epoch": 0.4107694060335147,
      "grad_norm": 7.84163236618042,
      "learning_rate": 3.947255815377098e-06,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8528471052646637,
      "num_tokens": 159397202.0,
      "step": 132510
    },
    {
      "entropy": 1.8840179577469827,
      "epoch": 0.4108004051585644,
      "grad_norm": 4.282788276672363,
      "learning_rate": 3.947106880873969e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8478341817855835,
      "num_tokens": 159410046.0,
      "step": 132520
    },
    {
      "entropy": 1.7982640989124774,
      "epoch": 0.4108314042836141,
      "grad_norm": 4.207269191741943,
      "learning_rate": 3.946957963227978e-06,
      "loss": 0.3782,
      "mean_token_accuracy": 0.8666106954216957,
      "num_tokens": 159422809.0,
      "step": 132530
    },
    {
      "entropy": 1.8728843182325363,
      "epoch": 0.4108624034086638,
      "grad_norm": 4.20776891708374,
      "learning_rate": 3.946809062435946e-06,
      "loss": 0.4235,
      "mean_token_accuracy": 0.8539015829563141,
      "num_tokens": 159435471.0,
      "step": 132540
    },
    {
      "entropy": 1.8800769940018653,
      "epoch": 0.4108934025337135,
      "grad_norm": 8.50875473022461,
      "learning_rate": 3.9466601784946935e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8496467992663383,
      "num_tokens": 159448258.0,
      "step": 132550
    },
    {
      "entropy": 1.824411703646183,
      "epoch": 0.4109244016587632,
      "grad_norm": 7.829589366912842,
      "learning_rate": 3.946511311401043e-06,
      "loss": 0.3826,
      "mean_token_accuracy": 0.8696093156933784,
      "num_tokens": 159460857.0,
      "step": 132560
    },
    {
      "entropy": 1.8826001703739166,
      "epoch": 0.4109554007838129,
      "grad_norm": 8.163921356201172,
      "learning_rate": 3.946362461151816e-06,
      "loss": 0.4364,
      "mean_token_accuracy": 0.8533704668283463,
      "num_tokens": 159473481.0,
      "step": 132570
    },
    {
      "entropy": 1.841157278418541,
      "epoch": 0.4109863999088626,
      "grad_norm": 8.525252342224121,
      "learning_rate": 3.946213627743839e-06,
      "loss": 0.3758,
      "mean_token_accuracy": 0.8639968425035477,
      "num_tokens": 159485282.0,
      "step": 132580
    },
    {
      "entropy": 1.903964453935623,
      "epoch": 0.41101739903391227,
      "grad_norm": 8.637951850891113,
      "learning_rate": 3.9460648111739346e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8393364146351814,
      "num_tokens": 159496784.0,
      "step": 132590
    },
    {
      "entropy": 1.8756361320614814,
      "epoch": 0.41104839815896194,
      "grad_norm": 9.585840225219727,
      "learning_rate": 3.945916011438926e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.8499387681484223,
      "num_tokens": 159508357.0,
      "step": 132600
    },
    {
      "entropy": 1.8971466958522796,
      "epoch": 0.41107939728401166,
      "grad_norm": 7.960050106048584,
      "learning_rate": 3.945767228535644e-06,
      "loss": 0.4611,
      "mean_token_accuracy": 0.8593509659171105,
      "num_tokens": 159519968.0,
      "step": 132610
    },
    {
      "entropy": 1.928882573544979,
      "epoch": 0.41111039640906133,
      "grad_norm": 7.798891544342041,
      "learning_rate": 3.9456184624609115e-06,
      "loss": 0.4672,
      "mean_token_accuracy": 0.8493491113185883,
      "num_tokens": 159531286.0,
      "step": 132620
    },
    {
      "entropy": 1.9064328223466873,
      "epoch": 0.41114139553411105,
      "grad_norm": 8.213712692260742,
      "learning_rate": 3.945469713211559e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.844736622273922,
      "num_tokens": 159543308.0,
      "step": 132630
    },
    {
      "entropy": 1.898288056254387,
      "epoch": 0.4111723946591607,
      "grad_norm": 7.8646039962768555,
      "learning_rate": 3.945320980784413e-06,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8415489286184311,
      "num_tokens": 159555394.0,
      "step": 132640
    },
    {
      "entropy": 1.8438436210155487,
      "epoch": 0.41120339378421045,
      "grad_norm": 8.93295955657959,
      "learning_rate": 3.945172265176305e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8616420149803161,
      "num_tokens": 159568585.0,
      "step": 132650
    },
    {
      "entropy": 1.909870770573616,
      "epoch": 0.4112343929092601,
      "grad_norm": 3.4213240146636963,
      "learning_rate": 3.945023566384064e-06,
      "loss": 0.4484,
      "mean_token_accuracy": 0.856622938811779,
      "num_tokens": 159580433.0,
      "step": 132660
    },
    {
      "entropy": 1.9142403885722161,
      "epoch": 0.41126539203430984,
      "grad_norm": 4.372912883758545,
      "learning_rate": 3.944874884404522e-06,
      "loss": 0.4383,
      "mean_token_accuracy": 0.8533108174800873,
      "num_tokens": 159592231.0,
      "step": 132670
    },
    {
      "entropy": 1.9566717997193337,
      "epoch": 0.4112963911593595,
      "grad_norm": 7.756010055541992,
      "learning_rate": 3.944726219234511e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8435320198535919,
      "num_tokens": 159603617.0,
      "step": 132680
    },
    {
      "entropy": 1.7626475676894189,
      "epoch": 0.41132739028440923,
      "grad_norm": 9.682427406311035,
      "learning_rate": 3.944577570870863e-06,
      "loss": 0.3736,
      "mean_token_accuracy": 0.8564101874828338,
      "num_tokens": 159617560.0,
      "step": 132690
    },
    {
      "entropy": 1.9389828056097032,
      "epoch": 0.4113583894094589,
      "grad_norm": 7.8840179443359375,
      "learning_rate": 3.944428939310412e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8546505197882652,
      "num_tokens": 159628901.0,
      "step": 132700
    },
    {
      "entropy": 1.7881031468510629,
      "epoch": 0.4113893885345086,
      "grad_norm": 3.9923946857452393,
      "learning_rate": 3.944280324549993e-06,
      "loss": 0.3477,
      "mean_token_accuracy": 0.8698825523257255,
      "num_tokens": 159642790.0,
      "step": 132710
    },
    {
      "entropy": 1.9210022300481797,
      "epoch": 0.4114203876595583,
      "grad_norm": 6.981124401092529,
      "learning_rate": 3.944131726586441e-06,
      "loss": 0.4662,
      "mean_token_accuracy": 0.842510312795639,
      "num_tokens": 159654565.0,
      "step": 132720
    },
    {
      "entropy": 1.7577592477202415,
      "epoch": 0.411451386784608,
      "grad_norm": 7.697649002075195,
      "learning_rate": 3.943983145416592e-06,
      "loss": 0.3523,
      "mean_token_accuracy": 0.8635148763656616,
      "num_tokens": 159668041.0,
      "step": 132730
    },
    {
      "entropy": 1.9331092268228531,
      "epoch": 0.4114823859096577,
      "grad_norm": 7.486753463745117,
      "learning_rate": 3.943834581037284e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8566341996192932,
      "num_tokens": 159679025.0,
      "step": 132740
    },
    {
      "entropy": 1.8456541523337364,
      "epoch": 0.4115133850347074,
      "grad_norm": 8.025050163269043,
      "learning_rate": 3.9436860334453535e-06,
      "loss": 0.3865,
      "mean_token_accuracy": 0.8562669888138771,
      "num_tokens": 159691580.0,
      "step": 132750
    },
    {
      "entropy": 1.8954646542668343,
      "epoch": 0.4115443841597571,
      "grad_norm": 8.532896995544434,
      "learning_rate": 3.94353750263764e-06,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8576274499297142,
      "num_tokens": 159702748.0,
      "step": 132760
    },
    {
      "entropy": 1.8293071150779725,
      "epoch": 0.4115753832848068,
      "grad_norm": 3.7488701343536377,
      "learning_rate": 3.943388988610982e-06,
      "loss": 0.3922,
      "mean_token_accuracy": 0.8620394229888916,
      "num_tokens": 159715325.0,
      "step": 132770
    },
    {
      "entropy": 1.8163642302155494,
      "epoch": 0.4116063824098565,
      "grad_norm": 11.977995872497559,
      "learning_rate": 3.943240491362222e-06,
      "loss": 0.4271,
      "mean_token_accuracy": 0.8522105798125267,
      "num_tokens": 159728453.0,
      "step": 132780
    },
    {
      "entropy": 1.8184501871466636,
      "epoch": 0.4116373815349062,
      "grad_norm": 7.925955772399902,
      "learning_rate": 3.9430920108882e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8594678580760956,
      "num_tokens": 159741671.0,
      "step": 132790
    },
    {
      "entropy": 1.8316495031118394,
      "epoch": 0.41166838065995587,
      "grad_norm": 8.928302764892578,
      "learning_rate": 3.942943547185757e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.86448425501585,
      "num_tokens": 159754290.0,
      "step": 132800
    },
    {
      "entropy": 1.8046483889222145,
      "epoch": 0.4116993797850056,
      "grad_norm": 3.8450210094451904,
      "learning_rate": 3.942795100251737e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.8553688034415246,
      "num_tokens": 159766765.0,
      "step": 132810
    },
    {
      "entropy": 1.888925837725401,
      "epoch": 0.41173037891005526,
      "grad_norm": 9.118193626403809,
      "learning_rate": 3.942646670082983e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8464773207902908,
      "num_tokens": 159779056.0,
      "step": 132820
    },
    {
      "entropy": 1.900699371099472,
      "epoch": 0.411761378035105,
      "grad_norm": 7.870625972747803,
      "learning_rate": 3.942498256676342e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8564991891384125,
      "num_tokens": 159790237.0,
      "step": 132830
    },
    {
      "entropy": 1.9068083673715592,
      "epoch": 0.41179237716015465,
      "grad_norm": 9.596370697021484,
      "learning_rate": 3.942349860028655e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8504379868507386,
      "num_tokens": 159802324.0,
      "step": 132840
    },
    {
      "entropy": 1.95052922219038,
      "epoch": 0.4118233762852043,
      "grad_norm": 7.9511494636535645,
      "learning_rate": 3.942201480136772e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8542814716696739,
      "num_tokens": 159813411.0,
      "step": 132850
    },
    {
      "entropy": 1.9478548653423786,
      "epoch": 0.41185437541025405,
      "grad_norm": 7.628299713134766,
      "learning_rate": 3.942053116997537e-06,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8497453570365906,
      "num_tokens": 159824950.0,
      "step": 132860
    },
    {
      "entropy": 1.9837032228708267,
      "epoch": 0.4118853745353037,
      "grad_norm": 8.014751434326172,
      "learning_rate": 3.941904770607801e-06,
      "loss": 0.5192,
      "mean_token_accuracy": 0.8386216104030609,
      "num_tokens": 159836448.0,
      "step": 132870
    },
    {
      "entropy": 1.881002813577652,
      "epoch": 0.41191637366035344,
      "grad_norm": 7.482250213623047,
      "learning_rate": 3.94175644096441e-06,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8465662628412247,
      "num_tokens": 159847845.0,
      "step": 132880
    },
    {
      "entropy": 1.8887635886669158,
      "epoch": 0.4119473727854031,
      "grad_norm": 3.9271512031555176,
      "learning_rate": 3.941608128064214e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8607404991984368,
      "num_tokens": 159860022.0,
      "step": 132890
    },
    {
      "entropy": 1.9076422840356826,
      "epoch": 0.41197837191045283,
      "grad_norm": 4.137758255004883,
      "learning_rate": 3.9414598319040645e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8553347662091255,
      "num_tokens": 159872269.0,
      "step": 132900
    },
    {
      "entropy": 1.8967055141925813,
      "epoch": 0.4120093710355025,
      "grad_norm": 6.800586700439453,
      "learning_rate": 3.941311552480813e-06,
      "loss": 0.4575,
      "mean_token_accuracy": 0.8462317153811455,
      "num_tokens": 159884036.0,
      "step": 132910
    },
    {
      "entropy": 1.9517858356237412,
      "epoch": 0.4120403701605522,
      "grad_norm": 8.87108039855957,
      "learning_rate": 3.941163289791309e-06,
      "loss": 0.4335,
      "mean_token_accuracy": 0.8600587204098702,
      "num_tokens": 159896130.0,
      "step": 132920
    },
    {
      "entropy": 1.9215604767203331,
      "epoch": 0.4120713692856019,
      "grad_norm": 8.849936485290527,
      "learning_rate": 3.941015043832408e-06,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8451809346675873,
      "num_tokens": 159908040.0,
      "step": 132930
    },
    {
      "entropy": 1.9565720595419407,
      "epoch": 0.4121023684106516,
      "grad_norm": 7.92255973815918,
      "learning_rate": 3.940866814600961e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8439367547631264,
      "num_tokens": 159919326.0,
      "step": 132940
    },
    {
      "entropy": 1.9593399614095688,
      "epoch": 0.4121333675357013,
      "grad_norm": 7.588318824768066,
      "learning_rate": 3.940718602093826e-06,
      "loss": 0.4888,
      "mean_token_accuracy": 0.850217179954052,
      "num_tokens": 159930913.0,
      "step": 132950
    },
    {
      "entropy": 1.9544413655996322,
      "epoch": 0.412164366660751,
      "grad_norm": 8.658806800842285,
      "learning_rate": 3.9405704063078546e-06,
      "loss": 0.5432,
      "mean_token_accuracy": 0.841646321117878,
      "num_tokens": 159942302.0,
      "step": 132960
    },
    {
      "entropy": 1.9482817143201827,
      "epoch": 0.4121953657858007,
      "grad_norm": 9.174203872680664,
      "learning_rate": 3.940422227239905e-06,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8407966732978821,
      "num_tokens": 159953706.0,
      "step": 132970
    },
    {
      "entropy": 1.8826787516474723,
      "epoch": 0.4122263649108504,
      "grad_norm": 10.95289421081543,
      "learning_rate": 3.9402740648868335e-06,
      "loss": 0.4539,
      "mean_token_accuracy": 0.84808798879385,
      "num_tokens": 159965894.0,
      "step": 132980
    },
    {
      "entropy": 1.8592276245355606,
      "epoch": 0.4122573640359001,
      "grad_norm": 9.670906066894531,
      "learning_rate": 3.940125919245498e-06,
      "loss": 0.4063,
      "mean_token_accuracy": 0.8619933322072029,
      "num_tokens": 159978522.0,
      "step": 132990
    },
    {
      "entropy": 1.9317376971244813,
      "epoch": 0.4122883631609498,
      "grad_norm": 7.894320011138916,
      "learning_rate": 3.939977790312759e-06,
      "loss": 0.4869,
      "mean_token_accuracy": 0.8445814654231072,
      "num_tokens": 159989620.0,
      "step": 133000
    },
    {
      "entropy": 2.002483421564102,
      "epoch": 0.41231936228599947,
      "grad_norm": 9.11700439453125,
      "learning_rate": 3.939829678085473e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8350201919674873,
      "num_tokens": 160001293.0,
      "step": 133010
    },
    {
      "entropy": 1.914002077281475,
      "epoch": 0.4123503614110492,
      "grad_norm": 8.952300071716309,
      "learning_rate": 3.939681582560501e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8454202443361283,
      "num_tokens": 160013020.0,
      "step": 133020
    },
    {
      "entropy": 1.9585654482245445,
      "epoch": 0.41238136053609886,
      "grad_norm": 7.388771057128906,
      "learning_rate": 3.939533503734705e-06,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8444287464022636,
      "num_tokens": 160024299.0,
      "step": 133030
    },
    {
      "entropy": 1.880546286702156,
      "epoch": 0.4124123596611486,
      "grad_norm": 3.908165454864502,
      "learning_rate": 3.939385441604947e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.859189136326313,
      "num_tokens": 160036416.0,
      "step": 133040
    },
    {
      "entropy": 1.8767894744873046,
      "epoch": 0.41244335878619826,
      "grad_norm": 7.4038472175598145,
      "learning_rate": 3.939237396168088e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8633315816521645,
      "num_tokens": 160048717.0,
      "step": 133050
    },
    {
      "entropy": 1.9026233911514283,
      "epoch": 0.412474357911248,
      "grad_norm": 8.312832832336426,
      "learning_rate": 3.939089367420993e-06,
      "loss": 0.4479,
      "mean_token_accuracy": 0.852534967660904,
      "num_tokens": 160060661.0,
      "step": 133060
    },
    {
      "entropy": 1.8906838029623032,
      "epoch": 0.41250535703629765,
      "grad_norm": 7.901620864868164,
      "learning_rate": 3.938941355360527e-06,
      "loss": 0.4,
      "mean_token_accuracy": 0.8649371221661568,
      "num_tokens": 160073055.0,
      "step": 133070
    },
    {
      "entropy": 1.9295494481921196,
      "epoch": 0.4125363561613474,
      "grad_norm": 7.450974941253662,
      "learning_rate": 3.938793359983554e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.8515502735972404,
      "num_tokens": 160084739.0,
      "step": 133080
    },
    {
      "entropy": 1.8649974435567855,
      "epoch": 0.41256735528639704,
      "grad_norm": 3.9373440742492676,
      "learning_rate": 3.93864538128694e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8602189511060715,
      "num_tokens": 160097401.0,
      "step": 133090
    },
    {
      "entropy": 1.88185980245471,
      "epoch": 0.4125983544114467,
      "grad_norm": 8.116096496582031,
      "learning_rate": 3.938497419267553e-06,
      "loss": 0.369,
      "mean_token_accuracy": 0.8672322928905487,
      "num_tokens": 160109518.0,
      "step": 133100
    },
    {
      "entropy": 1.9526149734854699,
      "epoch": 0.41262935353649643,
      "grad_norm": 10.256512641906738,
      "learning_rate": 3.938349473922259e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8465583026409149,
      "num_tokens": 160120718.0,
      "step": 133110
    },
    {
      "entropy": 1.8862829342484475,
      "epoch": 0.4126603526615461,
      "grad_norm": 7.902219295501709,
      "learning_rate": 3.938201545247929e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8557077690958976,
      "num_tokens": 160133234.0,
      "step": 133120
    },
    {
      "entropy": 1.9045727148652076,
      "epoch": 0.41269135178659583,
      "grad_norm": 8.710145950317383,
      "learning_rate": 3.93805363324143e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8461684197187423,
      "num_tokens": 160145332.0,
      "step": 133130
    },
    {
      "entropy": 1.8291308224201202,
      "epoch": 0.4127223509116455,
      "grad_norm": 9.70020866394043,
      "learning_rate": 3.937905737899633e-06,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8510018140077591,
      "num_tokens": 160158147.0,
      "step": 133140
    },
    {
      "entropy": 1.8897850081324576,
      "epoch": 0.4127533500366952,
      "grad_norm": 8.668664932250977,
      "learning_rate": 3.937757859219409e-06,
      "loss": 0.4277,
      "mean_token_accuracy": 0.8535225465893745,
      "num_tokens": 160169771.0,
      "step": 133150
    },
    {
      "entropy": 1.9326857790350913,
      "epoch": 0.4127843491617449,
      "grad_norm": 10.146557807922363,
      "learning_rate": 3.93760999719763e-06,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8501642525196076,
      "num_tokens": 160180380.0,
      "step": 133160
    },
    {
      "entropy": 1.8924714922904968,
      "epoch": 0.4128153482867946,
      "grad_norm": 8.729730606079102,
      "learning_rate": 3.937462151831168e-06,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8599536821246148,
      "num_tokens": 160192793.0,
      "step": 133170
    },
    {
      "entropy": 1.8484364092350005,
      "epoch": 0.4128463474118443,
      "grad_norm": 7.995754718780518,
      "learning_rate": 3.937314323116897e-06,
      "loss": 0.3844,
      "mean_token_accuracy": 0.8631056323647499,
      "num_tokens": 160204883.0,
      "step": 133180
    },
    {
      "entropy": 1.8777808651328087,
      "epoch": 0.412877346536894,
      "grad_norm": 9.567119598388672,
      "learning_rate": 3.937166511051691e-06,
      "loss": 0.4602,
      "mean_token_accuracy": 0.8523745030164719,
      "num_tokens": 160217041.0,
      "step": 133190
    },
    {
      "entropy": 1.8994059637188911,
      "epoch": 0.4129083456619437,
      "grad_norm": 7.305065155029297,
      "learning_rate": 3.937018715632426e-06,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8568759083747863,
      "num_tokens": 160229204.0,
      "step": 133200
    },
    {
      "entropy": 1.8381987407803535,
      "epoch": 0.4129393447869934,
      "grad_norm": 7.50701379776001,
      "learning_rate": 3.936870936855977e-06,
      "loss": 0.4089,
      "mean_token_accuracy": 0.8586773455142975,
      "num_tokens": 160242271.0,
      "step": 133210
    },
    {
      "entropy": 1.9254462212324142,
      "epoch": 0.41297034391204307,
      "grad_norm": 9.528829574584961,
      "learning_rate": 3.93672317471922e-06,
      "loss": 0.4837,
      "mean_token_accuracy": 0.8490356847643852,
      "num_tokens": 160253561.0,
      "step": 133220
    },
    {
      "entropy": 1.9055677652359009,
      "epoch": 0.4130013430370928,
      "grad_norm": 7.700775146484375,
      "learning_rate": 3.9365754292190345e-06,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8636551991105079,
      "num_tokens": 160264851.0,
      "step": 133230
    },
    {
      "entropy": 1.9295484229922295,
      "epoch": 0.41303234216214246,
      "grad_norm": 8.420736312866211,
      "learning_rate": 3.936427700352297e-06,
      "loss": 0.438,
      "mean_token_accuracy": 0.8591367319226265,
      "num_tokens": 160276017.0,
      "step": 133240
    },
    {
      "entropy": 1.9808582752943038,
      "epoch": 0.4130633412871922,
      "grad_norm": 6.451159954071045,
      "learning_rate": 3.936279988115888e-06,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8498171672224999,
      "num_tokens": 160286922.0,
      "step": 133250
    },
    {
      "entropy": 1.8775882482528687,
      "epoch": 0.41309434041224186,
      "grad_norm": 8.08848762512207,
      "learning_rate": 3.936132292506687e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8411854594945908,
      "num_tokens": 160299889.0,
      "step": 133260
    },
    {
      "entropy": 1.9185838371515274,
      "epoch": 0.4131253395372916,
      "grad_norm": 8.169017791748047,
      "learning_rate": 3.935984613521574e-06,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8455637603998184,
      "num_tokens": 160311587.0,
      "step": 133270
    },
    {
      "entropy": 1.927768488228321,
      "epoch": 0.41315633866234125,
      "grad_norm": 8.6692476272583,
      "learning_rate": 3.9358369511574325e-06,
      "loss": 0.4657,
      "mean_token_accuracy": 0.8541040673851967,
      "num_tokens": 160322859.0,
      "step": 133280
    },
    {
      "entropy": 1.980207970738411,
      "epoch": 0.413187337787391,
      "grad_norm": 9.063236236572266,
      "learning_rate": 3.935689305411144e-06,
      "loss": 0.5183,
      "mean_token_accuracy": 0.8438897222280503,
      "num_tokens": 160333949.0,
      "step": 133290
    },
    {
      "entropy": 1.8348941326141357,
      "epoch": 0.41321833691244064,
      "grad_norm": 11.424354553222656,
      "learning_rate": 3.935541676279592e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.857143734395504,
      "num_tokens": 160346373.0,
      "step": 133300
    },
    {
      "entropy": 1.9375634253025056,
      "epoch": 0.41324933603749037,
      "grad_norm": 7.8653435707092285,
      "learning_rate": 3.93539406375966e-06,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8492831483483314,
      "num_tokens": 160357451.0,
      "step": 133310
    },
    {
      "entropy": 1.8438090533018112,
      "epoch": 0.41328033516254004,
      "grad_norm": 8.466445922851562,
      "learning_rate": 3.9352464678482325e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8501735106110573,
      "num_tokens": 160370838.0,
      "step": 133320
    },
    {
      "entropy": 1.9386936947703362,
      "epoch": 0.41331133428758976,
      "grad_norm": 8.103391647338867,
      "learning_rate": 3.935098888542198e-06,
      "loss": 0.4446,
      "mean_token_accuracy": 0.8613961502909661,
      "num_tokens": 160382175.0,
      "step": 133330
    },
    {
      "entropy": 1.8318694144487382,
      "epoch": 0.41334233341263943,
      "grad_norm": 8.630756378173828,
      "learning_rate": 3.934951325838439e-06,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8484767213463783,
      "num_tokens": 160395178.0,
      "step": 133340
    },
    {
      "entropy": 1.9260124281048774,
      "epoch": 0.4133733325376891,
      "grad_norm": 7.492244720458984,
      "learning_rate": 3.934803779733846e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8476378932595253,
      "num_tokens": 160407014.0,
      "step": 133350
    },
    {
      "entropy": 1.8849865958094596,
      "epoch": 0.4134043316627388,
      "grad_norm": 3.519080400466919,
      "learning_rate": 3.934656250225307e-06,
      "loss": 0.4288,
      "mean_token_accuracy": 0.8558978796005249,
      "num_tokens": 160419398.0,
      "step": 133360
    },
    {
      "entropy": 1.8688034132122993,
      "epoch": 0.4134353307877885,
      "grad_norm": 6.8715033531188965,
      "learning_rate": 3.934508737309709e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8502852901816368,
      "num_tokens": 160431982.0,
      "step": 133370
    },
    {
      "entropy": 1.8360135570168494,
      "epoch": 0.4134663299128382,
      "grad_norm": 6.848766326904297,
      "learning_rate": 3.934361240983944e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8653775498270988,
      "num_tokens": 160444385.0,
      "step": 133380
    },
    {
      "entropy": 1.9299522519111634,
      "epoch": 0.4134973290378879,
      "grad_norm": 4.2813496589660645,
      "learning_rate": 3.934213761244901e-06,
      "loss": 0.5009,
      "mean_token_accuracy": 0.8370076134800911,
      "num_tokens": 160456540.0,
      "step": 133390
    },
    {
      "entropy": 1.8286241069436073,
      "epoch": 0.4135283281629376,
      "grad_norm": 8.64360237121582,
      "learning_rate": 3.934066298089472e-06,
      "loss": 0.4333,
      "mean_token_accuracy": 0.8586912482976914,
      "num_tokens": 160469565.0,
      "step": 133400
    },
    {
      "entropy": 1.9575915843248368,
      "epoch": 0.4135593272879873,
      "grad_norm": 5.0459370613098145,
      "learning_rate": 3.93391885151455e-06,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8582988694310189,
      "num_tokens": 160480566.0,
      "step": 133410
    },
    {
      "entropy": 1.9278161972761154,
      "epoch": 0.413590326413037,
      "grad_norm": 7.705987930297852,
      "learning_rate": 3.933771421517027e-06,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8494553253054619,
      "num_tokens": 160492405.0,
      "step": 133420
    },
    {
      "entropy": 1.9024669751524925,
      "epoch": 0.41362132553808667,
      "grad_norm": 7.893537998199463,
      "learning_rate": 3.9336240080937985e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8514198035001754,
      "num_tokens": 160504029.0,
      "step": 133430
    },
    {
      "entropy": 1.830211453139782,
      "epoch": 0.4136523246631364,
      "grad_norm": 5.1870222091674805,
      "learning_rate": 3.933476611241757e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.85890783816576,
      "num_tokens": 160516627.0,
      "step": 133440
    },
    {
      "entropy": 1.8257996559143066,
      "epoch": 0.41368332378818606,
      "grad_norm": 3.705263376235962,
      "learning_rate": 3.9333292309578e-06,
      "loss": 0.4155,
      "mean_token_accuracy": 0.8598002254962921,
      "num_tokens": 160530211.0,
      "step": 133450
    },
    {
      "entropy": 1.9126117467880248,
      "epoch": 0.4137143229132358,
      "grad_norm": 4.153369903564453,
      "learning_rate": 3.933181867238822e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8488003075122833,
      "num_tokens": 160541810.0,
      "step": 133460
    },
    {
      "entropy": 1.9465106219053268,
      "epoch": 0.41374532203828546,
      "grad_norm": 8.59892463684082,
      "learning_rate": 3.933034520081723e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8590655192732811,
      "num_tokens": 160552736.0,
      "step": 133470
    },
    {
      "entropy": 1.9508576810359954,
      "epoch": 0.4137763211633352,
      "grad_norm": 8.499300003051758,
      "learning_rate": 3.9328871894833975e-06,
      "loss": 0.4967,
      "mean_token_accuracy": 0.8416604056954384,
      "num_tokens": 160564088.0,
      "step": 133480
    },
    {
      "entropy": 1.9252055808901787,
      "epoch": 0.41380732028838485,
      "grad_norm": 8.029823303222656,
      "learning_rate": 3.932739875440747e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8455364406108856,
      "num_tokens": 160575490.0,
      "step": 133490
    },
    {
      "entropy": 1.8071866802871228,
      "epoch": 0.4138383194134346,
      "grad_norm": 8.960620880126953,
      "learning_rate": 3.93259257795067e-06,
      "loss": 0.4126,
      "mean_token_accuracy": 0.8656040906906128,
      "num_tokens": 160588680.0,
      "step": 133500
    },
    {
      "entropy": 1.877330508828163,
      "epoch": 0.41386931853848424,
      "grad_norm": 8.935325622558594,
      "learning_rate": 3.932445297010065e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8558456063270569,
      "num_tokens": 160600946.0,
      "step": 133510
    },
    {
      "entropy": 1.9224560901522636,
      "epoch": 0.41390031766353397,
      "grad_norm": 7.983498573303223,
      "learning_rate": 3.932298032615838e-06,
      "loss": 0.4876,
      "mean_token_accuracy": 0.8463954851031303,
      "num_tokens": 160612788.0,
      "step": 133520
    },
    {
      "entropy": 1.8346376448869706,
      "epoch": 0.41393131678858364,
      "grad_norm": 8.071921348571777,
      "learning_rate": 3.932150784764887e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8672112032771111,
      "num_tokens": 160625526.0,
      "step": 133530
    },
    {
      "entropy": 1.931402738392353,
      "epoch": 0.41396231591363336,
      "grad_norm": 3.571261405944824,
      "learning_rate": 3.932003553454117e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8496660724282264,
      "num_tokens": 160637476.0,
      "step": 133540
    },
    {
      "entropy": 1.9502887561917306,
      "epoch": 0.41399331503868303,
      "grad_norm": 7.512312412261963,
      "learning_rate": 3.93185633868043e-06,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8570262104272842,
      "num_tokens": 160648694.0,
      "step": 133550
    },
    {
      "entropy": 1.9070443853735923,
      "epoch": 0.41402431416373275,
      "grad_norm": 8.191246032714844,
      "learning_rate": 3.931709140440732e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8478936731815339,
      "num_tokens": 160660627.0,
      "step": 133560
    },
    {
      "entropy": 1.9022893160581589,
      "epoch": 0.4140553132887824,
      "grad_norm": 7.333972454071045,
      "learning_rate": 3.931561958731927e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8530973836779594,
      "num_tokens": 160672648.0,
      "step": 133570
    },
    {
      "entropy": 1.8672500133514405,
      "epoch": 0.41408631241383215,
      "grad_norm": 7.397851467132568,
      "learning_rate": 3.931414793550921e-06,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8604272216558456,
      "num_tokens": 160685095.0,
      "step": 133580
    },
    {
      "entropy": 1.911836712062359,
      "epoch": 0.4141173115388818,
      "grad_norm": 9.264906883239746,
      "learning_rate": 3.9312676448946225e-06,
      "loss": 0.4732,
      "mean_token_accuracy": 0.8507508277893067,
      "num_tokens": 160697135.0,
      "step": 133590
    },
    {
      "entropy": 1.892936834692955,
      "epoch": 0.4141483106639315,
      "grad_norm": 8.400348663330078,
      "learning_rate": 3.931120512759939e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8533075451850891,
      "num_tokens": 160708171.0,
      "step": 133600
    },
    {
      "entropy": 1.9460464030504228,
      "epoch": 0.4141793097889812,
      "grad_norm": 7.992114067077637,
      "learning_rate": 3.930973397143777e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.8554316237568855,
      "num_tokens": 160719263.0,
      "step": 133610
    },
    {
      "entropy": 1.8571830540895462,
      "epoch": 0.4142103089140309,
      "grad_norm": 9.004242897033691,
      "learning_rate": 3.930826298043048e-06,
      "loss": 0.3898,
      "mean_token_accuracy": 0.868439619243145,
      "num_tokens": 160731482.0,
      "step": 133620
    },
    {
      "entropy": 1.9615523904561996,
      "epoch": 0.4142413080390806,
      "grad_norm": 7.82743501663208,
      "learning_rate": 3.930679215454661e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8472507506608963,
      "num_tokens": 160741961.0,
      "step": 133630
    },
    {
      "entropy": 1.8198735401034356,
      "epoch": 0.41427230716413027,
      "grad_norm": 5.923557281494141,
      "learning_rate": 3.9305321493755265e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8585488334298134,
      "num_tokens": 160755417.0,
      "step": 133640
    },
    {
      "entropy": 1.8547986596822739,
      "epoch": 0.41430330628918,
      "grad_norm": 9.227656364440918,
      "learning_rate": 3.9303850998025586e-06,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8587556973099708,
      "num_tokens": 160767718.0,
      "step": 133650
    },
    {
      "entropy": 1.8604234397411346,
      "epoch": 0.41433430541422966,
      "grad_norm": 7.426607131958008,
      "learning_rate": 3.930238066732667e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8567753449082375,
      "num_tokens": 160780130.0,
      "step": 133660
    },
    {
      "entropy": 1.9464622527360915,
      "epoch": 0.4143653045392794,
      "grad_norm": 8.832720756530762,
      "learning_rate": 3.930091050162768e-06,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8490612745285034,
      "num_tokens": 160790549.0,
      "step": 133670
    },
    {
      "entropy": 1.738626065850258,
      "epoch": 0.41439630366432906,
      "grad_norm": 8.568602561950684,
      "learning_rate": 3.9299440500897725e-06,
      "loss": 0.3268,
      "mean_token_accuracy": 0.8719424903392792,
      "num_tokens": 160804288.0,
      "step": 133680
    },
    {
      "entropy": 1.8175345674157142,
      "epoch": 0.4144273027893788,
      "grad_norm": 3.655850410461426,
      "learning_rate": 3.929797066510598e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.8615160465240479,
      "num_tokens": 160816494.0,
      "step": 133690
    },
    {
      "entropy": 1.9714103803038596,
      "epoch": 0.41445830191442845,
      "grad_norm": 3.3211865425109863,
      "learning_rate": 3.92965009942216e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8397148564457894,
      "num_tokens": 160827569.0,
      "step": 133700
    },
    {
      "entropy": 1.8605849146842957,
      "epoch": 0.4144893010394782,
      "grad_norm": 4.2350921630859375,
      "learning_rate": 3.929503148821374e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.8473623186349869,
      "num_tokens": 160840031.0,
      "step": 133710
    },
    {
      "entropy": 1.8704538196325302,
      "epoch": 0.41452030016452784,
      "grad_norm": 4.739353656768799,
      "learning_rate": 3.929356214705158e-06,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8485003530979156,
      "num_tokens": 160851772.0,
      "step": 133720
    },
    {
      "entropy": 1.8266652062535287,
      "epoch": 0.41455129928957757,
      "grad_norm": 8.387866020202637,
      "learning_rate": 3.92920929707043e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8575938984751701,
      "num_tokens": 160864401.0,
      "step": 133730
    },
    {
      "entropy": 1.8190097853541374,
      "epoch": 0.41458229841462724,
      "grad_norm": 2.620790719985962,
      "learning_rate": 3.92906239591411e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.8680200964212418,
      "num_tokens": 160877175.0,
      "step": 133740
    },
    {
      "entropy": 1.9274157121777535,
      "epoch": 0.41461329753967696,
      "grad_norm": 8.17121696472168,
      "learning_rate": 3.928915511233117e-06,
      "loss": 0.481,
      "mean_token_accuracy": 0.8486859872937202,
      "num_tokens": 160889187.0,
      "step": 133750
    },
    {
      "entropy": 1.921906155347824,
      "epoch": 0.41464429666472663,
      "grad_norm": 8.626646041870117,
      "learning_rate": 3.928768643024372e-06,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8409358784556389,
      "num_tokens": 160899766.0,
      "step": 133760
    },
    {
      "entropy": 1.8961732387542725,
      "epoch": 0.41467529578977635,
      "grad_norm": 7.050107955932617,
      "learning_rate": 3.928621791284796e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.8544517681002617,
      "num_tokens": 160910866.0,
      "step": 133770
    },
    {
      "entropy": 1.9284267202019691,
      "epoch": 0.414706294914826,
      "grad_norm": 4.148862361907959,
      "learning_rate": 3.928474956011312e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8424947530031204,
      "num_tokens": 160922592.0,
      "step": 133780
    },
    {
      "entropy": 1.881701409816742,
      "epoch": 0.41473729403987575,
      "grad_norm": 6.7715229988098145,
      "learning_rate": 3.928328137200842e-06,
      "loss": 0.475,
      "mean_token_accuracy": 0.8563426598906517,
      "num_tokens": 160934482.0,
      "step": 133790
    },
    {
      "entropy": 1.7748121902346612,
      "epoch": 0.4147682931649254,
      "grad_norm": 9.0849609375,
      "learning_rate": 3.92818133485031e-06,
      "loss": 0.3697,
      "mean_token_accuracy": 0.870860530436039,
      "num_tokens": 160947645.0,
      "step": 133800
    },
    {
      "entropy": 1.9011081397533416,
      "epoch": 0.41479929228997514,
      "grad_norm": 8.376455307006836,
      "learning_rate": 3.928034548956642e-06,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8554087609052659,
      "num_tokens": 160959336.0,
      "step": 133810
    },
    {
      "entropy": 1.937306745350361,
      "epoch": 0.4148302914150248,
      "grad_norm": 9.280498504638672,
      "learning_rate": 3.927887779516763e-06,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8521208077669143,
      "num_tokens": 160970927.0,
      "step": 133820
    },
    {
      "entropy": 1.8931998923420905,
      "epoch": 0.4148612905400745,
      "grad_norm": 4.04230260848999,
      "learning_rate": 3.927741026527598e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8493711724877357,
      "num_tokens": 160982582.0,
      "step": 133830
    },
    {
      "entropy": 1.862531739473343,
      "epoch": 0.4148922896651242,
      "grad_norm": 8.157683372497559,
      "learning_rate": 3.927594289986076e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.849953505396843,
      "num_tokens": 160994545.0,
      "step": 133840
    },
    {
      "entropy": 1.8835390016436577,
      "epoch": 0.41492328879017387,
      "grad_norm": 8.619124412536621,
      "learning_rate": 3.927447569889122e-06,
      "loss": 0.4962,
      "mean_token_accuracy": 0.8389230579137802,
      "num_tokens": 161007106.0,
      "step": 133850
    },
    {
      "entropy": 1.8973518058657646,
      "epoch": 0.4149542879152236,
      "grad_norm": 7.793274402618408,
      "learning_rate": 3.9273008662336685e-06,
      "loss": 0.5054,
      "mean_token_accuracy": 0.8410894647240639,
      "num_tokens": 161018629.0,
      "step": 133860
    },
    {
      "entropy": 1.9573092699050902,
      "epoch": 0.41498528704027327,
      "grad_norm": 7.558224678039551,
      "learning_rate": 3.927154179016643e-06,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8450547724962234,
      "num_tokens": 161029814.0,
      "step": 133870
    },
    {
      "entropy": 1.8571864694356919,
      "epoch": 0.415016286165323,
      "grad_norm": 8.728129386901855,
      "learning_rate": 3.927007508234975e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8491732507944107,
      "num_tokens": 161041934.0,
      "step": 133880
    },
    {
      "entropy": 1.877596764266491,
      "epoch": 0.41504728529037266,
      "grad_norm": 4.276613235473633,
      "learning_rate": 3.9268608538855965e-06,
      "loss": 0.4656,
      "mean_token_accuracy": 0.8503337040543556,
      "num_tokens": 161053951.0,
      "step": 133890
    },
    {
      "entropy": 1.8341388761997224,
      "epoch": 0.4150782844154224,
      "grad_norm": 4.777545928955078,
      "learning_rate": 3.92671421596544e-06,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8406883478164673,
      "num_tokens": 161067201.0,
      "step": 133900
    },
    {
      "entropy": 1.8835119605064392,
      "epoch": 0.41510928354047205,
      "grad_norm": 7.586372375488281,
      "learning_rate": 3.926567594471437e-06,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8480971708893776,
      "num_tokens": 161079821.0,
      "step": 133910
    },
    {
      "entropy": 1.817841324210167,
      "epoch": 0.4151402826655218,
      "grad_norm": 8.434521675109863,
      "learning_rate": 3.926420989400522e-06,
      "loss": 0.4007,
      "mean_token_accuracy": 0.8662325829267502,
      "num_tokens": 161092583.0,
      "step": 133920
    },
    {
      "entropy": 1.9291996493935586,
      "epoch": 0.41517128179057144,
      "grad_norm": 8.39786434173584,
      "learning_rate": 3.926274400749629e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8544291496276856,
      "num_tokens": 161103847.0,
      "step": 133930
    },
    {
      "entropy": 1.8668764278292656,
      "epoch": 0.41520228091562117,
      "grad_norm": 3.983625650405884,
      "learning_rate": 3.926127828515693e-06,
      "loss": 0.417,
      "mean_token_accuracy": 0.8636351123452186,
      "num_tokens": 161115696.0,
      "step": 133940
    },
    {
      "entropy": 1.8500849679112434,
      "epoch": 0.41523328004067084,
      "grad_norm": 11.573062896728516,
      "learning_rate": 3.9259812726956495e-06,
      "loss": 0.4025,
      "mean_token_accuracy": 0.8647431343793869,
      "num_tokens": 161127966.0,
      "step": 133950
    },
    {
      "entropy": 1.8055422961711884,
      "epoch": 0.41526427916572056,
      "grad_norm": 2.1831018924713135,
      "learning_rate": 3.925834733286436e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8400851160287857,
      "num_tokens": 161141745.0,
      "step": 133960
    },
    {
      "entropy": 1.8568698644638062,
      "epoch": 0.41529527829077023,
      "grad_norm": 7.404766082763672,
      "learning_rate": 3.92568821028499e-06,
      "loss": 0.4149,
      "mean_token_accuracy": 0.8579725489020348,
      "num_tokens": 161153901.0,
      "step": 133970
    },
    {
      "entropy": 1.9167630136013032,
      "epoch": 0.41532627741581996,
      "grad_norm": 7.15609884262085,
      "learning_rate": 3.925541703688249e-06,
      "loss": 0.4677,
      "mean_token_accuracy": 0.8486376807093621,
      "num_tokens": 161165538.0,
      "step": 133980
    },
    {
      "entropy": 1.8915724635124207,
      "epoch": 0.4153572765408696,
      "grad_norm": 7.963951587677002,
      "learning_rate": 3.925395213493153e-06,
      "loss": 0.4811,
      "mean_token_accuracy": 0.8474412530660629,
      "num_tokens": 161178272.0,
      "step": 133990
    },
    {
      "entropy": 1.9221931904554368,
      "epoch": 0.41538827566591935,
      "grad_norm": 7.879276752471924,
      "learning_rate": 3.9252487396966406e-06,
      "loss": 0.4792,
      "mean_token_accuracy": 0.8496794193983078,
      "num_tokens": 161189021.0,
      "step": 134000
    },
    {
      "entropy": 1.8921538457274436,
      "epoch": 0.415419274790969,
      "grad_norm": 8.086331367492676,
      "learning_rate": 3.925102282295654e-06,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8499061942100525,
      "num_tokens": 161201131.0,
      "step": 134010
    },
    {
      "entropy": 1.825336940586567,
      "epoch": 0.41545027391601874,
      "grad_norm": 3.680633068084717,
      "learning_rate": 3.924955841287134e-06,
      "loss": 0.4024,
      "mean_token_accuracy": 0.8611365631222725,
      "num_tokens": 161213630.0,
      "step": 134020
    },
    {
      "entropy": 1.8338982105255126,
      "epoch": 0.4154812730410684,
      "grad_norm": 9.535050392150879,
      "learning_rate": 3.924809416668023e-06,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8613386616110802,
      "num_tokens": 161226206.0,
      "step": 134030
    },
    {
      "entropy": 1.8270452991127968,
      "epoch": 0.41551227216611814,
      "grad_norm": 4.551232814788818,
      "learning_rate": 3.924663008435264e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8666022554039955,
      "num_tokens": 161238486.0,
      "step": 134040
    },
    {
      "entropy": 1.9449387162923812,
      "epoch": 0.4155432712911678,
      "grad_norm": 8.532154083251953,
      "learning_rate": 3.924516616585802e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8555340111255646,
      "num_tokens": 161249333.0,
      "step": 134050
    },
    {
      "entropy": 1.9312722623348235,
      "epoch": 0.41557427041621753,
      "grad_norm": 8.082392692565918,
      "learning_rate": 3.9243702411165805e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8388245224952697,
      "num_tokens": 161260572.0,
      "step": 134060
    },
    {
      "entropy": 1.7703733541071416,
      "epoch": 0.4156052695412672,
      "grad_norm": 8.54536247253418,
      "learning_rate": 3.924223882024544e-06,
      "loss": 0.3865,
      "mean_token_accuracy": 0.8662910461425781,
      "num_tokens": 161274273.0,
      "step": 134070
    },
    {
      "entropy": 1.8904980316758155,
      "epoch": 0.41563626866631687,
      "grad_norm": 9.833793640136719,
      "learning_rate": 3.9240775393066405e-06,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8553065106272697,
      "num_tokens": 161285970.0,
      "step": 134080
    },
    {
      "entropy": 1.887739597260952,
      "epoch": 0.4156672677913666,
      "grad_norm": 6.797054767608643,
      "learning_rate": 3.923931212959817e-06,
      "loss": 0.4456,
      "mean_token_accuracy": 0.8516652911901474,
      "num_tokens": 161297693.0,
      "step": 134090
    },
    {
      "entropy": 1.9516427367925644,
      "epoch": 0.41569826691641626,
      "grad_norm": 7.91145658493042,
      "learning_rate": 3.923784902981022e-06,
      "loss": 0.4695,
      "mean_token_accuracy": 0.855071696639061,
      "num_tokens": 161308121.0,
      "step": 134100
    },
    {
      "entropy": 1.9453662782907486,
      "epoch": 0.415729266041466,
      "grad_norm": 7.931567668914795,
      "learning_rate": 3.923638609367202e-06,
      "loss": 0.5049,
      "mean_token_accuracy": 0.8384775072336197,
      "num_tokens": 161319934.0,
      "step": 134110
    },
    {
      "entropy": 1.9178491979837418,
      "epoch": 0.41576026516651565,
      "grad_norm": 8.443523406982422,
      "learning_rate": 3.923492332115307e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8489273399114609,
      "num_tokens": 161330638.0,
      "step": 134120
    },
    {
      "entropy": 1.9393232375383378,
      "epoch": 0.4157912642915654,
      "grad_norm": 6.289927005767822,
      "learning_rate": 3.9233460712222895e-06,
      "loss": 0.5225,
      "mean_token_accuracy": 0.8360320463776588,
      "num_tokens": 161342902.0,
      "step": 134130
    },
    {
      "entropy": 1.9341855004429818,
      "epoch": 0.41582226341661505,
      "grad_norm": 8.161822319030762,
      "learning_rate": 3.923199826685099e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.850852632522583,
      "num_tokens": 161354270.0,
      "step": 134140
    },
    {
      "entropy": 1.8818897396326064,
      "epoch": 0.41585326254166477,
      "grad_norm": 9.941109657287598,
      "learning_rate": 3.9230535985006865e-06,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8523253738880158,
      "num_tokens": 161366256.0,
      "step": 134150
    },
    {
      "entropy": 1.878204520046711,
      "epoch": 0.41588426166671444,
      "grad_norm": 9.152496337890625,
      "learning_rate": 3.9229073866660056e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8530999973416329,
      "num_tokens": 161378072.0,
      "step": 134160
    },
    {
      "entropy": 1.9943836033344269,
      "epoch": 0.41591526079176416,
      "grad_norm": 8.155722618103027,
      "learning_rate": 3.92276119117801e-06,
      "loss": 0.5221,
      "mean_token_accuracy": 0.8379778906702995,
      "num_tokens": 161389288.0,
      "step": 134170
    },
    {
      "entropy": 1.8447602733969688,
      "epoch": 0.41594625991681383,
      "grad_norm": 8.624749183654785,
      "learning_rate": 3.922615012033654e-06,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8538846731185913,
      "num_tokens": 161401938.0,
      "step": 134180
    },
    {
      "entropy": 1.8731971263885498,
      "epoch": 0.41597725904186356,
      "grad_norm": 8.428763389587402,
      "learning_rate": 3.922468849229893e-06,
      "loss": 0.4268,
      "mean_token_accuracy": 0.8539308935403824,
      "num_tokens": 161413429.0,
      "step": 134190
    },
    {
      "entropy": 1.9091638535261155,
      "epoch": 0.4160082581669132,
      "grad_norm": 8.18406867980957,
      "learning_rate": 3.922322702763682e-06,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8525436207652092,
      "num_tokens": 161424862.0,
      "step": 134200
    },
    {
      "entropy": 1.8178348287940025,
      "epoch": 0.41603925729196295,
      "grad_norm": 5.807666778564453,
      "learning_rate": 3.922176572631976e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.8637724488973617,
      "num_tokens": 161437571.0,
      "step": 134210
    },
    {
      "entropy": 1.8581533119082452,
      "epoch": 0.4160702564170126,
      "grad_norm": 5.807939052581787,
      "learning_rate": 3.922030458831736e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8546082183718682,
      "num_tokens": 161450359.0,
      "step": 134220
    },
    {
      "entropy": 1.9059478402137757,
      "epoch": 0.41610125554206234,
      "grad_norm": 8.28518009185791,
      "learning_rate": 3.921884361359918e-06,
      "loss": 0.4905,
      "mean_token_accuracy": 0.8530061066150665,
      "num_tokens": 161461493.0,
      "step": 134230
    },
    {
      "entropy": 1.8699321061372758,
      "epoch": 0.416132254667112,
      "grad_norm": 7.9905195236206055,
      "learning_rate": 3.921738280213482e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8566524386405945,
      "num_tokens": 161473498.0,
      "step": 134240
    },
    {
      "entropy": 1.88876011967659,
      "epoch": 0.41616325379216174,
      "grad_norm": 8.742656707763672,
      "learning_rate": 3.921592215389386e-06,
      "loss": 0.4385,
      "mean_token_accuracy": 0.8500191062688828,
      "num_tokens": 161485605.0,
      "step": 134250
    },
    {
      "entropy": 1.9045383304357528,
      "epoch": 0.4161942529172114,
      "grad_norm": 9.43719482421875,
      "learning_rate": 3.921446166884594e-06,
      "loss": 0.465,
      "mean_token_accuracy": 0.8534532591700554,
      "num_tokens": 161496981.0,
      "step": 134260
    },
    {
      "entropy": 1.9324941888451577,
      "epoch": 0.41622525204226113,
      "grad_norm": 7.1453680992126465,
      "learning_rate": 3.921300134696064e-06,
      "loss": 0.488,
      "mean_token_accuracy": 0.8457109674811363,
      "num_tokens": 161508264.0,
      "step": 134270
    },
    {
      "entropy": 1.8439267560839654,
      "epoch": 0.4162562511673108,
      "grad_norm": 3.715588331222534,
      "learning_rate": 3.921154118820759e-06,
      "loss": 0.43,
      "mean_token_accuracy": 0.8561937183141708,
      "num_tokens": 161521748.0,
      "step": 134280
    },
    {
      "entropy": 1.995403453707695,
      "epoch": 0.4162872502923605,
      "grad_norm": 8.698075294494629,
      "learning_rate": 3.921008119255644e-06,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8444655641913414,
      "num_tokens": 161532336.0,
      "step": 134290
    },
    {
      "entropy": 1.8547970443964004,
      "epoch": 0.4163182494174102,
      "grad_norm": 7.284455299377441,
      "learning_rate": 3.9208621359976816e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8494389861822128,
      "num_tokens": 161545319.0,
      "step": 134300
    },
    {
      "entropy": 1.9544886738061904,
      "epoch": 0.4163492485424599,
      "grad_norm": 8.434039115905762,
      "learning_rate": 3.920716169043834e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.8425055891275406,
      "num_tokens": 161556597.0,
      "step": 134310
    },
    {
      "entropy": 1.8419823780655862,
      "epoch": 0.4163802476675096,
      "grad_norm": 8.092065811157227,
      "learning_rate": 3.9205702183910696e-06,
      "loss": 0.3975,
      "mean_token_accuracy": 0.8516080200672149,
      "num_tokens": 161570185.0,
      "step": 134320
    },
    {
      "entropy": 1.8836676687002183,
      "epoch": 0.41641124679255925,
      "grad_norm": 7.677006721496582,
      "learning_rate": 3.920424284036354e-06,
      "loss": 0.4231,
      "mean_token_accuracy": 0.8599972873926163,
      "num_tokens": 161582993.0,
      "step": 134330
    },
    {
      "entropy": 1.910847471654415,
      "epoch": 0.416442245917609,
      "grad_norm": 10.631965637207031,
      "learning_rate": 3.920278365976653e-06,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8386124089360237,
      "num_tokens": 161594430.0,
      "step": 134340
    },
    {
      "entropy": 1.867373377084732,
      "epoch": 0.41647324504265865,
      "grad_norm": 4.6483259201049805,
      "learning_rate": 3.920132464208936e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8458052203059196,
      "num_tokens": 161607226.0,
      "step": 134350
    },
    {
      "entropy": 1.9246284693479538,
      "epoch": 0.41650424416770837,
      "grad_norm": 7.288817405700684,
      "learning_rate": 3.91998657873017e-06,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8457208350300789,
      "num_tokens": 161618272.0,
      "step": 134360
    },
    {
      "entropy": 1.8862150296568871,
      "epoch": 0.41653524329275804,
      "grad_norm": 5.131035804748535,
      "learning_rate": 3.919840709537325e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8593840152025223,
      "num_tokens": 161630204.0,
      "step": 134370
    },
    {
      "entropy": 1.876247802376747,
      "epoch": 0.41656624241780776,
      "grad_norm": 10.8263578414917,
      "learning_rate": 3.919694856627371e-06,
      "loss": 0.3749,
      "mean_token_accuracy": 0.869242000579834,
      "num_tokens": 161641771.0,
      "step": 134380
    },
    {
      "entropy": 1.9201571598649025,
      "epoch": 0.41659724154285743,
      "grad_norm": 8.346663475036621,
      "learning_rate": 3.919549019997278e-06,
      "loss": 0.4356,
      "mean_token_accuracy": 0.8538030579686164,
      "num_tokens": 161653401.0,
      "step": 134390
    },
    {
      "entropy": 1.8522045940160752,
      "epoch": 0.41662824066790716,
      "grad_norm": 3.982300281524658,
      "learning_rate": 3.919403199644019e-06,
      "loss": 0.4508,
      "mean_token_accuracy": 0.8470421150326729,
      "num_tokens": 161665679.0,
      "step": 134400
    },
    {
      "entropy": 1.8318135902285575,
      "epoch": 0.4166592397929568,
      "grad_norm": 4.39650297164917,
      "learning_rate": 3.919257395564566e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8538282006978989,
      "num_tokens": 161678834.0,
      "step": 134410
    },
    {
      "entropy": 1.8755167797207832,
      "epoch": 0.41669023891800655,
      "grad_norm": 9.182366371154785,
      "learning_rate": 3.919111607755892e-06,
      "loss": 0.4426,
      "mean_token_accuracy": 0.847786869108677,
      "num_tokens": 161691040.0,
      "step": 134420
    },
    {
      "entropy": 1.921954096853733,
      "epoch": 0.4167212380430562,
      "grad_norm": 8.545275688171387,
      "learning_rate": 3.918965836214972e-06,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8338611707091331,
      "num_tokens": 161703955.0,
      "step": 134430
    },
    {
      "entropy": 1.8536706238985061,
      "epoch": 0.41675223716810594,
      "grad_norm": 7.954626560211182,
      "learning_rate": 3.918820080938779e-06,
      "loss": 0.4029,
      "mean_token_accuracy": 0.8524566516280174,
      "num_tokens": 161716705.0,
      "step": 134440
    },
    {
      "entropy": 1.907846173644066,
      "epoch": 0.4167832362931556,
      "grad_norm": 6.290383338928223,
      "learning_rate": 3.91867434192429e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8462353229522706,
      "num_tokens": 161728755.0,
      "step": 134450
    },
    {
      "entropy": 1.9126424625515939,
      "epoch": 0.41681423541820534,
      "grad_norm": 7.351926803588867,
      "learning_rate": 3.918528619168481e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8547870278358459,
      "num_tokens": 161740140.0,
      "step": 134460
    },
    {
      "entropy": 1.7981932133436203,
      "epoch": 0.416845234543255,
      "grad_norm": 10.335866928100586,
      "learning_rate": 3.9183829126683305e-06,
      "loss": 0.3812,
      "mean_token_accuracy": 0.8719327822327614,
      "num_tokens": 161753546.0,
      "step": 134470
    },
    {
      "entropy": 1.9288191676139832,
      "epoch": 0.41687623366830473,
      "grad_norm": 8.642077445983887,
      "learning_rate": 3.918237222420813e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8515138506889344,
      "num_tokens": 161764200.0,
      "step": 134480
    },
    {
      "entropy": 1.8795452415943146,
      "epoch": 0.4169072327933544,
      "grad_norm": 7.725874900817871,
      "learning_rate": 3.918091548422912e-06,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8545924738049507,
      "num_tokens": 161776635.0,
      "step": 134490
    },
    {
      "entropy": 1.9160576537251472,
      "epoch": 0.4169382319184041,
      "grad_norm": 4.713511943817139,
      "learning_rate": 3.9179458906716036e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.854782110452652,
      "num_tokens": 161789062.0,
      "step": 134500
    },
    {
      "entropy": 1.891368383169174,
      "epoch": 0.4169692310434538,
      "grad_norm": 7.320746898651123,
      "learning_rate": 3.917800249163869e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8455578476190567,
      "num_tokens": 161801067.0,
      "step": 134510
    },
    {
      "entropy": 1.9198919370770455,
      "epoch": 0.4170002301685035,
      "grad_norm": 4.034561634063721,
      "learning_rate": 3.917654623896689e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8582830354571342,
      "num_tokens": 161813219.0,
      "step": 134520
    },
    {
      "entropy": 1.8274074256420136,
      "epoch": 0.4170312292935532,
      "grad_norm": 3.6742751598358154,
      "learning_rate": 3.917509014867046e-06,
      "loss": 0.3795,
      "mean_token_accuracy": 0.8572375267744065,
      "num_tokens": 161826737.0,
      "step": 134530
    },
    {
      "entropy": 1.96164468228817,
      "epoch": 0.4170622284186029,
      "grad_norm": 7.9983367919921875,
      "learning_rate": 3.917363422071923e-06,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8470546841621399,
      "num_tokens": 161837950.0,
      "step": 134540
    },
    {
      "entropy": 1.840873746573925,
      "epoch": 0.4170932275436526,
      "grad_norm": 8.481100082397461,
      "learning_rate": 3.917217845508302e-06,
      "loss": 0.4535,
      "mean_token_accuracy": 0.8508628994226456,
      "num_tokens": 161851286.0,
      "step": 134550
    },
    {
      "entropy": 1.8318819746375083,
      "epoch": 0.4171242266687023,
      "grad_norm": 6.554941654205322,
      "learning_rate": 3.917072285173169e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8574759662151337,
      "num_tokens": 161864090.0,
      "step": 134560
    },
    {
      "entropy": 1.8796166345477103,
      "epoch": 0.41715522579375197,
      "grad_norm": 7.1165266036987305,
      "learning_rate": 3.916926741063509e-06,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8539829269051552,
      "num_tokens": 161876656.0,
      "step": 134570
    },
    {
      "entropy": 1.8293213650584221,
      "epoch": 0.41718622491880164,
      "grad_norm": 8.498753547668457,
      "learning_rate": 3.916781213176306e-06,
      "loss": 0.3757,
      "mean_token_accuracy": 0.8652872815728188,
      "num_tokens": 161889042.0,
      "step": 134580
    },
    {
      "entropy": 1.9550884038209915,
      "epoch": 0.41721722404385136,
      "grad_norm": 9.21761417388916,
      "learning_rate": 3.916635701508549e-06,
      "loss": 0.5468,
      "mean_token_accuracy": 0.8403035715222359,
      "num_tokens": 161900992.0,
      "step": 134590
    },
    {
      "entropy": 1.9226959884166717,
      "epoch": 0.41724822316890103,
      "grad_norm": 7.0631842613220215,
      "learning_rate": 3.916490206057224e-06,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8605365335941315,
      "num_tokens": 161912060.0,
      "step": 134600
    },
    {
      "entropy": 1.7682531923055649,
      "epoch": 0.41727922229395076,
      "grad_norm": 8.758580207824707,
      "learning_rate": 3.91634472681932e-06,
      "loss": 0.4329,
      "mean_token_accuracy": 0.8549866124987602,
      "num_tokens": 161926308.0,
      "step": 134610
    },
    {
      "entropy": 1.805106556415558,
      "epoch": 0.4173102214190004,
      "grad_norm": 4.0905070304870605,
      "learning_rate": 3.916199263791824e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.8517909824848175,
      "num_tokens": 161939441.0,
      "step": 134620
    },
    {
      "entropy": 1.7959918841719626,
      "epoch": 0.41734122054405015,
      "grad_norm": 9.999496459960938,
      "learning_rate": 3.916053816971728e-06,
      "loss": 0.3566,
      "mean_token_accuracy": 0.863648708164692,
      "num_tokens": 161952525.0,
      "step": 134630
    },
    {
      "entropy": 1.8762843772768973,
      "epoch": 0.4173722196690998,
      "grad_norm": 3.626180648803711,
      "learning_rate": 3.9159083863560204e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8647943899035454,
      "num_tokens": 161964249.0,
      "step": 134640
    },
    {
      "entropy": 1.9006209999322892,
      "epoch": 0.41740321879414954,
      "grad_norm": 8.766674041748047,
      "learning_rate": 3.915762971941694e-06,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8434425577521324,
      "num_tokens": 161975450.0,
      "step": 134650
    },
    {
      "entropy": 1.959898342192173,
      "epoch": 0.4174342179191992,
      "grad_norm": 7.614290714263916,
      "learning_rate": 3.915617573725742e-06,
      "loss": 0.5172,
      "mean_token_accuracy": 0.847202044725418,
      "num_tokens": 161986492.0,
      "step": 134660
    },
    {
      "entropy": 1.9258347421884536,
      "epoch": 0.41746521704424894,
      "grad_norm": 7.911265850067139,
      "learning_rate": 3.9154721917051546e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.8465585187077522,
      "num_tokens": 161997495.0,
      "step": 134670
    },
    {
      "entropy": 1.8868989422917366,
      "epoch": 0.4174962161692986,
      "grad_norm": 3.8985397815704346,
      "learning_rate": 3.915326825876927e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.8470399528741837,
      "num_tokens": 162009972.0,
      "step": 134680
    },
    {
      "entropy": 1.961172890663147,
      "epoch": 0.41752721529434833,
      "grad_norm": 9.122645378112793,
      "learning_rate": 3.915181476238054e-06,
      "loss": 0.5025,
      "mean_token_accuracy": 0.8451171442866325,
      "num_tokens": 162021139.0,
      "step": 134690
    },
    {
      "entropy": 1.867299085855484,
      "epoch": 0.417558214419398,
      "grad_norm": 7.1753034591674805,
      "learning_rate": 3.91503614278553e-06,
      "loss": 0.4129,
      "mean_token_accuracy": 0.8628281652927399,
      "num_tokens": 162033229.0,
      "step": 134700
    },
    {
      "entropy": 1.8105697840452195,
      "epoch": 0.4175892135444477,
      "grad_norm": 3.7438528537750244,
      "learning_rate": 3.91489082551635e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8544130697846413,
      "num_tokens": 162046064.0,
      "step": 134710
    },
    {
      "entropy": 1.916566787660122,
      "epoch": 0.4176202126694974,
      "grad_norm": 8.405454635620117,
      "learning_rate": 3.914745524427513e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.8509135708212853,
      "num_tokens": 162056998.0,
      "step": 134720
    },
    {
      "entropy": 1.883594536781311,
      "epoch": 0.4176512117945471,
      "grad_norm": 9.41535472869873,
      "learning_rate": 3.914600239516016e-06,
      "loss": 0.4559,
      "mean_token_accuracy": 0.8488730236887931,
      "num_tokens": 162068897.0,
      "step": 134730
    },
    {
      "entropy": 1.8739972546696664,
      "epoch": 0.4176822109195968,
      "grad_norm": 4.211245059967041,
      "learning_rate": 3.9144549707788556e-06,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8516407638788224,
      "num_tokens": 162081030.0,
      "step": 134740
    },
    {
      "entropy": 1.9788037657737731,
      "epoch": 0.4177132100446465,
      "grad_norm": 8.905488967895508,
      "learning_rate": 3.914309718213034e-06,
      "loss": 0.5444,
      "mean_token_accuracy": 0.8368604257702827,
      "num_tokens": 162091716.0,
      "step": 134750
    },
    {
      "entropy": 1.8498676016926765,
      "epoch": 0.4177442091696962,
      "grad_norm": 2.3638017177581787,
      "learning_rate": 3.914164481815549e-06,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8562711879611016,
      "num_tokens": 162103721.0,
      "step": 134760
    },
    {
      "entropy": 1.9256371967494488,
      "epoch": 0.4177752082947459,
      "grad_norm": 8.89646053314209,
      "learning_rate": 3.9140192615833995e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.8388048008084297,
      "num_tokens": 162115431.0,
      "step": 134770
    },
    {
      "entropy": 1.8755139395594598,
      "epoch": 0.41780620741979557,
      "grad_norm": 7.007622718811035,
      "learning_rate": 3.91387405751359e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.8632483765482902,
      "num_tokens": 162127975.0,
      "step": 134780
    },
    {
      "entropy": 1.865597426891327,
      "epoch": 0.4178372065448453,
      "grad_norm": 4.80448579788208,
      "learning_rate": 3.913728869603122e-06,
      "loss": 0.4438,
      "mean_token_accuracy": 0.8586938053369522,
      "num_tokens": 162139430.0,
      "step": 134790
    },
    {
      "entropy": 1.9837687581777572,
      "epoch": 0.41786820566989497,
      "grad_norm": 7.113491058349609,
      "learning_rate": 3.913583697848999e-06,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8430691197514534,
      "num_tokens": 162150117.0,
      "step": 134800
    },
    {
      "entropy": 1.8740553870797156,
      "epoch": 0.4178992047949447,
      "grad_norm": 8.047861099243164,
      "learning_rate": 3.913438542248223e-06,
      "loss": 0.4269,
      "mean_token_accuracy": 0.8576863974332809,
      "num_tokens": 162161930.0,
      "step": 134810
    },
    {
      "entropy": 1.955976441502571,
      "epoch": 0.41793020391999436,
      "grad_norm": 8.8277006149292,
      "learning_rate": 3.913293402797799e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8523634567856788,
      "num_tokens": 162173265.0,
      "step": 134820
    },
    {
      "entropy": 1.9804795682430267,
      "epoch": 0.417961203045044,
      "grad_norm": 7.893826007843018,
      "learning_rate": 3.913148279494734e-06,
      "loss": 0.5503,
      "mean_token_accuracy": 0.8313609048724174,
      "num_tokens": 162184039.0,
      "step": 134830
    },
    {
      "entropy": 1.8900707513093948,
      "epoch": 0.41799220217009375,
      "grad_norm": 7.8856329917907715,
      "learning_rate": 3.913003172336033e-06,
      "loss": 0.4777,
      "mean_token_accuracy": 0.8358192846179009,
      "num_tokens": 162196357.0,
      "step": 134840
    },
    {
      "entropy": 1.93702944368124,
      "epoch": 0.4180232012951434,
      "grad_norm": 7.011838912963867,
      "learning_rate": 3.912858081318703e-06,
      "loss": 0.4432,
      "mean_token_accuracy": 0.854140405356884,
      "num_tokens": 162207437.0,
      "step": 134850
    },
    {
      "entropy": 1.9064744263887405,
      "epoch": 0.41805420042019314,
      "grad_norm": 3.6196608543395996,
      "learning_rate": 3.912713006439751e-06,
      "loss": 0.5159,
      "mean_token_accuracy": 0.8377542212605477,
      "num_tokens": 162218620.0,
      "step": 134860
    },
    {
      "entropy": 1.8770883545279502,
      "epoch": 0.4180851995452428,
      "grad_norm": 8.111139297485352,
      "learning_rate": 3.912567947696187e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.858944533765316,
      "num_tokens": 162230501.0,
      "step": 134870
    },
    {
      "entropy": 1.9460731491446495,
      "epoch": 0.41811619867029254,
      "grad_norm": 10.321117401123047,
      "learning_rate": 3.912422905085019e-06,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8436010986566543,
      "num_tokens": 162241864.0,
      "step": 134880
    },
    {
      "entropy": 1.9751074135303497,
      "epoch": 0.4181471977953422,
      "grad_norm": 7.4099273681640625,
      "learning_rate": 3.912277878603257e-06,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8479305684566498,
      "num_tokens": 162252709.0,
      "step": 134890
    },
    {
      "entropy": 1.8740838050842286,
      "epoch": 0.41817819692039193,
      "grad_norm": 3.3563060760498047,
      "learning_rate": 3.9121328682479126e-06,
      "loss": 0.4222,
      "mean_token_accuracy": 0.8611055985093117,
      "num_tokens": 162264722.0,
      "step": 134900
    },
    {
      "entropy": 1.8472968205809592,
      "epoch": 0.4182091960454416,
      "grad_norm": 4.131076812744141,
      "learning_rate": 3.911987874015997e-06,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8558681800961494,
      "num_tokens": 162277616.0,
      "step": 134910
    },
    {
      "entropy": 1.9546231135725975,
      "epoch": 0.4182401951704913,
      "grad_norm": 3.539958953857422,
      "learning_rate": 3.911842895904522e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8383591964840889,
      "num_tokens": 162288986.0,
      "step": 134920
    },
    {
      "entropy": 1.8494961842894555,
      "epoch": 0.418271194295541,
      "grad_norm": 7.521032810211182,
      "learning_rate": 3.911697933910501e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8557368025183678,
      "num_tokens": 162301754.0,
      "step": 134930
    },
    {
      "entropy": 1.8642336532473565,
      "epoch": 0.4183021934205907,
      "grad_norm": 4.32247257232666,
      "learning_rate": 3.911552988030949e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8557508587837219,
      "num_tokens": 162313972.0,
      "step": 134940
    },
    {
      "entropy": 1.94805389046669,
      "epoch": 0.4183331925456404,
      "grad_norm": 9.484874725341797,
      "learning_rate": 3.911408058262879e-06,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8523627325892449,
      "num_tokens": 162324752.0,
      "step": 134950
    },
    {
      "entropy": 1.8355311632156373,
      "epoch": 0.4183641916706901,
      "grad_norm": 16.279993057250977,
      "learning_rate": 3.911263144603306e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.8622775912284851,
      "num_tokens": 162337826.0,
      "step": 134960
    },
    {
      "entropy": 1.8673854805529118,
      "epoch": 0.4183951907957398,
      "grad_norm": 7.603078365325928,
      "learning_rate": 3.911118247049249e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8558634921908379,
      "num_tokens": 162350433.0,
      "step": 134970
    },
    {
      "entropy": 1.9007538348436355,
      "epoch": 0.4184261899207895,
      "grad_norm": 7.9676337242126465,
      "learning_rate": 3.9109733655977225e-06,
      "loss": 0.5029,
      "mean_token_accuracy": 0.8374831721186637,
      "num_tokens": 162362674.0,
      "step": 134980
    },
    {
      "entropy": 1.8434744045138358,
      "epoch": 0.4184571890458392,
      "grad_norm": 7.775720119476318,
      "learning_rate": 3.910828500245745e-06,
      "loss": 0.4214,
      "mean_token_accuracy": 0.8564092546701432,
      "num_tokens": 162375250.0,
      "step": 134990
    },
    {
      "entropy": 1.918935863673687,
      "epoch": 0.4184881881708889,
      "grad_norm": 7.319066524505615,
      "learning_rate": 3.910683650990335e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.861502180993557,
      "num_tokens": 162386475.0,
      "step": 135000
    },
    {
      "entropy": 1.8640062302350997,
      "epoch": 0.41851918729593857,
      "grad_norm": 8.106133460998535,
      "learning_rate": 3.910538817828512e-06,
      "loss": 0.4262,
      "mean_token_accuracy": 0.854218167066574,
      "num_tokens": 162398652.0,
      "step": 135010
    },
    {
      "entropy": 1.7948240421712398,
      "epoch": 0.4185501864209883,
      "grad_norm": 3.9371562004089355,
      "learning_rate": 3.910394000757297e-06,
      "loss": 0.4051,
      "mean_token_accuracy": 0.8555122867226601,
      "num_tokens": 162411080.0,
      "step": 135020
    },
    {
      "entropy": 1.8870450854301453,
      "epoch": 0.41858118554603796,
      "grad_norm": 7.394895076751709,
      "learning_rate": 3.910249199773708e-06,
      "loss": 0.4217,
      "mean_token_accuracy": 0.8639406561851501,
      "num_tokens": 162422198.0,
      "step": 135030
    },
    {
      "entropy": 1.928053417801857,
      "epoch": 0.4186121846710877,
      "grad_norm": 7.574821472167969,
      "learning_rate": 3.910104414874769e-06,
      "loss": 0.5109,
      "mean_token_accuracy": 0.84600929915905,
      "num_tokens": 162432961.0,
      "step": 135040
    },
    {
      "entropy": 1.9174908056855202,
      "epoch": 0.41864318379613735,
      "grad_norm": 8.21877384185791,
      "learning_rate": 3.909959646057503e-06,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8488267034292221,
      "num_tokens": 162443874.0,
      "step": 135050
    },
    {
      "entropy": 1.8227450162172318,
      "epoch": 0.4186741829211871,
      "grad_norm": 9.875720024108887,
      "learning_rate": 3.90981489331893e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8520313039422035,
      "num_tokens": 162457203.0,
      "step": 135060
    },
    {
      "entropy": 1.8200043380260467,
      "epoch": 0.41870518204623675,
      "grad_norm": 10.07118034362793,
      "learning_rate": 3.909670156656076e-06,
      "loss": 0.4238,
      "mean_token_accuracy": 0.8482141956686974,
      "num_tokens": 162469740.0,
      "step": 135070
    },
    {
      "entropy": 1.9454358339309692,
      "epoch": 0.4187361811712864,
      "grad_norm": 9.135685920715332,
      "learning_rate": 3.909525436065966e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8446780741214752,
      "num_tokens": 162480950.0,
      "step": 135080
    },
    {
      "entropy": 1.802914521098137,
      "epoch": 0.41876718029633614,
      "grad_norm": 7.378044605255127,
      "learning_rate": 3.909380731545625e-06,
      "loss": 0.3828,
      "mean_token_accuracy": 0.8730568781495094,
      "num_tokens": 162493924.0,
      "step": 135090
    },
    {
      "entropy": 1.762102383375168,
      "epoch": 0.4187981794213858,
      "grad_norm": 8.075687408447266,
      "learning_rate": 3.90923604309208e-06,
      "loss": 0.3709,
      "mean_token_accuracy": 0.8703906610608101,
      "num_tokens": 162507010.0,
      "step": 135100
    },
    {
      "entropy": 1.890143983066082,
      "epoch": 0.41882917854643553,
      "grad_norm": 8.960722923278809,
      "learning_rate": 3.9090913707023566e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8449656277894974,
      "num_tokens": 162518570.0,
      "step": 135110
    },
    {
      "entropy": 1.8910117477178574,
      "epoch": 0.4188601776714852,
      "grad_norm": 3.260122776031494,
      "learning_rate": 3.908946714373483e-06,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8450547441840172,
      "num_tokens": 162530768.0,
      "step": 135120
    },
    {
      "entropy": 1.8741901010274886,
      "epoch": 0.4188911767965349,
      "grad_norm": 3.713733196258545,
      "learning_rate": 3.908802074102489e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8599935069680213,
      "num_tokens": 162542560.0,
      "step": 135130
    },
    {
      "entropy": 1.8225472882390021,
      "epoch": 0.4189221759215846,
      "grad_norm": 2.8528435230255127,
      "learning_rate": 3.908657449886402e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8527352377772331,
      "num_tokens": 162555416.0,
      "step": 135140
    },
    {
      "entropy": 1.9285516187548637,
      "epoch": 0.4189531750466343,
      "grad_norm": 9.028520584106445,
      "learning_rate": 3.908512841722253e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8442423656582833,
      "num_tokens": 162566844.0,
      "step": 135150
    },
    {
      "entropy": 1.9082002013921737,
      "epoch": 0.418984174171684,
      "grad_norm": 8.118371963500977,
      "learning_rate": 3.908368249607073e-06,
      "loss": 0.5018,
      "mean_token_accuracy": 0.8459517538547516,
      "num_tokens": 162578692.0,
      "step": 135160
    },
    {
      "entropy": 1.9068205058574677,
      "epoch": 0.4190151732967337,
      "grad_norm": 8.231268882751465,
      "learning_rate": 3.908223673537895e-06,
      "loss": 0.451,
      "mean_token_accuracy": 0.8569748342037201,
      "num_tokens": 162590372.0,
      "step": 135170
    },
    {
      "entropy": 1.7947299778461456,
      "epoch": 0.4190461724217834,
      "grad_norm": 7.387755393981934,
      "learning_rate": 3.908079113511748e-06,
      "loss": 0.4068,
      "mean_token_accuracy": 0.8611501380801201,
      "num_tokens": 162603085.0,
      "step": 135180
    },
    {
      "entropy": 1.830442163348198,
      "epoch": 0.4190771715468331,
      "grad_norm": 3.5146117210388184,
      "learning_rate": 3.907934569525668e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8458912119269371,
      "num_tokens": 162616037.0,
      "step": 135190
    },
    {
      "entropy": 1.8631732419133187,
      "epoch": 0.4191081706718828,
      "grad_norm": 8.939651489257812,
      "learning_rate": 3.907790041576687e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8587908774614335,
      "num_tokens": 162627961.0,
      "step": 135200
    },
    {
      "entropy": 1.891756534576416,
      "epoch": 0.4191391697969325,
      "grad_norm": 7.691137313842773,
      "learning_rate": 3.907645529661842e-06,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8453534230589866,
      "num_tokens": 162640527.0,
      "step": 135210
    },
    {
      "entropy": 1.880522905290127,
      "epoch": 0.41917016892198217,
      "grad_norm": 8.83800983428955,
      "learning_rate": 3.907501033778167e-06,
      "loss": 0.4631,
      "mean_token_accuracy": 0.8536536455154419,
      "num_tokens": 162652160.0,
      "step": 135220
    },
    {
      "entropy": 1.9340814992785453,
      "epoch": 0.4192011680470319,
      "grad_norm": 9.312424659729004,
      "learning_rate": 3.907356553922698e-06,
      "loss": 0.4707,
      "mean_token_accuracy": 0.8431622445583343,
      "num_tokens": 162663724.0,
      "step": 135230
    },
    {
      "entropy": 1.7783160164952279,
      "epoch": 0.41923216717208156,
      "grad_norm": 8.399624824523926,
      "learning_rate": 3.907212090092472e-06,
      "loss": 0.3552,
      "mean_token_accuracy": 0.8627290308475495,
      "num_tokens": 162677173.0,
      "step": 135240
    },
    {
      "entropy": 1.9121920481324195,
      "epoch": 0.4192631662971313,
      "grad_norm": 4.327278137207031,
      "learning_rate": 3.907067642284528e-06,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8431808009743691,
      "num_tokens": 162688069.0,
      "step": 135250
    },
    {
      "entropy": 1.90309626609087,
      "epoch": 0.41929416542218095,
      "grad_norm": 7.39229679107666,
      "learning_rate": 3.906923210495903e-06,
      "loss": 0.4782,
      "mean_token_accuracy": 0.8486498475074769,
      "num_tokens": 162699438.0,
      "step": 135260
    },
    {
      "entropy": 1.7719358682632447,
      "epoch": 0.4193251645472307,
      "grad_norm": 2.3753769397735596,
      "learning_rate": 3.9067787947236376e-06,
      "loss": 0.3807,
      "mean_token_accuracy": 0.8691380977630615,
      "num_tokens": 162711906.0,
      "step": 135270
    },
    {
      "entropy": 1.7657958284020423,
      "epoch": 0.41935616367228035,
      "grad_norm": 8.348288536071777,
      "learning_rate": 3.906634394964771e-06,
      "loss": 0.399,
      "mean_token_accuracy": 0.8593086928129197,
      "num_tokens": 162725490.0,
      "step": 135280
    },
    {
      "entropy": 1.8791421324014663,
      "epoch": 0.41938716279733007,
      "grad_norm": 5.080158233642578,
      "learning_rate": 3.906490011216344e-06,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8483076050877572,
      "num_tokens": 162737225.0,
      "step": 135290
    },
    {
      "entropy": 1.77224540412426,
      "epoch": 0.41941816192237974,
      "grad_norm": 9.42541790008545,
      "learning_rate": 3.9063456434754005e-06,
      "loss": 0.3504,
      "mean_token_accuracy": 0.8711794734001159,
      "num_tokens": 162750989.0,
      "step": 135300
    },
    {
      "entropy": 1.909542678296566,
      "epoch": 0.41944916104742946,
      "grad_norm": 8.59212589263916,
      "learning_rate": 3.906201291738979e-06,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8415026575326919,
      "num_tokens": 162762817.0,
      "step": 135310
    },
    {
      "entropy": 1.8708134442567825,
      "epoch": 0.41948016017247913,
      "grad_norm": 7.089857578277588,
      "learning_rate": 3.906056956004125e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8642296716570854,
      "num_tokens": 162774634.0,
      "step": 135320
    },
    {
      "entropy": 1.8708177879452705,
      "epoch": 0.4195111592975288,
      "grad_norm": 7.2748823165893555,
      "learning_rate": 3.905912636267882e-06,
      "loss": 0.3918,
      "mean_token_accuracy": 0.8618823811411858,
      "num_tokens": 162786925.0,
      "step": 135330
    },
    {
      "entropy": 1.9178523600101471,
      "epoch": 0.4195421584225785,
      "grad_norm": 7.701486110687256,
      "learning_rate": 3.905768332527295e-06,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8496042251586914,
      "num_tokens": 162799600.0,
      "step": 135340
    },
    {
      "entropy": 1.8545253962278365,
      "epoch": 0.4195731575476282,
      "grad_norm": 8.929347038269043,
      "learning_rate": 3.905624044779408e-06,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8579249680042267,
      "num_tokens": 162811881.0,
      "step": 135350
    },
    {
      "entropy": 1.8590249836444854,
      "epoch": 0.4196041566726779,
      "grad_norm": 7.88853645324707,
      "learning_rate": 3.905479773021269e-06,
      "loss": 0.4949,
      "mean_token_accuracy": 0.843219818174839,
      "num_tokens": 162824189.0,
      "step": 135360
    },
    {
      "entropy": 1.8993327513337135,
      "epoch": 0.4196351557977276,
      "grad_norm": 7.894437789916992,
      "learning_rate": 3.905335517249924e-06,
      "loss": 0.4708,
      "mean_token_accuracy": 0.8462705582380294,
      "num_tokens": 162836157.0,
      "step": 135370
    },
    {
      "entropy": 1.924989990890026,
      "epoch": 0.4196661549227773,
      "grad_norm": 3.602142333984375,
      "learning_rate": 3.9051912774624215e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8598493561148643,
      "num_tokens": 162847849.0,
      "step": 135380
    },
    {
      "entropy": 1.8464587591588497,
      "epoch": 0.419697154047827,
      "grad_norm": 9.821682929992676,
      "learning_rate": 3.905047053655809e-06,
      "loss": 0.3962,
      "mean_token_accuracy": 0.8622051507234574,
      "num_tokens": 162860582.0,
      "step": 135390
    },
    {
      "entropy": 1.8456949055194856,
      "epoch": 0.4197281531728767,
      "grad_norm": 3.930516242980957,
      "learning_rate": 3.904902845827135e-06,
      "loss": 0.401,
      "mean_token_accuracy": 0.8577687725424766,
      "num_tokens": 162873376.0,
      "step": 135400
    },
    {
      "entropy": 1.805974441766739,
      "epoch": 0.4197591522979264,
      "grad_norm": 7.337070941925049,
      "learning_rate": 3.904758653973452e-06,
      "loss": 0.3964,
      "mean_token_accuracy": 0.8626691713929177,
      "num_tokens": 162886100.0,
      "step": 135410
    },
    {
      "entropy": 1.9164967209100723,
      "epoch": 0.4197901514229761,
      "grad_norm": 9.788459777832031,
      "learning_rate": 3.904614478091809e-06,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8476973131299019,
      "num_tokens": 162897812.0,
      "step": 135420
    },
    {
      "entropy": 1.9670021116733551,
      "epoch": 0.41982115054802577,
      "grad_norm": 9.283044815063477,
      "learning_rate": 3.904470318179257e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8294244706630707,
      "num_tokens": 162909541.0,
      "step": 135430
    },
    {
      "entropy": 1.9219158321619034,
      "epoch": 0.4198521496730755,
      "grad_norm": 8.37589168548584,
      "learning_rate": 3.9043261742328505e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8363066896796226,
      "num_tokens": 162921516.0,
      "step": 135440
    },
    {
      "entropy": 1.953237357735634,
      "epoch": 0.41988314879812516,
      "grad_norm": 9.618903160095215,
      "learning_rate": 3.904182046249641e-06,
      "loss": 0.534,
      "mean_token_accuracy": 0.8366630434989929,
      "num_tokens": 162932956.0,
      "step": 135450
    },
    {
      "entropy": 1.9217358708381653,
      "epoch": 0.4199141479231749,
      "grad_norm": 7.9734649658203125,
      "learning_rate": 3.904037934226683e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8531175389885902,
      "num_tokens": 162944286.0,
      "step": 135460
    },
    {
      "entropy": 1.799736338853836,
      "epoch": 0.41994514704822455,
      "grad_norm": 8.599648475646973,
      "learning_rate": 3.903893838161029e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.8724782764911652,
      "num_tokens": 162957587.0,
      "step": 135470
    },
    {
      "entropy": 1.8858862951397897,
      "epoch": 0.4199761461732743,
      "grad_norm": 3.428628444671631,
      "learning_rate": 3.903749758049738e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8496242195367814,
      "num_tokens": 162969224.0,
      "step": 135480
    },
    {
      "entropy": 1.8270292654633522,
      "epoch": 0.42000714529832395,
      "grad_norm": 2.5544705390930176,
      "learning_rate": 3.903605693889863e-06,
      "loss": 0.3985,
      "mean_token_accuracy": 0.8701824352145195,
      "num_tokens": 162982342.0,
      "step": 135490
    },
    {
      "entropy": 1.8943888053297997,
      "epoch": 0.42003814442337367,
      "grad_norm": 9.184906005859375,
      "learning_rate": 3.90346164567846e-06,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8437836706638336,
      "num_tokens": 162994089.0,
      "step": 135500
    },
    {
      "entropy": 1.8288681022822857,
      "epoch": 0.42006914354842334,
      "grad_norm": 8.831469535827637,
      "learning_rate": 3.903317613412592e-06,
      "loss": 0.394,
      "mean_token_accuracy": 0.867976401746273,
      "num_tokens": 163006600.0,
      "step": 135510
    },
    {
      "entropy": 1.9004840448498725,
      "epoch": 0.42010014267347306,
      "grad_norm": 8.030474662780762,
      "learning_rate": 3.903173597089313e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.8580782458186149,
      "num_tokens": 163017668.0,
      "step": 135520
    },
    {
      "entropy": 1.9328984439373016,
      "epoch": 0.42013114179852273,
      "grad_norm": 8.515178680419922,
      "learning_rate": 3.903029596705682e-06,
      "loss": 0.4733,
      "mean_token_accuracy": 0.8430827930569649,
      "num_tokens": 163029533.0,
      "step": 135530
    },
    {
      "entropy": 1.903455564379692,
      "epoch": 0.42016214092357246,
      "grad_norm": 8.266940116882324,
      "learning_rate": 3.90288561225876e-06,
      "loss": 0.4381,
      "mean_token_accuracy": 0.8574344128370285,
      "num_tokens": 163041861.0,
      "step": 135540
    },
    {
      "entropy": 1.7817657575011254,
      "epoch": 0.4201931400486221,
      "grad_norm": 8.015380859375,
      "learning_rate": 3.902741643745609e-06,
      "loss": 0.3766,
      "mean_token_accuracy": 0.8572424992918968,
      "num_tokens": 163054619.0,
      "step": 135550
    },
    {
      "entropy": 1.8124181941151618,
      "epoch": 0.4202241391736718,
      "grad_norm": 5.867700576782227,
      "learning_rate": 3.902597691163288e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8469098672270775,
      "num_tokens": 163067835.0,
      "step": 135560
    },
    {
      "entropy": 1.9572464644908905,
      "epoch": 0.4202551382987215,
      "grad_norm": 8.415164947509766,
      "learning_rate": 3.902453754508861e-06,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8418419510126114,
      "num_tokens": 163079318.0,
      "step": 135570
    },
    {
      "entropy": 1.9142266780138015,
      "epoch": 0.4202861374237712,
      "grad_norm": 7.7311577796936035,
      "learning_rate": 3.902309833779389e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8493567198514939,
      "num_tokens": 163090456.0,
      "step": 135580
    },
    {
      "entropy": 1.9059449434280396,
      "epoch": 0.4203171365488209,
      "grad_norm": 7.837739944458008,
      "learning_rate": 3.902165928971938e-06,
      "loss": 0.4521,
      "mean_token_accuracy": 0.857314033806324,
      "num_tokens": 163101560.0,
      "step": 135590
    },
    {
      "entropy": 1.9108774542808533,
      "epoch": 0.4203481356738706,
      "grad_norm": 7.050215244293213,
      "learning_rate": 3.90202204008357e-06,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8525308519601822,
      "num_tokens": 163113176.0,
      "step": 135600
    },
    {
      "entropy": 1.8393698036670685,
      "epoch": 0.4203791347989203,
      "grad_norm": 3.4507510662078857,
      "learning_rate": 3.901878167111353e-06,
      "loss": 0.4471,
      "mean_token_accuracy": 0.8612019449472428,
      "num_tokens": 163126007.0,
      "step": 135610
    },
    {
      "entropy": 1.9607517927885056,
      "epoch": 0.42041013392397,
      "grad_norm": 7.289004325866699,
      "learning_rate": 3.9017343100523505e-06,
      "loss": 0.5067,
      "mean_token_accuracy": 0.8425389587879181,
      "num_tokens": 163136804.0,
      "step": 135620
    },
    {
      "entropy": 1.8896097630262374,
      "epoch": 0.4204411330490197,
      "grad_norm": 8.581795692443848,
      "learning_rate": 3.90159046890363e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8448723956942559,
      "num_tokens": 163148832.0,
      "step": 135630
    },
    {
      "entropy": 1.922749936580658,
      "epoch": 0.42047213217406937,
      "grad_norm": 8.489492416381836,
      "learning_rate": 3.90144664366226e-06,
      "loss": 0.4862,
      "mean_token_accuracy": 0.8492430374026299,
      "num_tokens": 163160997.0,
      "step": 135640
    },
    {
      "entropy": 1.9693446904420853,
      "epoch": 0.4205031312991191,
      "grad_norm": 9.86866283416748,
      "learning_rate": 3.9013028343253065e-06,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8453862652182579,
      "num_tokens": 163172061.0,
      "step": 135650
    },
    {
      "entropy": 1.9001538813114167,
      "epoch": 0.42053413042416876,
      "grad_norm": 7.654898166656494,
      "learning_rate": 3.901159040889842e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8523498460650444,
      "num_tokens": 163184244.0,
      "step": 135660
    },
    {
      "entropy": 1.945447364449501,
      "epoch": 0.4205651295492185,
      "grad_norm": 8.955438613891602,
      "learning_rate": 3.901015263352933e-06,
      "loss": 0.5053,
      "mean_token_accuracy": 0.8504986420273781,
      "num_tokens": 163195436.0,
      "step": 135670
    },
    {
      "entropy": 1.9072694763541223,
      "epoch": 0.42059612867426815,
      "grad_norm": 8.15773868560791,
      "learning_rate": 3.900871501711651e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8516003981232643,
      "num_tokens": 163206912.0,
      "step": 135680
    },
    {
      "entropy": 1.8923516809940337,
      "epoch": 0.4206271277993179,
      "grad_norm": 8.769623756408691,
      "learning_rate": 3.900727755963067e-06,
      "loss": 0.4825,
      "mean_token_accuracy": 0.848148649930954,
      "num_tokens": 163218689.0,
      "step": 135690
    },
    {
      "entropy": 1.8840610541403293,
      "epoch": 0.42065812692436755,
      "grad_norm": 7.660909652709961,
      "learning_rate": 3.9005840261042535e-06,
      "loss": 0.4562,
      "mean_token_accuracy": 0.8552731052041054,
      "num_tokens": 163231024.0,
      "step": 135700
    },
    {
      "entropy": 1.903901606798172,
      "epoch": 0.4206891260494173,
      "grad_norm": 3.740813732147217,
      "learning_rate": 3.900440312132283e-06,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8453383758664131,
      "num_tokens": 163242740.0,
      "step": 135710
    },
    {
      "entropy": 1.9122828841209412,
      "epoch": 0.42072012517446694,
      "grad_norm": 4.723738670349121,
      "learning_rate": 3.90029661404423e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8527891039848328,
      "num_tokens": 163254707.0,
      "step": 135720
    },
    {
      "entropy": 1.9014658220112324,
      "epoch": 0.42075112429951667,
      "grad_norm": 3.560215711593628,
      "learning_rate": 3.900152931837168e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8482464477419853,
      "num_tokens": 163266816.0,
      "step": 135730
    },
    {
      "entropy": 1.9249530613422394,
      "epoch": 0.42078212342456633,
      "grad_norm": 8.004035949707031,
      "learning_rate": 3.900009265508172e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8406677231192589,
      "num_tokens": 163278761.0,
      "step": 135740
    },
    {
      "entropy": 1.8970685094594955,
      "epoch": 0.42081312254961606,
      "grad_norm": 8.880062103271484,
      "learning_rate": 3.899865615054318e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8497609063982964,
      "num_tokens": 163291090.0,
      "step": 135750
    },
    {
      "entropy": 1.828018756210804,
      "epoch": 0.4208441216746657,
      "grad_norm": 7.22142219543457,
      "learning_rate": 3.8997219804726815e-06,
      "loss": 0.43,
      "mean_token_accuracy": 0.8671692669391632,
      "num_tokens": 163302749.0,
      "step": 135760
    },
    {
      "entropy": 1.8527932271361351,
      "epoch": 0.42087512079971545,
      "grad_norm": 7.9586286544799805,
      "learning_rate": 3.899578361760341e-06,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8436411425471306,
      "num_tokens": 163315230.0,
      "step": 135770
    },
    {
      "entropy": 1.8658724144101142,
      "epoch": 0.4209061199247651,
      "grad_norm": 4.109507083892822,
      "learning_rate": 3.899434758914374e-06,
      "loss": 0.4303,
      "mean_token_accuracy": 0.8581762701272965,
      "num_tokens": 163327266.0,
      "step": 135780
    },
    {
      "entropy": 1.842541829496622,
      "epoch": 0.42093711904981485,
      "grad_norm": 9.778244972229004,
      "learning_rate": 3.89929117193186e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8542135626077652,
      "num_tokens": 163339658.0,
      "step": 135790
    },
    {
      "entropy": 1.8465433612465858,
      "epoch": 0.4209681181748645,
      "grad_norm": 8.6669282913208,
      "learning_rate": 3.899147600809877e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8487094223499299,
      "num_tokens": 163352124.0,
      "step": 135800
    },
    {
      "entropy": 1.90444867759943,
      "epoch": 0.4209991172999142,
      "grad_norm": 8.977686882019043,
      "learning_rate": 3.899004045545507e-06,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8524438515305519,
      "num_tokens": 163363934.0,
      "step": 135810
    },
    {
      "entropy": 1.8843096554279328,
      "epoch": 0.4210301164249639,
      "grad_norm": 8.393902778625488,
      "learning_rate": 3.898860506135832e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8609971582889557,
      "num_tokens": 163375865.0,
      "step": 135820
    },
    {
      "entropy": 1.9457574129104613,
      "epoch": 0.4210611155500136,
      "grad_norm": 7.407272815704346,
      "learning_rate": 3.898716982577929e-06,
      "loss": 0.5116,
      "mean_token_accuracy": 0.8447777807712555,
      "num_tokens": 163386634.0,
      "step": 135830
    },
    {
      "entropy": 1.916204272210598,
      "epoch": 0.4210921146750633,
      "grad_norm": 7.169795513153076,
      "learning_rate": 3.898573474868886e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8451360926032067,
      "num_tokens": 163398136.0,
      "step": 135840
    },
    {
      "entropy": 1.887849335372448,
      "epoch": 0.42112311380011297,
      "grad_norm": 9.207273483276367,
      "learning_rate": 3.898429983005783e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.849080765247345,
      "num_tokens": 163409693.0,
      "step": 135850
    },
    {
      "entropy": 1.8300122573971749,
      "epoch": 0.4211541129251627,
      "grad_norm": 8.275792121887207,
      "learning_rate": 3.898286506985706e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8691477358341217,
      "num_tokens": 163422481.0,
      "step": 135860
    },
    {
      "entropy": 1.9414837792515756,
      "epoch": 0.42118511205021236,
      "grad_norm": 9.043680191040039,
      "learning_rate": 3.898143046805739e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8596002608537674,
      "num_tokens": 163433827.0,
      "step": 135870
    },
    {
      "entropy": 1.8458356723189353,
      "epoch": 0.4212161111752621,
      "grad_norm": 8.126672744750977,
      "learning_rate": 3.897999602462968e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.8593410521745681,
      "num_tokens": 163446168.0,
      "step": 135880
    },
    {
      "entropy": 1.925001895427704,
      "epoch": 0.42124711030031176,
      "grad_norm": 8.667831420898438,
      "learning_rate": 3.897856173954477e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8532567232847214,
      "num_tokens": 163457240.0,
      "step": 135890
    },
    {
      "entropy": 1.8693645611405372,
      "epoch": 0.4212781094253615,
      "grad_norm": 6.114764213562012,
      "learning_rate": 3.897712761277357e-06,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8476710587739944,
      "num_tokens": 163470439.0,
      "step": 135900
    },
    {
      "entropy": 1.928558248281479,
      "epoch": 0.42130910855041115,
      "grad_norm": 9.09399127960205,
      "learning_rate": 3.897569364428692e-06,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8500400453805923,
      "num_tokens": 163481986.0,
      "step": 135910
    },
    {
      "entropy": 1.9676057755947114,
      "epoch": 0.4213401076754609,
      "grad_norm": 7.342076301574707,
      "learning_rate": 3.897425983405573e-06,
      "loss": 0.494,
      "mean_token_accuracy": 0.8489060357213021,
      "num_tokens": 163492836.0,
      "step": 135920
    },
    {
      "entropy": 1.8425328716635705,
      "epoch": 0.42137110680051054,
      "grad_norm": 4.709118843078613,
      "learning_rate": 3.897282618205089e-06,
      "loss": 0.4112,
      "mean_token_accuracy": 0.8543885082006455,
      "num_tokens": 163505935.0,
      "step": 135930
    },
    {
      "entropy": 1.875008788704872,
      "epoch": 0.42140210592556027,
      "grad_norm": 2.8072144985198975,
      "learning_rate": 3.89713926882433e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8433866515755654,
      "num_tokens": 163518513.0,
      "step": 135940
    },
    {
      "entropy": 1.907314045727253,
      "epoch": 0.42143310505060994,
      "grad_norm": 7.362760543823242,
      "learning_rate": 3.896995935260386e-06,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8597197920084,
      "num_tokens": 163531303.0,
      "step": 135950
    },
    {
      "entropy": 1.9355842828750611,
      "epoch": 0.42146410417565966,
      "grad_norm": 3.6717031002044678,
      "learning_rate": 3.896852617510349e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8499437302350998,
      "num_tokens": 163542827.0,
      "step": 135960
    },
    {
      "entropy": 1.9380091413855554,
      "epoch": 0.42149510330070933,
      "grad_norm": 7.6753621101379395,
      "learning_rate": 3.896709315571311e-06,
      "loss": 0.4788,
      "mean_token_accuracy": 0.834736093878746,
      "num_tokens": 163555197.0,
      "step": 135970
    },
    {
      "entropy": 1.8424546226859093,
      "epoch": 0.42152610242575905,
      "grad_norm": 3.7391586303710938,
      "learning_rate": 3.896566029440366e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.865011602640152,
      "num_tokens": 163567600.0,
      "step": 135980
    },
    {
      "entropy": 1.830755239725113,
      "epoch": 0.4215571015508087,
      "grad_norm": 10.804282188415527,
      "learning_rate": 3.8964227591146075e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8500369563698769,
      "num_tokens": 163580961.0,
      "step": 135990
    },
    {
      "entropy": 1.8739851251244546,
      "epoch": 0.42158810067585845,
      "grad_norm": 8.843622207641602,
      "learning_rate": 3.89627950459113e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.853018419444561,
      "num_tokens": 163593529.0,
      "step": 136000
    },
    {
      "entropy": 1.8963138684630394,
      "epoch": 0.4216190998009081,
      "grad_norm": 7.0039191246032715,
      "learning_rate": 3.8961362658670284e-06,
      "loss": 0.4491,
      "mean_token_accuracy": 0.849344827234745,
      "num_tokens": 163605267.0,
      "step": 136010
    },
    {
      "entropy": 1.8308332130312919,
      "epoch": 0.42165009892595784,
      "grad_norm": 9.128500938415527,
      "learning_rate": 3.895993042939398e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8566051840782165,
      "num_tokens": 163618128.0,
      "step": 136020
    },
    {
      "entropy": 1.9336943060159684,
      "epoch": 0.4216810980510075,
      "grad_norm": 8.359414100646973,
      "learning_rate": 3.895849835805338e-06,
      "loss": 0.4617,
      "mean_token_accuracy": 0.8571832865476608,
      "num_tokens": 163629610.0,
      "step": 136030
    },
    {
      "entropy": 1.81222113519907,
      "epoch": 0.42171209717605723,
      "grad_norm": 7.943345546722412,
      "learning_rate": 3.8957066444619444e-06,
      "loss": 0.3755,
      "mean_token_accuracy": 0.86582732796669,
      "num_tokens": 163642383.0,
      "step": 136040
    },
    {
      "entropy": 1.8568004190921783,
      "epoch": 0.4217430963011069,
      "grad_norm": 5.962125778198242,
      "learning_rate": 3.895563468906315e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.8517604455351829,
      "num_tokens": 163654588.0,
      "step": 136050
    },
    {
      "entropy": 1.9352711230516433,
      "epoch": 0.42177409542615657,
      "grad_norm": 9.666674613952637,
      "learning_rate": 3.89542030913555e-06,
      "loss": 0.4991,
      "mean_token_accuracy": 0.8422815144062042,
      "num_tokens": 163665434.0,
      "step": 136060
    },
    {
      "entropy": 1.9193247556686401,
      "epoch": 0.4218050945512063,
      "grad_norm": 9.703880310058594,
      "learning_rate": 3.895277165146748e-06,
      "loss": 0.517,
      "mean_token_accuracy": 0.8462235674262046,
      "num_tokens": 163677012.0,
      "step": 136070
    },
    {
      "entropy": 1.9230276107788087,
      "epoch": 0.42183609367625596,
      "grad_norm": 9.869771003723145,
      "learning_rate": 3.895134036937011e-06,
      "loss": 0.4913,
      "mean_token_accuracy": 0.8501724734902382,
      "num_tokens": 163687569.0,
      "step": 136080
    },
    {
      "entropy": 1.9550448417663575,
      "epoch": 0.4218670928013057,
      "grad_norm": 8.047429084777832,
      "learning_rate": 3.89499092450344e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.843185929954052,
      "num_tokens": 163698183.0,
      "step": 136090
    },
    {
      "entropy": 1.8611886352300644,
      "epoch": 0.42189809192635536,
      "grad_norm": 8.042689323425293,
      "learning_rate": 3.894847827843135e-06,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8485627174377441,
      "num_tokens": 163710589.0,
      "step": 136100
    },
    {
      "entropy": 1.8729781568050385,
      "epoch": 0.4219290910514051,
      "grad_norm": 8.031086921691895,
      "learning_rate": 3.894704746953201e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8670966506004334,
      "num_tokens": 163722848.0,
      "step": 136110
    },
    {
      "entropy": 1.9010607331991196,
      "epoch": 0.42196009017645475,
      "grad_norm": 8.790847778320312,
      "learning_rate": 3.894561681830741e-06,
      "loss": 0.4238,
      "mean_token_accuracy": 0.85305155813694,
      "num_tokens": 163734838.0,
      "step": 136120
    },
    {
      "entropy": 1.869620206952095,
      "epoch": 0.4219910893015045,
      "grad_norm": 7.369499206542969,
      "learning_rate": 3.89441863247286e-06,
      "loss": 0.3984,
      "mean_token_accuracy": 0.8668390452861786,
      "num_tokens": 163747691.0,
      "step": 136130
    },
    {
      "entropy": 1.8736939072608947,
      "epoch": 0.42202208842655414,
      "grad_norm": 8.191452026367188,
      "learning_rate": 3.89427559887666e-06,
      "loss": 0.443,
      "mean_token_accuracy": 0.8507094740867615,
      "num_tokens": 163759715.0,
      "step": 136140
    },
    {
      "entropy": 1.891453741490841,
      "epoch": 0.42205308755160387,
      "grad_norm": 4.542218208312988,
      "learning_rate": 3.89413258103925e-06,
      "loss": 0.4505,
      "mean_token_accuracy": 0.8499180749058723,
      "num_tokens": 163772049.0,
      "step": 136150
    },
    {
      "entropy": 1.8985579848289489,
      "epoch": 0.42208408667665354,
      "grad_norm": 6.974556922912598,
      "learning_rate": 3.8939895789577355e-06,
      "loss": 0.4075,
      "mean_token_accuracy": 0.866258729994297,
      "num_tokens": 163783799.0,
      "step": 136160
    },
    {
      "entropy": 1.8413588017225266,
      "epoch": 0.42211508580170326,
      "grad_norm": 3.5806539058685303,
      "learning_rate": 3.893846592629224e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.856754244863987,
      "num_tokens": 163796194.0,
      "step": 136170
    },
    {
      "entropy": 1.8396148562431336,
      "epoch": 0.42214608492675293,
      "grad_norm": 7.897609710693359,
      "learning_rate": 3.893703622050822e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8517332747578621,
      "num_tokens": 163808614.0,
      "step": 136180
    },
    {
      "entropy": 1.8847883358597755,
      "epoch": 0.42217708405180265,
      "grad_norm": 8.880624771118164,
      "learning_rate": 3.89356066721964e-06,
      "loss": 0.4671,
      "mean_token_accuracy": 0.848950457572937,
      "num_tokens": 163820626.0,
      "step": 136190
    },
    {
      "entropy": 1.9313714131712914,
      "epoch": 0.4222080831768523,
      "grad_norm": 7.545229911804199,
      "learning_rate": 3.893417728132786e-06,
      "loss": 0.4886,
      "mean_token_accuracy": 0.8502359837293625,
      "num_tokens": 163832281.0,
      "step": 136200
    },
    {
      "entropy": 1.942906777560711,
      "epoch": 0.42223908230190205,
      "grad_norm": 8.471320152282715,
      "learning_rate": 3.8932748047873715e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8459069982171059,
      "num_tokens": 163843591.0,
      "step": 136210
    },
    {
      "entropy": 1.8104972153902055,
      "epoch": 0.4222700814269517,
      "grad_norm": 7.06924295425415,
      "learning_rate": 3.893131897180506e-06,
      "loss": 0.3943,
      "mean_token_accuracy": 0.8590885251760483,
      "num_tokens": 163856366.0,
      "step": 136220
    },
    {
      "entropy": 1.8111203506588935,
      "epoch": 0.42230108055200144,
      "grad_norm": 3.74467396736145,
      "learning_rate": 3.892989005309303e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8676251381635666,
      "num_tokens": 163870058.0,
      "step": 136230
    },
    {
      "entropy": 1.8452250599861144,
      "epoch": 0.4223320796770511,
      "grad_norm": 9.069093704223633,
      "learning_rate": 3.892846129170875e-06,
      "loss": 0.4211,
      "mean_token_accuracy": 0.8483367681503295,
      "num_tokens": 163883160.0,
      "step": 136240
    },
    {
      "entropy": 1.9414851903915404,
      "epoch": 0.42236307880210083,
      "grad_norm": 9.163724899291992,
      "learning_rate": 3.892703268762333e-06,
      "loss": 0.5084,
      "mean_token_accuracy": 0.8401763945817947,
      "num_tokens": 163894584.0,
      "step": 136250
    },
    {
      "entropy": 1.8744617268443107,
      "epoch": 0.4223940779271505,
      "grad_norm": 4.958732604980469,
      "learning_rate": 3.892560424080792e-06,
      "loss": 0.4281,
      "mean_token_accuracy": 0.8564569279551506,
      "num_tokens": 163906968.0,
      "step": 136260
    },
    {
      "entropy": 1.9449343591928483,
      "epoch": 0.4224250770522002,
      "grad_norm": 7.974160671234131,
      "learning_rate": 3.892417595123367e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8489760622382164,
      "num_tokens": 163918188.0,
      "step": 136270
    },
    {
      "entropy": 1.8679928988218308,
      "epoch": 0.4224560761772499,
      "grad_norm": 9.126168251037598,
      "learning_rate": 3.892274781887172e-06,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8490209579467773,
      "num_tokens": 163930662.0,
      "step": 136280
    },
    {
      "entropy": 1.8945604875683784,
      "epoch": 0.4224870753022996,
      "grad_norm": 7.6865363121032715,
      "learning_rate": 3.892131984369326e-06,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8592929482460022,
      "num_tokens": 163942417.0,
      "step": 136290
    },
    {
      "entropy": 1.9043464064598083,
      "epoch": 0.4225180744273493,
      "grad_norm": 9.015311241149902,
      "learning_rate": 3.891989202566944e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8410688266158104,
      "num_tokens": 163955136.0,
      "step": 136300
    },
    {
      "entropy": 1.9317906364798545,
      "epoch": 0.42254907355239896,
      "grad_norm": 9.661110877990723,
      "learning_rate": 3.891846436477144e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8499211475253106,
      "num_tokens": 163966625.0,
      "step": 136310
    },
    {
      "entropy": 1.9347480967640878,
      "epoch": 0.4225800726774487,
      "grad_norm": 4.00605583190918,
      "learning_rate": 3.891703686097043e-06,
      "loss": 0.5011,
      "mean_token_accuracy": 0.8386645168066025,
      "num_tokens": 163978413.0,
      "step": 136320
    },
    {
      "entropy": 1.8491806223988534,
      "epoch": 0.42261107180249835,
      "grad_norm": 6.554616451263428,
      "learning_rate": 3.891560951423763e-06,
      "loss": 0.4156,
      "mean_token_accuracy": 0.8566528141498566,
      "num_tokens": 163990762.0,
      "step": 136330
    },
    {
      "entropy": 1.9310165911912918,
      "epoch": 0.4226420709275481,
      "grad_norm": 10.993410110473633,
      "learning_rate": 3.891418232454421e-06,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8411554425954819,
      "num_tokens": 164003208.0,
      "step": 136340
    },
    {
      "entropy": 1.8688613146543502,
      "epoch": 0.42267307005259774,
      "grad_norm": 4.234386920928955,
      "learning_rate": 3.891275529186138e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8530001133680344,
      "num_tokens": 164015901.0,
      "step": 136350
    },
    {
      "entropy": 1.8955225050449371,
      "epoch": 0.42270406917764747,
      "grad_norm": 7.75908899307251,
      "learning_rate": 3.891132841616038e-06,
      "loss": 0.3968,
      "mean_token_accuracy": 0.8615424484014511,
      "num_tokens": 164027911.0,
      "step": 136360
    },
    {
      "entropy": 1.90107059776783,
      "epoch": 0.42273506830269714,
      "grad_norm": 4.065981388092041,
      "learning_rate": 3.890990169741239e-06,
      "loss": 0.4431,
      "mean_token_accuracy": 0.8597174286842346,
      "num_tokens": 164039942.0,
      "step": 136370
    },
    {
      "entropy": 1.971032439172268,
      "epoch": 0.42276606742774686,
      "grad_norm": 8.14873218536377,
      "learning_rate": 3.890847513558867e-06,
      "loss": 0.5376,
      "mean_token_accuracy": 0.8287710756063461,
      "num_tokens": 164051132.0,
      "step": 136380
    },
    {
      "entropy": 1.933982428908348,
      "epoch": 0.42279706655279653,
      "grad_norm": 9.093382835388184,
      "learning_rate": 3.890704873066045e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8538150131702423,
      "num_tokens": 164063375.0,
      "step": 136390
    },
    {
      "entropy": 1.9692260310053826,
      "epoch": 0.42282806567784625,
      "grad_norm": 7.917427062988281,
      "learning_rate": 3.890562248259897e-06,
      "loss": 0.4822,
      "mean_token_accuracy": 0.842354828119278,
      "num_tokens": 164074799.0,
      "step": 136400
    },
    {
      "entropy": 1.949564391374588,
      "epoch": 0.4228590648028959,
      "grad_norm": 8.609591484069824,
      "learning_rate": 3.890419639137547e-06,
      "loss": 0.5208,
      "mean_token_accuracy": 0.8415541231632233,
      "num_tokens": 164086636.0,
      "step": 136410
    },
    {
      "entropy": 1.8990302249789237,
      "epoch": 0.42289006392794565,
      "grad_norm": 8.789817810058594,
      "learning_rate": 3.890277045696122e-06,
      "loss": 0.4632,
      "mean_token_accuracy": 0.8372734472155571,
      "num_tokens": 164099437.0,
      "step": 136420
    },
    {
      "entropy": 1.9440916180610657,
      "epoch": 0.4229210630529953,
      "grad_norm": 8.302007675170898,
      "learning_rate": 3.890134467932746e-06,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8474587753415108,
      "num_tokens": 164111134.0,
      "step": 136430
    },
    {
      "entropy": 1.9455933913588523,
      "epoch": 0.42295206217804504,
      "grad_norm": 9.274413108825684,
      "learning_rate": 3.889991905844551e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.84911008477211,
      "num_tokens": 164122818.0,
      "step": 136440
    },
    {
      "entropy": 1.8401254639029503,
      "epoch": 0.4229830613030947,
      "grad_norm": 7.885955810546875,
      "learning_rate": 3.88984935942866e-06,
      "loss": 0.474,
      "mean_token_accuracy": 0.8547663778066635,
      "num_tokens": 164135328.0,
      "step": 136450
    },
    {
      "entropy": 1.8858144536614418,
      "epoch": 0.42301406042814443,
      "grad_norm": 7.3441338539123535,
      "learning_rate": 3.8897068286822046e-06,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8430177375674248,
      "num_tokens": 164147713.0,
      "step": 136460
    },
    {
      "entropy": 1.9624179631471634,
      "epoch": 0.4230450595531941,
      "grad_norm": 7.776828289031982,
      "learning_rate": 3.8895643136023146e-06,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8557393714785576,
      "num_tokens": 164159157.0,
      "step": 136470
    },
    {
      "entropy": 1.9207000091671944,
      "epoch": 0.4230760586782438,
      "grad_norm": 8.962823867797852,
      "learning_rate": 3.889421814186118e-06,
      "loss": 0.5688,
      "mean_token_accuracy": 0.8365709602832794,
      "num_tokens": 164170834.0,
      "step": 136480
    },
    {
      "entropy": 1.9185936331748963,
      "epoch": 0.4231070578032935,
      "grad_norm": 6.344710826873779,
      "learning_rate": 3.889279330430746e-06,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8482104346156121,
      "num_tokens": 164183232.0,
      "step": 136490
    },
    {
      "entropy": 1.908946332335472,
      "epoch": 0.4231380569283432,
      "grad_norm": 7.712813854217529,
      "learning_rate": 3.889136862333333e-06,
      "loss": 0.4495,
      "mean_token_accuracy": 0.8569162204861641,
      "num_tokens": 164195812.0,
      "step": 136500
    },
    {
      "entropy": 1.9538255900144577,
      "epoch": 0.4231690560533929,
      "grad_norm": 8.661401748657227,
      "learning_rate": 3.888994409891007e-06,
      "loss": 0.5378,
      "mean_token_accuracy": 0.8376554921269417,
      "num_tokens": 164207031.0,
      "step": 136510
    },
    {
      "entropy": 1.9065537318587302,
      "epoch": 0.4232000551784426,
      "grad_norm": 3.5878145694732666,
      "learning_rate": 3.8888519731009065e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8541804388165474,
      "num_tokens": 164219113.0,
      "step": 136520
    },
    {
      "entropy": 1.9570256814360618,
      "epoch": 0.4232310543034923,
      "grad_norm": 11.070653915405273,
      "learning_rate": 3.8887095519601594e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8492454007267952,
      "num_tokens": 164230241.0,
      "step": 136530
    },
    {
      "entropy": 1.9088216736912726,
      "epoch": 0.423262053428542,
      "grad_norm": 5.85700798034668,
      "learning_rate": 3.888567146465905e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8500758215785027,
      "num_tokens": 164242548.0,
      "step": 136540
    },
    {
      "entropy": 1.9074143424630166,
      "epoch": 0.4232930525535917,
      "grad_norm": 7.5828938484191895,
      "learning_rate": 3.888424756615277e-06,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8519628167152404,
      "num_tokens": 164254254.0,
      "step": 136550
    },
    {
      "entropy": 1.958798161149025,
      "epoch": 0.42332405167864134,
      "grad_norm": 7.560995101928711,
      "learning_rate": 3.888282382405411e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8531105071306229,
      "num_tokens": 164265550.0,
      "step": 136560
    },
    {
      "entropy": 1.9580877602100373,
      "epoch": 0.42335505080369107,
      "grad_norm": 7.767574787139893,
      "learning_rate": 3.888140023833444e-06,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8387601390480995,
      "num_tokens": 164277156.0,
      "step": 136570
    },
    {
      "entropy": 1.9372137054800986,
      "epoch": 0.42338604992874074,
      "grad_norm": 7.815626621246338,
      "learning_rate": 3.887997680896513e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8588192850351334,
      "num_tokens": 164288771.0,
      "step": 136580
    },
    {
      "entropy": 1.8666774332523346,
      "epoch": 0.42341704905379046,
      "grad_norm": 11.0514554977417,
      "learning_rate": 3.887855353591757e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8574911624193191,
      "num_tokens": 164302139.0,
      "step": 136590
    },
    {
      "entropy": 1.8457847326993941,
      "epoch": 0.42344804817884013,
      "grad_norm": 3.481797695159912,
      "learning_rate": 3.887713041916315e-06,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8640109539031983,
      "num_tokens": 164315100.0,
      "step": 136600
    },
    {
      "entropy": 1.8275635659694671,
      "epoch": 0.42347904730388986,
      "grad_norm": 7.928697109222412,
      "learning_rate": 3.887570745867327e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8522553309798241,
      "num_tokens": 164328806.0,
      "step": 136610
    },
    {
      "entropy": 1.945831647515297,
      "epoch": 0.4235100464289395,
      "grad_norm": 8.89621353149414,
      "learning_rate": 3.887428465441934e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8539063900709152,
      "num_tokens": 164340003.0,
      "step": 136620
    },
    {
      "entropy": 1.9486369535326957,
      "epoch": 0.42354104555398925,
      "grad_norm": 9.575989723205566,
      "learning_rate": 3.8872862006372745e-06,
      "loss": 0.4362,
      "mean_token_accuracy": 0.8562672764062882,
      "num_tokens": 164350643.0,
      "step": 136630
    },
    {
      "entropy": 1.8524845391511917,
      "epoch": 0.4235720446790389,
      "grad_norm": 6.633891582489014,
      "learning_rate": 3.887143951450493e-06,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8535214066505432,
      "num_tokens": 164363583.0,
      "step": 136640
    },
    {
      "entropy": 1.833214531838894,
      "epoch": 0.42360304380408864,
      "grad_norm": 7.453033924102783,
      "learning_rate": 3.887001717878731e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8589364722371101,
      "num_tokens": 164377000.0,
      "step": 136650
    },
    {
      "entropy": 1.7080215141177177,
      "epoch": 0.4236340429291383,
      "grad_norm": 7.440715312957764,
      "learning_rate": 3.886859499919133e-06,
      "loss": 0.356,
      "mean_token_accuracy": 0.8675283506512642,
      "num_tokens": 164390708.0,
      "step": 136660
    },
    {
      "entropy": 1.8073251724243165,
      "epoch": 0.42366504205418803,
      "grad_norm": 4.673215389251709,
      "learning_rate": 3.886717297568841e-06,
      "loss": 0.4071,
      "mean_token_accuracy": 0.8596639856696129,
      "num_tokens": 164404186.0,
      "step": 136670
    },
    {
      "entropy": 1.7819759905338288,
      "epoch": 0.4236960411792377,
      "grad_norm": 4.5533552169799805,
      "learning_rate": 3.8865751108250015e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8706469982862473,
      "num_tokens": 164418745.0,
      "step": 136680
    },
    {
      "entropy": 1.931021724641323,
      "epoch": 0.42372704030428743,
      "grad_norm": 7.136261463165283,
      "learning_rate": 3.88643293968476e-06,
      "loss": 0.4592,
      "mean_token_accuracy": 0.851832240819931,
      "num_tokens": 164431143.0,
      "step": 136690
    },
    {
      "entropy": 1.9522256717085837,
      "epoch": 0.4237580394293371,
      "grad_norm": 6.650088787078857,
      "learning_rate": 3.886290784145263e-06,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8387707456946373,
      "num_tokens": 164442769.0,
      "step": 136700
    },
    {
      "entropy": 1.9686541602015495,
      "epoch": 0.4237890385543868,
      "grad_norm": 4.377057075500488,
      "learning_rate": 3.886148644203657e-06,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8511147424578667,
      "num_tokens": 164454299.0,
      "step": 136710
    },
    {
      "entropy": 1.907273495197296,
      "epoch": 0.4238200376794365,
      "grad_norm": 9.434427261352539,
      "learning_rate": 3.886006519857088e-06,
      "loss": 0.4182,
      "mean_token_accuracy": 0.8593376144766808,
      "num_tokens": 164466567.0,
      "step": 136720
    },
    {
      "entropy": 1.8965537667274475,
      "epoch": 0.4238510368044862,
      "grad_norm": 7.356637477874756,
      "learning_rate": 3.885864411102708e-06,
      "loss": 0.4372,
      "mean_token_accuracy": 0.8546531453728676,
      "num_tokens": 164478726.0,
      "step": 136730
    },
    {
      "entropy": 1.8856528677046298,
      "epoch": 0.4238820359295359,
      "grad_norm": 9.142583847045898,
      "learning_rate": 3.885722317937665e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8572150781750679,
      "num_tokens": 164491306.0,
      "step": 136740
    },
    {
      "entropy": 1.8360870122909545,
      "epoch": 0.4239130350545856,
      "grad_norm": 9.105968475341797,
      "learning_rate": 3.885580240359107e-06,
      "loss": 0.4475,
      "mean_token_accuracy": 0.851330541074276,
      "num_tokens": 164503683.0,
      "step": 136750
    },
    {
      "entropy": 1.926066693663597,
      "epoch": 0.4239440341796353,
      "grad_norm": 6.350192546844482,
      "learning_rate": 3.885438178364187e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8608698025345802,
      "num_tokens": 164515517.0,
      "step": 136760
    },
    {
      "entropy": 1.9915336534380912,
      "epoch": 0.423975033304685,
      "grad_norm": 7.947946548461914,
      "learning_rate": 3.885296131950055e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8465639933943748,
      "num_tokens": 164527079.0,
      "step": 136770
    },
    {
      "entropy": 1.9105818301439286,
      "epoch": 0.42400603242973467,
      "grad_norm": 9.564684867858887,
      "learning_rate": 3.885154101113865e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8596938267350197,
      "num_tokens": 164538814.0,
      "step": 136780
    },
    {
      "entropy": 1.9386409044265747,
      "epoch": 0.4240370315547844,
      "grad_norm": 10.250873565673828,
      "learning_rate": 3.885012085852768e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.853357446193695,
      "num_tokens": 164550954.0,
      "step": 136790
    },
    {
      "entropy": 1.9958611935377122,
      "epoch": 0.42406803067983406,
      "grad_norm": 8.585638999938965,
      "learning_rate": 3.884870086163918e-06,
      "loss": 0.4789,
      "mean_token_accuracy": 0.852650736272335,
      "num_tokens": 164561994.0,
      "step": 136800
    },
    {
      "entropy": 1.8948075383901597,
      "epoch": 0.42409902980488373,
      "grad_norm": 8.89138412475586,
      "learning_rate": 3.88472810204447e-06,
      "loss": 0.4955,
      "mean_token_accuracy": 0.8488554358482361,
      "num_tokens": 164573952.0,
      "step": 136810
    },
    {
      "entropy": 1.9540094196796418,
      "epoch": 0.42413002892993346,
      "grad_norm": 7.017063140869141,
      "learning_rate": 3.884586133491578e-06,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8478411644697189,
      "num_tokens": 164585607.0,
      "step": 136820
    },
    {
      "entropy": 1.8501681357622146,
      "epoch": 0.4241610280549831,
      "grad_norm": 8.608563423156738,
      "learning_rate": 3.884444180502399e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8567686438560486,
      "num_tokens": 164598434.0,
      "step": 136830
    },
    {
      "entropy": 1.9179615780711174,
      "epoch": 0.42419202718003285,
      "grad_norm": 7.781816482543945,
      "learning_rate": 3.884302243074088e-06,
      "loss": 0.5059,
      "mean_token_accuracy": 0.8461335703730584,
      "num_tokens": 164609861.0,
      "step": 136840
    },
    {
      "entropy": 1.9359031677246095,
      "epoch": 0.4242230263050825,
      "grad_norm": 9.121841430664062,
      "learning_rate": 3.884160321203804e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8523103460669518,
      "num_tokens": 164622151.0,
      "step": 136850
    },
    {
      "entropy": 1.923108348250389,
      "epoch": 0.42425402543013224,
      "grad_norm": 3.747391700744629,
      "learning_rate": 3.884018414888704e-06,
      "loss": 0.4437,
      "mean_token_accuracy": 0.856217896938324,
      "num_tokens": 164633849.0,
      "step": 136860
    },
    {
      "entropy": 2.0000748217105864,
      "epoch": 0.4242850245551819,
      "grad_norm": 9.04723834991455,
      "learning_rate": 3.883876524125946e-06,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8461816415190697,
      "num_tokens": 164645091.0,
      "step": 136870
    },
    {
      "entropy": 1.9932536423206328,
      "epoch": 0.42431602368023164,
      "grad_norm": 9.16733169555664,
      "learning_rate": 3.883734648912692e-06,
      "loss": 0.5414,
      "mean_token_accuracy": 0.8346955180168152,
      "num_tokens": 164655606.0,
      "step": 136880
    },
    {
      "entropy": 1.936814832687378,
      "epoch": 0.4243470228052813,
      "grad_norm": 7.927580833435059,
      "learning_rate": 3.883592789246098e-06,
      "loss": 0.4945,
      "mean_token_accuracy": 0.848675799369812,
      "num_tokens": 164667161.0,
      "step": 136890
    },
    {
      "entropy": 1.9285539746284486,
      "epoch": 0.42437802193033103,
      "grad_norm": 3.780477523803711,
      "learning_rate": 3.883450945123329e-06,
      "loss": 0.4747,
      "mean_token_accuracy": 0.8493298560380935,
      "num_tokens": 164678581.0,
      "step": 136900
    },
    {
      "entropy": 1.8320277720689773,
      "epoch": 0.4244090210553807,
      "grad_norm": 9.713822364807129,
      "learning_rate": 3.883309116541545e-06,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8653510987758637,
      "num_tokens": 164690881.0,
      "step": 136910
    },
    {
      "entropy": 1.9016039952635766,
      "epoch": 0.4244400201804304,
      "grad_norm": 8.08753776550293,
      "learning_rate": 3.883167303497907e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8574401244521141,
      "num_tokens": 164703031.0,
      "step": 136920
    },
    {
      "entropy": 1.8986100777983665,
      "epoch": 0.4244710193054801,
      "grad_norm": 3.745185375213623,
      "learning_rate": 3.88302550598958e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8573842152953148,
      "num_tokens": 164715311.0,
      "step": 136930
    },
    {
      "entropy": 1.9534535899758338,
      "epoch": 0.4245020184305298,
      "grad_norm": 10.045483589172363,
      "learning_rate": 3.882883724013727e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8531460225582123,
      "num_tokens": 164727139.0,
      "step": 136940
    },
    {
      "entropy": 1.931221941113472,
      "epoch": 0.4245330175555795,
      "grad_norm": 9.615496635437012,
      "learning_rate": 3.882741957567513e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.8572171211242676,
      "num_tokens": 164738964.0,
      "step": 136950
    },
    {
      "entropy": 1.9113097980618476,
      "epoch": 0.4245640166806292,
      "grad_norm": 4.518940448760986,
      "learning_rate": 3.882600206648101e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8442388892173767,
      "num_tokens": 164750751.0,
      "step": 136960
    },
    {
      "entropy": 1.884780551493168,
      "epoch": 0.4245950158056789,
      "grad_norm": 10.810199737548828,
      "learning_rate": 3.88245847125266e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8605505630373955,
      "num_tokens": 164762720.0,
      "step": 136970
    },
    {
      "entropy": 1.8848048835992812,
      "epoch": 0.4246260149307286,
      "grad_norm": 7.640275478363037,
      "learning_rate": 3.882316751378355e-06,
      "loss": 0.4696,
      "mean_token_accuracy": 0.8543965950608253,
      "num_tokens": 164774673.0,
      "step": 136980
    },
    {
      "entropy": 1.9594868689775466,
      "epoch": 0.42465701405577827,
      "grad_norm": 8.8289213180542,
      "learning_rate": 3.882175047022354e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8498617529869079,
      "num_tokens": 164785684.0,
      "step": 136990
    },
    {
      "entropy": 1.9494032636284828,
      "epoch": 0.424688013180828,
      "grad_norm": 9.075950622558594,
      "learning_rate": 3.8820333581818245e-06,
      "loss": 0.487,
      "mean_token_accuracy": 0.8439081847667694,
      "num_tokens": 164797327.0,
      "step": 137000
    },
    {
      "entropy": 1.8996917128562927,
      "epoch": 0.42471901230587766,
      "grad_norm": 9.116379737854004,
      "learning_rate": 3.881891684853936e-06,
      "loss": 0.509,
      "mean_token_accuracy": 0.8461435765028,
      "num_tokens": 164809654.0,
      "step": 137010
    },
    {
      "entropy": 1.927696566283703,
      "epoch": 0.4247500114309274,
      "grad_norm": 8.284031867980957,
      "learning_rate": 3.881750027035857e-06,
      "loss": 0.4344,
      "mean_token_accuracy": 0.8543187946081161,
      "num_tokens": 164820690.0,
      "step": 137020
    },
    {
      "entropy": 1.773264393210411,
      "epoch": 0.42478101055597706,
      "grad_norm": 7.941432476043701,
      "learning_rate": 3.88160838472476e-06,
      "loss": 0.389,
      "mean_token_accuracy": 0.8707416027784347,
      "num_tokens": 164834760.0,
      "step": 137030
    },
    {
      "entropy": 1.798407319188118,
      "epoch": 0.4248120096810267,
      "grad_norm": 3.9334545135498047,
      "learning_rate": 3.881466757917814e-06,
      "loss": 0.3334,
      "mean_token_accuracy": 0.8698904514312744,
      "num_tokens": 164848105.0,
      "step": 137040
    },
    {
      "entropy": 1.8838206291198731,
      "epoch": 0.42484300880607645,
      "grad_norm": 7.942371845245361,
      "learning_rate": 3.8813251466121905e-06,
      "loss": 0.4114,
      "mean_token_accuracy": 0.8569322526454926,
      "num_tokens": 164860645.0,
      "step": 137050
    },
    {
      "entropy": 1.9465039163827895,
      "epoch": 0.4248740079311261,
      "grad_norm": 9.43044662475586,
      "learning_rate": 3.881183550805064e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8527957573533058,
      "num_tokens": 164871767.0,
      "step": 137060
    },
    {
      "entropy": 1.948851892352104,
      "epoch": 0.42490500705617584,
      "grad_norm": 7.4881110191345215,
      "learning_rate": 3.881041970493606e-06,
      "loss": 0.465,
      "mean_token_accuracy": 0.8495731383562088,
      "num_tokens": 164883389.0,
      "step": 137070
    },
    {
      "entropy": 1.7817671418190002,
      "epoch": 0.4249360061812255,
      "grad_norm": 3.924825429916382,
      "learning_rate": 3.88090040567499e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8674560770392418,
      "num_tokens": 164896540.0,
      "step": 137080
    },
    {
      "entropy": 1.9194593280553818,
      "epoch": 0.42496700530627524,
      "grad_norm": 7.824845790863037,
      "learning_rate": 3.8807588563463924e-06,
      "loss": 0.5412,
      "mean_token_accuracy": 0.8526005268096923,
      "num_tokens": 164908834.0,
      "step": 137090
    },
    {
      "entropy": 1.9632477343082428,
      "epoch": 0.4249980044313249,
      "grad_norm": 7.450333595275879,
      "learning_rate": 3.880617322504987e-06,
      "loss": 0.5201,
      "mean_token_accuracy": 0.8478045627474785,
      "num_tokens": 164919727.0,
      "step": 137100
    },
    {
      "entropy": 1.9021240234375,
      "epoch": 0.42502900355637463,
      "grad_norm": 8.733455657958984,
      "learning_rate": 3.8804758041479515e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.853900471329689,
      "num_tokens": 164931228.0,
      "step": 137110
    },
    {
      "entropy": 1.9155962690711021,
      "epoch": 0.4250600026814243,
      "grad_norm": 8.310453414916992,
      "learning_rate": 3.880334301272461e-06,
      "loss": 0.5104,
      "mean_token_accuracy": 0.8396357014775276,
      "num_tokens": 164943673.0,
      "step": 137120
    },
    {
      "entropy": 1.8234332576394081,
      "epoch": 0.425091001806474,
      "grad_norm": 8.392141342163086,
      "learning_rate": 3.880192813875693e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8463705331087112,
      "num_tokens": 164957038.0,
      "step": 137130
    },
    {
      "entropy": 1.8326028779149055,
      "epoch": 0.4251220009315237,
      "grad_norm": 4.345094203948975,
      "learning_rate": 3.880051341954828e-06,
      "loss": 0.3922,
      "mean_token_accuracy": 0.870023638010025,
      "num_tokens": 164970423.0,
      "step": 137140
    },
    {
      "entropy": 1.881044802069664,
      "epoch": 0.4251530000565734,
      "grad_norm": 8.87633991241455,
      "learning_rate": 3.879909885507042e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8461710780858993,
      "num_tokens": 164982236.0,
      "step": 137150
    },
    {
      "entropy": 1.8549317836761474,
      "epoch": 0.4251839991816231,
      "grad_norm": 9.388912200927734,
      "learning_rate": 3.879768444529517e-06,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8555863380432129,
      "num_tokens": 164995126.0,
      "step": 137160
    },
    {
      "entropy": 1.8838744007050992,
      "epoch": 0.4252149983066728,
      "grad_norm": 7.678390979766846,
      "learning_rate": 3.879627019019431e-06,
      "loss": 0.4511,
      "mean_token_accuracy": 0.8485071495175361,
      "num_tokens": 165007465.0,
      "step": 137170
    },
    {
      "entropy": 1.9211634308099748,
      "epoch": 0.4252459974317225,
      "grad_norm": 8.58484172821045,
      "learning_rate": 3.879485608973968e-06,
      "loss": 0.4207,
      "mean_token_accuracy": 0.8675820276141166,
      "num_tokens": 165018031.0,
      "step": 137180
    },
    {
      "entropy": 1.7676224395632745,
      "epoch": 0.4252769965567722,
      "grad_norm": 8.151896476745605,
      "learning_rate": 3.879344214390308e-06,
      "loss": 0.416,
      "mean_token_accuracy": 0.8637456342577934,
      "num_tokens": 165032409.0,
      "step": 137190
    },
    {
      "entropy": 1.883620023727417,
      "epoch": 0.42530799568182187,
      "grad_norm": 7.578486442565918,
      "learning_rate": 3.879202835265633e-06,
      "loss": 0.4462,
      "mean_token_accuracy": 0.8562824308872223,
      "num_tokens": 165044847.0,
      "step": 137200
    },
    {
      "entropy": 1.8436495438218117,
      "epoch": 0.4253389948068716,
      "grad_norm": 8.562501907348633,
      "learning_rate": 3.879061471597127e-06,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8575531423091889,
      "num_tokens": 165058103.0,
      "step": 137210
    },
    {
      "entropy": 1.8444740131497384,
      "epoch": 0.42536999393192126,
      "grad_norm": 7.836605072021484,
      "learning_rate": 3.878920123381976e-06,
      "loss": 0.4366,
      "mean_token_accuracy": 0.8577763974666596,
      "num_tokens": 165070717.0,
      "step": 137220
    },
    {
      "entropy": 1.747816914319992,
      "epoch": 0.425400993056971,
      "grad_norm": 8.393543243408203,
      "learning_rate": 3.878778790617362e-06,
      "loss": 0.3329,
      "mean_token_accuracy": 0.8659502550959587,
      "num_tokens": 165085002.0,
      "step": 137230
    },
    {
      "entropy": 1.9642114371061326,
      "epoch": 0.42543199218202066,
      "grad_norm": 8.948190689086914,
      "learning_rate": 3.87863747330047e-06,
      "loss": 0.5287,
      "mean_token_accuracy": 0.8416478574275971,
      "num_tokens": 165095906.0,
      "step": 137240
    },
    {
      "entropy": 1.8498766794800758,
      "epoch": 0.4254629913070704,
      "grad_norm": 4.120171070098877,
      "learning_rate": 3.8784961714284885e-06,
      "loss": 0.4044,
      "mean_token_accuracy": 0.8605809271335602,
      "num_tokens": 165108454.0,
      "step": 137250
    },
    {
      "entropy": 1.9248084783554078,
      "epoch": 0.42549399043212005,
      "grad_norm": 3.3840432167053223,
      "learning_rate": 3.878354884998603e-06,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8541036680340767,
      "num_tokens": 165120547.0,
      "step": 137260
    },
    {
      "entropy": 1.8695127993822098,
      "epoch": 0.4255249895571698,
      "grad_norm": 7.736443519592285,
      "learning_rate": 3.878213614008001e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8537966668605804,
      "num_tokens": 165132994.0,
      "step": 137270
    },
    {
      "entropy": 1.889861761033535,
      "epoch": 0.42555598868221944,
      "grad_norm": 9.705860137939453,
      "learning_rate": 3.878072358453872e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8553769558668136,
      "num_tokens": 165144856.0,
      "step": 137280
    },
    {
      "entropy": 1.8385322004556657,
      "epoch": 0.4255869878072691,
      "grad_norm": 4.066993236541748,
      "learning_rate": 3.877931118333403e-06,
      "loss": 0.3621,
      "mean_token_accuracy": 0.8684134423732758,
      "num_tokens": 165157887.0,
      "step": 137290
    },
    {
      "entropy": 1.9168552801012992,
      "epoch": 0.42561798693231884,
      "grad_norm": 7.822579383850098,
      "learning_rate": 3.877789893643785e-06,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8533562257885933,
      "num_tokens": 165170433.0,
      "step": 137300
    },
    {
      "entropy": 1.8463567778468133,
      "epoch": 0.4256489860573685,
      "grad_norm": 7.959437847137451,
      "learning_rate": 3.877648684382209e-06,
      "loss": 0.362,
      "mean_token_accuracy": 0.8728025883436203,
      "num_tokens": 165183224.0,
      "step": 137310
    },
    {
      "entropy": 1.896780176460743,
      "epoch": 0.42567998518241823,
      "grad_norm": 3.448885679244995,
      "learning_rate": 3.877507490545866e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8589633658528328,
      "num_tokens": 165194990.0,
      "step": 137320
    },
    {
      "entropy": 1.9273283243179322,
      "epoch": 0.4257109843074679,
      "grad_norm": 8.690948486328125,
      "learning_rate": 3.877366312131946e-06,
      "loss": 0.478,
      "mean_token_accuracy": 0.8396705284714698,
      "num_tokens": 165207611.0,
      "step": 137330
    },
    {
      "entropy": 1.9326757207512855,
      "epoch": 0.4257419834325176,
      "grad_norm": 9.278722763061523,
      "learning_rate": 3.877225149137642e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8490404456853866,
      "num_tokens": 165219648.0,
      "step": 137340
    },
    {
      "entropy": 1.7946316838264464,
      "epoch": 0.4257729825575673,
      "grad_norm": 8.48944091796875,
      "learning_rate": 3.877084001560149e-06,
      "loss": 0.3903,
      "mean_token_accuracy": 0.8692017003893853,
      "num_tokens": 165232876.0,
      "step": 137350
    },
    {
      "entropy": 1.9322822287678718,
      "epoch": 0.425803981682617,
      "grad_norm": 7.394003391265869,
      "learning_rate": 3.87694286939666e-06,
      "loss": 0.4537,
      "mean_token_accuracy": 0.85526312738657,
      "num_tokens": 165244530.0,
      "step": 137360
    },
    {
      "entropy": 1.8217709109187126,
      "epoch": 0.4258349808076667,
      "grad_norm": 8.290609359741211,
      "learning_rate": 3.876801752644371e-06,
      "loss": 0.394,
      "mean_token_accuracy": 0.8580706313252449,
      "num_tokens": 165257526.0,
      "step": 137370
    },
    {
      "entropy": 1.8738437041640281,
      "epoch": 0.4258659799327164,
      "grad_norm": 8.288384437561035,
      "learning_rate": 3.876660651300476e-06,
      "loss": 0.475,
      "mean_token_accuracy": 0.8446152061223984,
      "num_tokens": 165269243.0,
      "step": 137380
    },
    {
      "entropy": 1.8645242124795913,
      "epoch": 0.4258969790577661,
      "grad_norm": 7.856457710266113,
      "learning_rate": 3.876519565362171e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8648913115262985,
      "num_tokens": 165281328.0,
      "step": 137390
    },
    {
      "entropy": 1.8050186708569527,
      "epoch": 0.4259279781828158,
      "grad_norm": 6.527402400970459,
      "learning_rate": 3.876378494826653e-06,
      "loss": 0.3733,
      "mean_token_accuracy": 0.8636491522192955,
      "num_tokens": 165294695.0,
      "step": 137400
    },
    {
      "entropy": 1.88536017537117,
      "epoch": 0.42595897730786547,
      "grad_norm": 2.9641103744506836,
      "learning_rate": 3.8762374396911215e-06,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8582195937633514,
      "num_tokens": 165307479.0,
      "step": 137410
    },
    {
      "entropy": 1.846780201792717,
      "epoch": 0.4259899764329152,
      "grad_norm": 5.190714359283447,
      "learning_rate": 3.876096399952772e-06,
      "loss": 0.3712,
      "mean_token_accuracy": 0.8630301207304001,
      "num_tokens": 165320970.0,
      "step": 137420
    },
    {
      "entropy": 1.9380200251936912,
      "epoch": 0.42602097555796486,
      "grad_norm": 7.597511291503906,
      "learning_rate": 3.875955375608804e-06,
      "loss": 0.5248,
      "mean_token_accuracy": 0.8328088164329529,
      "num_tokens": 165333314.0,
      "step": 137430
    },
    {
      "entropy": 1.9373356685042382,
      "epoch": 0.4260519746830146,
      "grad_norm": 7.61322021484375,
      "learning_rate": 3.875814366656419e-06,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8467313498258591,
      "num_tokens": 165345441.0,
      "step": 137440
    },
    {
      "entropy": 1.849155631661415,
      "epoch": 0.42608297380806426,
      "grad_norm": 8.47956371307373,
      "learning_rate": 3.875673373092818e-06,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8476941645145416,
      "num_tokens": 165358430.0,
      "step": 137450
    },
    {
      "entropy": 1.9534918010234832,
      "epoch": 0.426113972933114,
      "grad_norm": 7.988271236419678,
      "learning_rate": 3.875532394915199e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8520032778382302,
      "num_tokens": 165369269.0,
      "step": 137460
    },
    {
      "entropy": 1.9056456133723259,
      "epoch": 0.42614497205816365,
      "grad_norm": 9.272027969360352,
      "learning_rate": 3.875391432120765e-06,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8481805965304374,
      "num_tokens": 165380985.0,
      "step": 137470
    },
    {
      "entropy": 1.939954537153244,
      "epoch": 0.4261759711832134,
      "grad_norm": 9.503613471984863,
      "learning_rate": 3.87525048470672e-06,
      "loss": 0.5082,
      "mean_token_accuracy": 0.8481825277209282,
      "num_tokens": 165392220.0,
      "step": 137480
    },
    {
      "entropy": 1.9774063229560852,
      "epoch": 0.42620697030826304,
      "grad_norm": 8.539331436157227,
      "learning_rate": 3.875109552670266e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8497935310006142,
      "num_tokens": 165403622.0,
      "step": 137490
    },
    {
      "entropy": 1.9746006906032563,
      "epoch": 0.42623796943331277,
      "grad_norm": 7.517263412475586,
      "learning_rate": 3.874968636008607e-06,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8602147474884987,
      "num_tokens": 165414792.0,
      "step": 137500
    },
    {
      "entropy": 1.9843586146831513,
      "epoch": 0.42626896855836244,
      "grad_norm": 9.263333320617676,
      "learning_rate": 3.874827734718949e-06,
      "loss": 0.5149,
      "mean_token_accuracy": 0.8430792987346649,
      "num_tokens": 165425288.0,
      "step": 137510
    },
    {
      "entropy": 1.8610298484563828,
      "epoch": 0.42629996768341216,
      "grad_norm": 8.659163475036621,
      "learning_rate": 3.8746868487984955e-06,
      "loss": 0.4769,
      "mean_token_accuracy": 0.8577761441469193,
      "num_tokens": 165437939.0,
      "step": 137520
    },
    {
      "entropy": 1.9919404417276383,
      "epoch": 0.42633096680846183,
      "grad_norm": 7.622340202331543,
      "learning_rate": 3.874545978244454e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8462273955345154,
      "num_tokens": 165449331.0,
      "step": 137530
    },
    {
      "entropy": 1.883239607512951,
      "epoch": 0.4263619659335115,
      "grad_norm": 8.540725708007812,
      "learning_rate": 3.87440512305403e-06,
      "loss": 0.4444,
      "mean_token_accuracy": 0.8450183764100074,
      "num_tokens": 165462052.0,
      "step": 137540
    },
    {
      "entropy": 1.761160995066166,
      "epoch": 0.4263929650585612,
      "grad_norm": 3.7234740257263184,
      "learning_rate": 3.874264283224433e-06,
      "loss": 0.341,
      "mean_token_accuracy": 0.877360138297081,
      "num_tokens": 165476174.0,
      "step": 137550
    },
    {
      "entropy": 1.8949896410107612,
      "epoch": 0.4264239641836109,
      "grad_norm": 4.375694274902344,
      "learning_rate": 3.874123458752871e-06,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8546592324972153,
      "num_tokens": 165488036.0,
      "step": 137560
    },
    {
      "entropy": 1.970250654220581,
      "epoch": 0.4264549633086606,
      "grad_norm": 7.3062357902526855,
      "learning_rate": 3.873982649636551e-06,
      "loss": 0.4985,
      "mean_token_accuracy": 0.8475343957543373,
      "num_tokens": 165498774.0,
      "step": 137570
    },
    {
      "entropy": 1.931829509139061,
      "epoch": 0.4264859624337103,
      "grad_norm": 8.701257705688477,
      "learning_rate": 3.8738418558726845e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8460123270750046,
      "num_tokens": 165510856.0,
      "step": 137580
    },
    {
      "entropy": 1.9778954088687897,
      "epoch": 0.42651696155876,
      "grad_norm": 10.640881538391113,
      "learning_rate": 3.873701077458481e-06,
      "loss": 0.5176,
      "mean_token_accuracy": 0.8416032001376152,
      "num_tokens": 165521684.0,
      "step": 137590
    },
    {
      "entropy": 1.871251691877842,
      "epoch": 0.4265479606838097,
      "grad_norm": 7.633983612060547,
      "learning_rate": 3.8735603143911525e-06,
      "loss": 0.418,
      "mean_token_accuracy": 0.8568643853068352,
      "num_tokens": 165533887.0,
      "step": 137600
    },
    {
      "entropy": 1.9317373171448708,
      "epoch": 0.4265789598088594,
      "grad_norm": 7.2197465896606445,
      "learning_rate": 3.873419566667911e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.845657354593277,
      "num_tokens": 165545740.0,
      "step": 137610
    },
    {
      "entropy": 1.8812014043331147,
      "epoch": 0.4266099589339091,
      "grad_norm": 8.381542205810547,
      "learning_rate": 3.873278834285967e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8589608117938041,
      "num_tokens": 165557441.0,
      "step": 137620
    },
    {
      "entropy": 1.9113554194569589,
      "epoch": 0.4266409580589588,
      "grad_norm": 8.038044929504395,
      "learning_rate": 3.8731381172425355e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8493774205446243,
      "num_tokens": 165568993.0,
      "step": 137630
    },
    {
      "entropy": 1.891987682878971,
      "epoch": 0.42667195718400847,
      "grad_norm": 8.052876472473145,
      "learning_rate": 3.87299741553483e-06,
      "loss": 0.4313,
      "mean_token_accuracy": 0.8627904951572418,
      "num_tokens": 165580905.0,
      "step": 137640
    },
    {
      "entropy": 1.9532914757728577,
      "epoch": 0.4267029563090582,
      "grad_norm": 8.188140869140625,
      "learning_rate": 3.872856729160065e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8446842655539513,
      "num_tokens": 165591744.0,
      "step": 137650
    },
    {
      "entropy": 1.9147072404623031,
      "epoch": 0.42673395543410786,
      "grad_norm": 7.08876895904541,
      "learning_rate": 3.872716058115456e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8530980065464974,
      "num_tokens": 165603429.0,
      "step": 137660
    },
    {
      "entropy": 1.950710503757,
      "epoch": 0.4267649545591576,
      "grad_norm": 9.763349533081055,
      "learning_rate": 3.87257540239822e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8458059638738632,
      "num_tokens": 165614886.0,
      "step": 137670
    },
    {
      "entropy": 1.8922182023525238,
      "epoch": 0.42679595368420725,
      "grad_norm": 7.75042724609375,
      "learning_rate": 3.872434762005572e-06,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8595366209745408,
      "num_tokens": 165626421.0,
      "step": 137680
    },
    {
      "entropy": 1.9189585834741592,
      "epoch": 0.426826952809257,
      "grad_norm": 5.677242755889893,
      "learning_rate": 3.872294136934731e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8460530653595925,
      "num_tokens": 165638484.0,
      "step": 137690
    },
    {
      "entropy": 1.8581395775079728,
      "epoch": 0.42685795193430665,
      "grad_norm": 8.89156723022461,
      "learning_rate": 3.872153527182914e-06,
      "loss": 0.4251,
      "mean_token_accuracy": 0.851922070980072,
      "num_tokens": 165652365.0,
      "step": 137700
    },
    {
      "entropy": 1.9807763159275056,
      "epoch": 0.42688895105935637,
      "grad_norm": 9.592456817626953,
      "learning_rate": 3.87201293274734e-06,
      "loss": 0.5234,
      "mean_token_accuracy": 0.8444359451532364,
      "num_tokens": 165663039.0,
      "step": 137710
    },
    {
      "entropy": 1.9604480162262916,
      "epoch": 0.42691995018440604,
      "grad_norm": 6.907299518585205,
      "learning_rate": 3.87187235362523e-06,
      "loss": 0.5137,
      "mean_token_accuracy": 0.8401239201426506,
      "num_tokens": 165674987.0,
      "step": 137720
    },
    {
      "entropy": 1.8310967803001403,
      "epoch": 0.42695094930945576,
      "grad_norm": 7.537128925323486,
      "learning_rate": 3.871731789813803e-06,
      "loss": 0.3615,
      "mean_token_accuracy": 0.8688787788152694,
      "num_tokens": 165688212.0,
      "step": 137730
    },
    {
      "entropy": 1.8531201511621476,
      "epoch": 0.42698194843450543,
      "grad_norm": 3.874274253845215,
      "learning_rate": 3.87159124131028e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8694923147559166,
      "num_tokens": 165700648.0,
      "step": 137740
    },
    {
      "entropy": 1.8400910899043084,
      "epoch": 0.42701294755955516,
      "grad_norm": 8.50154972076416,
      "learning_rate": 3.871450708111883e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8610082730650902,
      "num_tokens": 165713968.0,
      "step": 137750
    },
    {
      "entropy": 1.9427133277058601,
      "epoch": 0.4270439466846048,
      "grad_norm": 9.070932388305664,
      "learning_rate": 3.8713101902158354e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8499548882246017,
      "num_tokens": 165725879.0,
      "step": 137760
    },
    {
      "entropy": 1.913887146115303,
      "epoch": 0.42707494580965455,
      "grad_norm": 7.201402187347412,
      "learning_rate": 3.871169687619359e-06,
      "loss": 0.4384,
      "mean_token_accuracy": 0.8556391328573227,
      "num_tokens": 165737902.0,
      "step": 137770
    },
    {
      "entropy": 1.8545256823301315,
      "epoch": 0.4271059449347042,
      "grad_norm": 7.537273406982422,
      "learning_rate": 3.871029200319679e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8587723672389984,
      "num_tokens": 165750054.0,
      "step": 137780
    },
    {
      "entropy": 1.938425388932228,
      "epoch": 0.4271369440597539,
      "grad_norm": 8.340550422668457,
      "learning_rate": 3.870888728314018e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8487496972084045,
      "num_tokens": 165760805.0,
      "step": 137790
    },
    {
      "entropy": 1.7333957627415657,
      "epoch": 0.4271679431848036,
      "grad_norm": 8.770150184631348,
      "learning_rate": 3.870748271599602e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8617977604269982,
      "num_tokens": 165774883.0,
      "step": 137800
    },
    {
      "entropy": 1.924593235552311,
      "epoch": 0.4271989423098533,
      "grad_norm": 4.3831400871276855,
      "learning_rate": 3.870607830173659e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8520621985197068,
      "num_tokens": 165786764.0,
      "step": 137810
    },
    {
      "entropy": 1.856243497133255,
      "epoch": 0.427229941434903,
      "grad_norm": 8.254294395446777,
      "learning_rate": 3.8704674040334124e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8571957781910896,
      "num_tokens": 165799501.0,
      "step": 137820
    },
    {
      "entropy": 1.9511059790849685,
      "epoch": 0.4272609405599527,
      "grad_norm": 7.887767314910889,
      "learning_rate": 3.8703269931760905e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8668577119708061,
      "num_tokens": 165810071.0,
      "step": 137830
    },
    {
      "entropy": 1.854556131362915,
      "epoch": 0.4272919396850024,
      "grad_norm": 7.153961658477783,
      "learning_rate": 3.870186597598924e-06,
      "loss": 0.4058,
      "mean_token_accuracy": 0.8615784257650375,
      "num_tokens": 165822533.0,
      "step": 137840
    },
    {
      "entropy": 1.8858709022402764,
      "epoch": 0.42732293881005207,
      "grad_norm": 3.6662893295288086,
      "learning_rate": 3.870046217299139e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8626757502555847,
      "num_tokens": 165834055.0,
      "step": 137850
    },
    {
      "entropy": 1.977617959678173,
      "epoch": 0.4273539379351018,
      "grad_norm": 10.866979598999023,
      "learning_rate": 3.869905852273965e-06,
      "loss": 0.5671,
      "mean_token_accuracy": 0.8238250657916069,
      "num_tokens": 165846081.0,
      "step": 137860
    },
    {
      "entropy": 1.808312650024891,
      "epoch": 0.42738493706015146,
      "grad_norm": 3.0479211807250977,
      "learning_rate": 3.869765502520633e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8619560331106186,
      "num_tokens": 165858762.0,
      "step": 137870
    },
    {
      "entropy": 1.966909298300743,
      "epoch": 0.4274159361852012,
      "grad_norm": 8.616495132446289,
      "learning_rate": 3.869625168036374e-06,
      "loss": 0.5079,
      "mean_token_accuracy": 0.8432893455028534,
      "num_tokens": 165869711.0,
      "step": 137880
    },
    {
      "entropy": 1.8776897192001343,
      "epoch": 0.42744693531025085,
      "grad_norm": 5.014652252197266,
      "learning_rate": 3.8694848488184185e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8473319381475448,
      "num_tokens": 165882325.0,
      "step": 137890
    },
    {
      "entropy": 1.9359253600239754,
      "epoch": 0.4274779344353006,
      "grad_norm": 9.926865577697754,
      "learning_rate": 3.869344544864e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8564401522278786,
      "num_tokens": 165893416.0,
      "step": 137900
    },
    {
      "entropy": 1.9416421443223952,
      "epoch": 0.42750893356035025,
      "grad_norm": 4.446148872375488,
      "learning_rate": 3.869204256170351e-06,
      "loss": 0.5114,
      "mean_token_accuracy": 0.8427053913474083,
      "num_tokens": 165904568.0,
      "step": 137910
    },
    {
      "entropy": 1.8732975110411645,
      "epoch": 0.42753993268539997,
      "grad_norm": 7.7177019119262695,
      "learning_rate": 3.869063982734706e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.860724626481533,
      "num_tokens": 165916369.0,
      "step": 137920
    },
    {
      "entropy": 1.8843568712472916,
      "epoch": 0.42757093181044964,
      "grad_norm": 8.525490760803223,
      "learning_rate": 3.868923724554298e-06,
      "loss": 0.4442,
      "mean_token_accuracy": 0.8500266253948212,
      "num_tokens": 165928150.0,
      "step": 137930
    },
    {
      "entropy": 1.9407880812883378,
      "epoch": 0.42760193093549936,
      "grad_norm": 8.2413969039917,
      "learning_rate": 3.868783481626363e-06,
      "loss": 0.4822,
      "mean_token_accuracy": 0.850470757484436,
      "num_tokens": 165938815.0,
      "step": 137940
    },
    {
      "entropy": 1.9635769337415696,
      "epoch": 0.42763293006054903,
      "grad_norm": 8.470434188842773,
      "learning_rate": 3.868643253948136e-06,
      "loss": 0.4737,
      "mean_token_accuracy": 0.853185373544693,
      "num_tokens": 165949479.0,
      "step": 137950
    },
    {
      "entropy": 1.778677401691675,
      "epoch": 0.42766392918559876,
      "grad_norm": 9.556220054626465,
      "learning_rate": 3.868503041516854e-06,
      "loss": 0.3995,
      "mean_token_accuracy": 0.8522502809762955,
      "num_tokens": 165962742.0,
      "step": 137960
    },
    {
      "entropy": 1.9046275839209557,
      "epoch": 0.4276949283106484,
      "grad_norm": 8.775141716003418,
      "learning_rate": 3.8683628443297554e-06,
      "loss": 0.5006,
      "mean_token_accuracy": 0.8469727069139481,
      "num_tokens": 165974044.0,
      "step": 137970
    },
    {
      "entropy": 1.9093778878450394,
      "epoch": 0.42772592743569815,
      "grad_norm": 4.199886798858643,
      "learning_rate": 3.868222662384076e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8507871389389038,
      "num_tokens": 165985295.0,
      "step": 137980
    },
    {
      "entropy": 1.858989103138447,
      "epoch": 0.4277569265607478,
      "grad_norm": 8.536810874938965,
      "learning_rate": 3.868082495677056e-06,
      "loss": 0.4509,
      "mean_token_accuracy": 0.8496241062879563,
      "num_tokens": 165998258.0,
      "step": 137990
    },
    {
      "entropy": 1.8789437100291253,
      "epoch": 0.42778792568579754,
      "grad_norm": 8.451990127563477,
      "learning_rate": 3.867942344205934e-06,
      "loss": 0.4667,
      "mean_token_accuracy": 0.8546844542026519,
      "num_tokens": 166009620.0,
      "step": 138000
    },
    {
      "entropy": 1.8833610072731972,
      "epoch": 0.4278189248108472,
      "grad_norm": 7.35070276260376,
      "learning_rate": 3.867802207967949e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8602321013808251,
      "num_tokens": 166021208.0,
      "step": 138010
    },
    {
      "entropy": 1.8281296089291572,
      "epoch": 0.42784992393589694,
      "grad_norm": 7.972667694091797,
      "learning_rate": 3.867662086960344e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8570877268910408,
      "num_tokens": 166033804.0,
      "step": 138020
    },
    {
      "entropy": 1.7907851234078407,
      "epoch": 0.4278809230609466,
      "grad_norm": 8.16515827178955,
      "learning_rate": 3.8675219811803586e-06,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8523140147328376,
      "num_tokens": 166047092.0,
      "step": 138030
    },
    {
      "entropy": 1.7590288013219832,
      "epoch": 0.4279119221859963,
      "grad_norm": 3.805769443511963,
      "learning_rate": 3.8673818906252354e-06,
      "loss": 0.349,
      "mean_token_accuracy": 0.8651872128248215,
      "num_tokens": 166060694.0,
      "step": 138040
    },
    {
      "entropy": 1.8908022806048392,
      "epoch": 0.427942921311046,
      "grad_norm": 3.611936092376709,
      "learning_rate": 3.867241815292218e-06,
      "loss": 0.4588,
      "mean_token_accuracy": 0.851640222966671,
      "num_tokens": 166072485.0,
      "step": 138050
    },
    {
      "entropy": 1.9297591403126717,
      "epoch": 0.42797392043609567,
      "grad_norm": 7.440158843994141,
      "learning_rate": 3.867101755178548e-06,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8443147346377373,
      "num_tokens": 166083693.0,
      "step": 138060
    },
    {
      "entropy": 1.9145255535840988,
      "epoch": 0.4280049195611454,
      "grad_norm": 7.726010799407959,
      "learning_rate": 3.866961710281471e-06,
      "loss": 0.4287,
      "mean_token_accuracy": 0.8665438875555992,
      "num_tokens": 166095128.0,
      "step": 138070
    },
    {
      "entropy": 1.8834168136119842,
      "epoch": 0.42803591868619506,
      "grad_norm": 8.636183738708496,
      "learning_rate": 3.866821680598232e-06,
      "loss": 0.4411,
      "mean_token_accuracy": 0.8588782355189324,
      "num_tokens": 166107020.0,
      "step": 138080
    },
    {
      "entropy": 1.890410417318344,
      "epoch": 0.4280669178112448,
      "grad_norm": 9.407487869262695,
      "learning_rate": 3.8666816661260766e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8558942392468453,
      "num_tokens": 166118533.0,
      "step": 138090
    },
    {
      "entropy": 1.895900882035494,
      "epoch": 0.42809791693629445,
      "grad_norm": 9.499615669250488,
      "learning_rate": 3.86654166686225e-06,
      "loss": 0.4494,
      "mean_token_accuracy": 0.8502351224422455,
      "num_tokens": 166130282.0,
      "step": 138100
    },
    {
      "entropy": 1.899457646906376,
      "epoch": 0.4281289160613442,
      "grad_norm": 8.837594032287598,
      "learning_rate": 3.866401682804001e-06,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8482611507177353,
      "num_tokens": 166142276.0,
      "step": 138110
    },
    {
      "entropy": 1.9217953234910965,
      "epoch": 0.42815991518639385,
      "grad_norm": 8.12231731414795,
      "learning_rate": 3.866261713948576e-06,
      "loss": 0.5005,
      "mean_token_accuracy": 0.847544614970684,
      "num_tokens": 166154311.0,
      "step": 138120
    },
    {
      "entropy": 1.975175791978836,
      "epoch": 0.42819091431144357,
      "grad_norm": 7.083240509033203,
      "learning_rate": 3.866121760293223e-06,
      "loss": 0.5289,
      "mean_token_accuracy": 0.8364548414945603,
      "num_tokens": 166165093.0,
      "step": 138130
    },
    {
      "entropy": 1.9826736554503441,
      "epoch": 0.42822191343649324,
      "grad_norm": 9.184717178344727,
      "learning_rate": 3.865981821835192e-06,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8572745114564896,
      "num_tokens": 166176186.0,
      "step": 138140
    },
    {
      "entropy": 1.8134810969233512,
      "epoch": 0.42825291256154296,
      "grad_norm": 9.861612319946289,
      "learning_rate": 3.8658418985717325e-06,
      "loss": 0.4013,
      "mean_token_accuracy": 0.8668722435832024,
      "num_tokens": 166188879.0,
      "step": 138150
    },
    {
      "entropy": 1.884695628285408,
      "epoch": 0.42828391168659263,
      "grad_norm": 7.422934532165527,
      "learning_rate": 3.865701990500095e-06,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8617542907595634,
      "num_tokens": 166201719.0,
      "step": 138160
    },
    {
      "entropy": 1.9591299802064897,
      "epoch": 0.42831491081164236,
      "grad_norm": 7.2439165115356445,
      "learning_rate": 3.865562097617531e-06,
      "loss": 0.4762,
      "mean_token_accuracy": 0.8428146690130234,
      "num_tokens": 166213159.0,
      "step": 138170
    },
    {
      "entropy": 1.9952568769454957,
      "epoch": 0.428345909936692,
      "grad_norm": 8.97078800201416,
      "learning_rate": 3.865422219921291e-06,
      "loss": 0.51,
      "mean_token_accuracy": 0.8497517094016075,
      "num_tokens": 166224006.0,
      "step": 138180
    },
    {
      "entropy": 1.924196371436119,
      "epoch": 0.42837690906174175,
      "grad_norm": 7.9839186668396,
      "learning_rate": 3.865282357408629e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.84932641685009,
      "num_tokens": 166235138.0,
      "step": 138190
    },
    {
      "entropy": 1.8705265656113625,
      "epoch": 0.4284079081867914,
      "grad_norm": 8.166479110717773,
      "learning_rate": 3.8651425100767995e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8603829473257065,
      "num_tokens": 166246952.0,
      "step": 138200
    },
    {
      "entropy": 1.8786292031407357,
      "epoch": 0.42843890731184114,
      "grad_norm": 3.4489612579345703,
      "learning_rate": 3.865002677923053e-06,
      "loss": 0.3917,
      "mean_token_accuracy": 0.8639238491654396,
      "num_tokens": 166259522.0,
      "step": 138210
    },
    {
      "entropy": 1.924619671702385,
      "epoch": 0.4284699064368908,
      "grad_norm": 7.831303596496582,
      "learning_rate": 3.864862860944646e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8498168155550957,
      "num_tokens": 166270568.0,
      "step": 138220
    },
    {
      "entropy": 1.8376672357320785,
      "epoch": 0.42850090556194054,
      "grad_norm": 8.016936302185059,
      "learning_rate": 3.864723059138835e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8555368795990944,
      "num_tokens": 166284006.0,
      "step": 138230
    },
    {
      "entropy": 1.8461035266518593,
      "epoch": 0.4285319046869902,
      "grad_norm": 7.379615783691406,
      "learning_rate": 3.864583272502873e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.8619070097804069,
      "num_tokens": 166296817.0,
      "step": 138240
    },
    {
      "entropy": 1.8854956120252608,
      "epoch": 0.42856290381203993,
      "grad_norm": 3.884890079498291,
      "learning_rate": 3.86444350103402e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8470316350460052,
      "num_tokens": 166309348.0,
      "step": 138250
    },
    {
      "entropy": 1.7654275074601173,
      "epoch": 0.4285939029370896,
      "grad_norm": 9.349808692932129,
      "learning_rate": 3.864303744729532e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.8671312019228935,
      "num_tokens": 166322832.0,
      "step": 138260
    },
    {
      "entropy": 1.8022211775183679,
      "epoch": 0.4286249020621393,
      "grad_norm": 9.219120979309082,
      "learning_rate": 3.864164003586666e-06,
      "loss": 0.3904,
      "mean_token_accuracy": 0.8593027204275131,
      "num_tokens": 166335905.0,
      "step": 138270
    },
    {
      "entropy": 1.8638395503163339,
      "epoch": 0.428655901187189,
      "grad_norm": 7.9525628089904785,
      "learning_rate": 3.864024277602683e-06,
      "loss": 0.4317,
      "mean_token_accuracy": 0.8568304657936097,
      "num_tokens": 166348632.0,
      "step": 138280
    },
    {
      "entropy": 1.954586958885193,
      "epoch": 0.42868690031223866,
      "grad_norm": 4.062179088592529,
      "learning_rate": 3.863884566774842e-06,
      "loss": 0.4813,
      "mean_token_accuracy": 0.8409469664096832,
      "num_tokens": 166360316.0,
      "step": 138290
    },
    {
      "entropy": 1.9421510837972165,
      "epoch": 0.4287178994372884,
      "grad_norm": 2.5580577850341797,
      "learning_rate": 3.863744871100402e-06,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8459225296974182,
      "num_tokens": 166372255.0,
      "step": 138300
    },
    {
      "entropy": 1.9440967753529548,
      "epoch": 0.42874889856233805,
      "grad_norm": 8.030793190002441,
      "learning_rate": 3.863605190576625e-06,
      "loss": 0.4716,
      "mean_token_accuracy": 0.8478009730577469,
      "num_tokens": 166383633.0,
      "step": 138310
    },
    {
      "entropy": 1.9596870571374894,
      "epoch": 0.4287798976873878,
      "grad_norm": 8.272663116455078,
      "learning_rate": 3.863465525200771e-06,
      "loss": 0.5012,
      "mean_token_accuracy": 0.855411772429943,
      "num_tokens": 166393996.0,
      "step": 138320
    },
    {
      "entropy": 1.9576293110847474,
      "epoch": 0.42881089681243745,
      "grad_norm": 8.87380313873291,
      "learning_rate": 3.863325874970105e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8526660829782486,
      "num_tokens": 166405246.0,
      "step": 138330
    },
    {
      "entropy": 1.857802005112171,
      "epoch": 0.42884189593748717,
      "grad_norm": 7.251513957977295,
      "learning_rate": 3.863186239881888e-06,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8620542094111443,
      "num_tokens": 166416790.0,
      "step": 138340
    },
    {
      "entropy": 1.888051538169384,
      "epoch": 0.42887289506253684,
      "grad_norm": 5.157383441925049,
      "learning_rate": 3.863046619933384e-06,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8537735804915428,
      "num_tokens": 166429648.0,
      "step": 138350
    },
    {
      "entropy": 1.8693744093179703,
      "epoch": 0.42890389418758657,
      "grad_norm": 3.8320672512054443,
      "learning_rate": 3.862907015121856e-06,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8648984044790268,
      "num_tokens": 166441105.0,
      "step": 138360
    },
    {
      "entropy": 1.993468850851059,
      "epoch": 0.42893489331263623,
      "grad_norm": 7.150916576385498,
      "learning_rate": 3.8627674254445724e-06,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8378593668341636,
      "num_tokens": 166452361.0,
      "step": 138370
    },
    {
      "entropy": 1.9181072056293487,
      "epoch": 0.42896589243768596,
      "grad_norm": 7.089375972747803,
      "learning_rate": 3.862627850898797e-06,
      "loss": 0.5196,
      "mean_token_accuracy": 0.8403525799512863,
      "num_tokens": 166463855.0,
      "step": 138380
    },
    {
      "entropy": 1.967288474738598,
      "epoch": 0.4289968915627356,
      "grad_norm": 7.978488445281982,
      "learning_rate": 3.8624882914817964e-06,
      "loss": 0.5225,
      "mean_token_accuracy": 0.838663375377655,
      "num_tokens": 166475665.0,
      "step": 138390
    },
    {
      "entropy": 1.8494392395019532,
      "epoch": 0.42902789068778535,
      "grad_norm": 8.566108703613281,
      "learning_rate": 3.8623487471908375e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.844138278067112,
      "num_tokens": 166488310.0,
      "step": 138400
    },
    {
      "entropy": 1.8647007137537002,
      "epoch": 0.429058889812835,
      "grad_norm": 8.269538879394531,
      "learning_rate": 3.862209218023188e-06,
      "loss": 0.4146,
      "mean_token_accuracy": 0.8606134533882142,
      "num_tokens": 166500577.0,
      "step": 138410
    },
    {
      "entropy": 1.8854542568325996,
      "epoch": 0.42908988893788474,
      "grad_norm": 8.041147232055664,
      "learning_rate": 3.862069703976116e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8451815471053123,
      "num_tokens": 166512317.0,
      "step": 138420
    },
    {
      "entropy": 1.885179491341114,
      "epoch": 0.4291208880629344,
      "grad_norm": 7.563552379608154,
      "learning_rate": 3.861930205046893e-06,
      "loss": 0.425,
      "mean_token_accuracy": 0.8753857642412186,
      "num_tokens": 166524119.0,
      "step": 138430
    },
    {
      "entropy": 1.8435839340090752,
      "epoch": 0.42915188718798414,
      "grad_norm": 3.8470942974090576,
      "learning_rate": 3.861790721232786e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.865519005060196,
      "num_tokens": 166536053.0,
      "step": 138440
    },
    {
      "entropy": 1.943047122657299,
      "epoch": 0.4291828863130338,
      "grad_norm": 7.736998558044434,
      "learning_rate": 3.861651252531068e-06,
      "loss": 0.4898,
      "mean_token_accuracy": 0.8467626482248306,
      "num_tokens": 166547902.0,
      "step": 138450
    },
    {
      "entropy": 1.9400530129671096,
      "epoch": 0.42921388543808353,
      "grad_norm": 7.892675399780273,
      "learning_rate": 3.861511798939009e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8521466955542565,
      "num_tokens": 166559133.0,
      "step": 138460
    },
    {
      "entropy": 1.8419177174568175,
      "epoch": 0.4292448845631332,
      "grad_norm": 8.258288383483887,
      "learning_rate": 3.8613723604538814e-06,
      "loss": 0.4093,
      "mean_token_accuracy": 0.8595252349972725,
      "num_tokens": 166571712.0,
      "step": 138470
    },
    {
      "entropy": 1.918247513473034,
      "epoch": 0.4292758836881829,
      "grad_norm": 8.431722640991211,
      "learning_rate": 3.8612329370729566e-06,
      "loss": 0.4883,
      "mean_token_accuracy": 0.8501754224300384,
      "num_tokens": 166583492.0,
      "step": 138480
    },
    {
      "entropy": 1.922251921892166,
      "epoch": 0.4293068828132326,
      "grad_norm": 7.291534423828125,
      "learning_rate": 3.861093528793509e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8490281432867051,
      "num_tokens": 166595760.0,
      "step": 138490
    },
    {
      "entropy": 1.9234865739941598,
      "epoch": 0.4293378819382823,
      "grad_norm": 8.550054550170898,
      "learning_rate": 3.860954135612814e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8617073327302933,
      "num_tokens": 166607418.0,
      "step": 138500
    },
    {
      "entropy": 1.9898644000291825,
      "epoch": 0.429368881063332,
      "grad_norm": 7.687295913696289,
      "learning_rate": 3.860814757528143e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8465102508664131,
      "num_tokens": 166618579.0,
      "step": 138510
    },
    {
      "entropy": 1.9373308807611465,
      "epoch": 0.42939988018838166,
      "grad_norm": 8.360345840454102,
      "learning_rate": 3.8606753945367746e-06,
      "loss": 0.4598,
      "mean_token_accuracy": 0.849955253303051,
      "num_tokens": 166629584.0,
      "step": 138520
    },
    {
      "entropy": 1.8768581017851829,
      "epoch": 0.4294308793134314,
      "grad_norm": 8.298843383789062,
      "learning_rate": 3.860536046635983e-06,
      "loss": 0.4401,
      "mean_token_accuracy": 0.8533497661352157,
      "num_tokens": 166641699.0,
      "step": 138530
    },
    {
      "entropy": 1.8640162199735641,
      "epoch": 0.42946187843848105,
      "grad_norm": 8.439657211303711,
      "learning_rate": 3.8603967138230456e-06,
      "loss": 0.4794,
      "mean_token_accuracy": 0.8546614408493042,
      "num_tokens": 166654557.0,
      "step": 138540
    },
    {
      "entropy": 1.7630941659212112,
      "epoch": 0.4294928775635308,
      "grad_norm": 4.3905744552612305,
      "learning_rate": 3.8602573960952405e-06,
      "loss": 0.3375,
      "mean_token_accuracy": 0.8672227919101715,
      "num_tokens": 166669057.0,
      "step": 138550
    },
    {
      "entropy": 1.883773235976696,
      "epoch": 0.42952387668858044,
      "grad_norm": 8.244091987609863,
      "learning_rate": 3.860118093449845e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8554739341139793,
      "num_tokens": 166680914.0,
      "step": 138560
    },
    {
      "entropy": 1.9457030475139618,
      "epoch": 0.42955487581363017,
      "grad_norm": 9.337789535522461,
      "learning_rate": 3.8599788058841385e-06,
      "loss": 0.4938,
      "mean_token_accuracy": 0.8439801335334778,
      "num_tokens": 166692421.0,
      "step": 138570
    },
    {
      "entropy": 1.8535506889224052,
      "epoch": 0.42958587493867983,
      "grad_norm": 10.059792518615723,
      "learning_rate": 3.859839533395399e-06,
      "loss": 0.3947,
      "mean_token_accuracy": 0.8560854658484459,
      "num_tokens": 166704599.0,
      "step": 138580
    },
    {
      "entropy": 1.9137897729873656,
      "epoch": 0.42961687406372956,
      "grad_norm": 8.756423950195312,
      "learning_rate": 3.859700275980909e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8534789383411407,
      "num_tokens": 166716671.0,
      "step": 138590
    },
    {
      "entropy": 1.8182189181447028,
      "epoch": 0.42964787318877923,
      "grad_norm": 4.194288730621338,
      "learning_rate": 3.859561033637948e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8575180351734162,
      "num_tokens": 166729546.0,
      "step": 138600
    },
    {
      "entropy": 1.9256513655185699,
      "epoch": 0.42967887231382895,
      "grad_norm": 8.066496849060059,
      "learning_rate": 3.859421806363798e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8577398300170899,
      "num_tokens": 166741327.0,
      "step": 138610
    },
    {
      "entropy": 1.9436537489295005,
      "epoch": 0.4297098714388786,
      "grad_norm": 8.44263744354248,
      "learning_rate": 3.859282594155741e-06,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8500909224152565,
      "num_tokens": 166752956.0,
      "step": 138620
    },
    {
      "entropy": 1.8246969357132912,
      "epoch": 0.42974087056392835,
      "grad_norm": 7.600648880004883,
      "learning_rate": 3.859143397011061e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.8686912745237351,
      "num_tokens": 166766150.0,
      "step": 138630
    },
    {
      "entropy": 1.8569831773638725,
      "epoch": 0.429771869688978,
      "grad_norm": 3.839282989501953,
      "learning_rate": 3.85900421492704e-06,
      "loss": 0.4299,
      "mean_token_accuracy": 0.8597505047917366,
      "num_tokens": 166779296.0,
      "step": 138640
    },
    {
      "entropy": 1.9191448912024498,
      "epoch": 0.42980286881402774,
      "grad_norm": 8.564552307128906,
      "learning_rate": 3.858865047900964e-06,
      "loss": 0.5094,
      "mean_token_accuracy": 0.8487473547458648,
      "num_tokens": 166790635.0,
      "step": 138650
    },
    {
      "entropy": 1.9488032266497612,
      "epoch": 0.4298338679390774,
      "grad_norm": 7.2074360847473145,
      "learning_rate": 3.858725895930116e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8469724044203758,
      "num_tokens": 166802382.0,
      "step": 138660
    },
    {
      "entropy": 1.9374908640980721,
      "epoch": 0.42986486706412713,
      "grad_norm": 8.300436973571777,
      "learning_rate": 3.8585867590117845e-06,
      "loss": 0.445,
      "mean_token_accuracy": 0.8558620125055313,
      "num_tokens": 166813964.0,
      "step": 138670
    },
    {
      "entropy": 1.9663563668727875,
      "epoch": 0.4298958661891768,
      "grad_norm": 9.31306266784668,
      "learning_rate": 3.858447637143254e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8509336993098259,
      "num_tokens": 166825330.0,
      "step": 138680
    },
    {
      "entropy": 1.944647277891636,
      "epoch": 0.4299268653142265,
      "grad_norm": 9.174337387084961,
      "learning_rate": 3.858308530321811e-06,
      "loss": 0.4679,
      "mean_token_accuracy": 0.8518739119172096,
      "num_tokens": 166836729.0,
      "step": 138690
    },
    {
      "entropy": 1.9004219979047776,
      "epoch": 0.4299578644392762,
      "grad_norm": 9.097691535949707,
      "learning_rate": 3.858169438544745e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8654677256941795,
      "num_tokens": 166848221.0,
      "step": 138700
    },
    {
      "entropy": 1.905510926246643,
      "epoch": 0.4299888635643259,
      "grad_norm": 7.829078674316406,
      "learning_rate": 3.858030361809342e-06,
      "loss": 0.4285,
      "mean_token_accuracy": 0.8583436101675034,
      "num_tokens": 166860834.0,
      "step": 138710
    },
    {
      "entropy": 2.0032723784446715,
      "epoch": 0.4300198626893756,
      "grad_norm": 9.5745267868042,
      "learning_rate": 3.857891300112894e-06,
      "loss": 0.5005,
      "mean_token_accuracy": 0.845015361905098,
      "num_tokens": 166871860.0,
      "step": 138720
    },
    {
      "entropy": 1.9505044639110565,
      "epoch": 0.4300508618144253,
      "grad_norm": 8.139888763427734,
      "learning_rate": 3.85775225345269e-06,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8441401451826096,
      "num_tokens": 166883775.0,
      "step": 138730
    },
    {
      "entropy": 1.996666456758976,
      "epoch": 0.430081860939475,
      "grad_norm": 8.10516357421875,
      "learning_rate": 3.857613221826021e-06,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8445516094565392,
      "num_tokens": 166894824.0,
      "step": 138740
    },
    {
      "entropy": 1.9581088334321977,
      "epoch": 0.4301128600645247,
      "grad_norm": 8.987552642822266,
      "learning_rate": 3.8574742052301755e-06,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8530200630426407,
      "num_tokens": 166906284.0,
      "step": 138750
    },
    {
      "entropy": 1.8697850778698921,
      "epoch": 0.4301438591895744,
      "grad_norm": 7.841629981994629,
      "learning_rate": 3.857335203662448e-06,
      "loss": 0.4098,
      "mean_token_accuracy": 0.8616936087608338,
      "num_tokens": 166919062.0,
      "step": 138760
    },
    {
      "entropy": 1.8914014741778373,
      "epoch": 0.43017485831462404,
      "grad_norm": 3.872026205062866,
      "learning_rate": 3.85719621712013e-06,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8455238342285156,
      "num_tokens": 166930561.0,
      "step": 138770
    },
    {
      "entropy": 1.8692807108163834,
      "epoch": 0.43020585743967377,
      "grad_norm": 2.8080806732177734,
      "learning_rate": 3.857057245600515e-06,
      "loss": 0.3985,
      "mean_token_accuracy": 0.8588801845908165,
      "num_tokens": 166944394.0,
      "step": 138780
    },
    {
      "entropy": 1.8834291696548462,
      "epoch": 0.43023685656472344,
      "grad_norm": 8.107061386108398,
      "learning_rate": 3.856918289100897e-06,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8372646197676659,
      "num_tokens": 166957123.0,
      "step": 138790
    },
    {
      "entropy": 1.9186320677399635,
      "epoch": 0.43026785568977316,
      "grad_norm": 6.207789421081543,
      "learning_rate": 3.856779347618569e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8523793607950211,
      "num_tokens": 166969149.0,
      "step": 138800
    },
    {
      "entropy": 1.9312793225049973,
      "epoch": 0.43029885481482283,
      "grad_norm": 9.400903701782227,
      "learning_rate": 3.85664042115083e-06,
      "loss": 0.4098,
      "mean_token_accuracy": 0.8667648211121559,
      "num_tokens": 166981732.0,
      "step": 138810
    },
    {
      "entropy": 1.7980728089809417,
      "epoch": 0.43032985393987255,
      "grad_norm": 5.016563415527344,
      "learning_rate": 3.8565015096949724e-06,
      "loss": 0.3996,
      "mean_token_accuracy": 0.8642199486494064,
      "num_tokens": 166995822.0,
      "step": 138820
    },
    {
      "entropy": 1.8924931958317757,
      "epoch": 0.4303608530649222,
      "grad_norm": 8.750584602355957,
      "learning_rate": 3.856362613248295e-06,
      "loss": 0.4858,
      "mean_token_accuracy": 0.8461324706673622,
      "num_tokens": 167008749.0,
      "step": 138830
    },
    {
      "entropy": 1.840745933353901,
      "epoch": 0.43039185218997195,
      "grad_norm": 10.046074867248535,
      "learning_rate": 3.856223731808094e-06,
      "loss": 0.4145,
      "mean_token_accuracy": 0.8469096630811691,
      "num_tokens": 167022329.0,
      "step": 138840
    },
    {
      "entropy": 1.8810099840164185,
      "epoch": 0.4304228513150216,
      "grad_norm": 8.733558654785156,
      "learning_rate": 3.856084865371667e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8611682400107383,
      "num_tokens": 167035334.0,
      "step": 138850
    },
    {
      "entropy": 1.8504212602972985,
      "epoch": 0.43045385044007134,
      "grad_norm": 2.9879043102264404,
      "learning_rate": 3.855946013936313e-06,
      "loss": 0.4424,
      "mean_token_accuracy": 0.8579250067472458,
      "num_tokens": 167048594.0,
      "step": 138860
    },
    {
      "entropy": 1.9671182319521905,
      "epoch": 0.430484849565121,
      "grad_norm": 8.825550079345703,
      "learning_rate": 3.855807177499334e-06,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8505840554833413,
      "num_tokens": 167060152.0,
      "step": 138870
    },
    {
      "entropy": 1.8691889211535453,
      "epoch": 0.43051584869017073,
      "grad_norm": 7.807781219482422,
      "learning_rate": 3.855668356058026e-06,
      "loss": 0.4117,
      "mean_token_accuracy": 0.8700853690505028,
      "num_tokens": 167072813.0,
      "step": 138880
    },
    {
      "entropy": 1.89640132188797,
      "epoch": 0.4305468478152204,
      "grad_norm": 7.267297267913818,
      "learning_rate": 3.855529549609692e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8511320620775222,
      "num_tokens": 167084361.0,
      "step": 138890
    },
    {
      "entropy": 1.8999717831611633,
      "epoch": 0.4305778469402701,
      "grad_norm": 8.101717948913574,
      "learning_rate": 3.855390758151633e-06,
      "loss": 0.4316,
      "mean_token_accuracy": 0.8580278515815735,
      "num_tokens": 167096540.0,
      "step": 138900
    },
    {
      "entropy": 1.9178320989012718,
      "epoch": 0.4306088460653198,
      "grad_norm": 6.982521057128906,
      "learning_rate": 3.855251981681151e-06,
      "loss": 0.4057,
      "mean_token_accuracy": 0.8600656524300575,
      "num_tokens": 167107872.0,
      "step": 138910
    },
    {
      "entropy": 1.9452101722359658,
      "epoch": 0.4306398451903695,
      "grad_norm": 10.741822242736816,
      "learning_rate": 3.855113220195549e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8533175542950631,
      "num_tokens": 167119118.0,
      "step": 138920
    },
    {
      "entropy": 1.893312358856201,
      "epoch": 0.4306708443154192,
      "grad_norm": 8.916707992553711,
      "learning_rate": 3.8549744736921305e-06,
      "loss": 0.4095,
      "mean_token_accuracy": 0.8634658083319664,
      "num_tokens": 167130595.0,
      "step": 138930
    },
    {
      "entropy": 1.8818313643336295,
      "epoch": 0.4307018434404689,
      "grad_norm": 9.812067985534668,
      "learning_rate": 3.854835742168199e-06,
      "loss": 0.4265,
      "mean_token_accuracy": 0.8561764359474182,
      "num_tokens": 167143490.0,
      "step": 138940
    },
    {
      "entropy": 1.8934076741337775,
      "epoch": 0.4307328425655186,
      "grad_norm": 6.764780521392822,
      "learning_rate": 3.85469702562106e-06,
      "loss": 0.4771,
      "mean_token_accuracy": 0.8511760547757149,
      "num_tokens": 167156082.0,
      "step": 138950
    },
    {
      "entropy": 1.983777368068695,
      "epoch": 0.4307638416905683,
      "grad_norm": 7.675331115722656,
      "learning_rate": 3.854558324048018e-06,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8494358032941818,
      "num_tokens": 167166878.0,
      "step": 138960
    },
    {
      "entropy": 1.8882510006427764,
      "epoch": 0.430794840815618,
      "grad_norm": 3.8300962448120117,
      "learning_rate": 3.85441963744638e-06,
      "loss": 0.4472,
      "mean_token_accuracy": 0.8561283707618713,
      "num_tokens": 167178890.0,
      "step": 138970
    },
    {
      "entropy": 1.9575580522418021,
      "epoch": 0.4308258399406677,
      "grad_norm": 10.387866020202637,
      "learning_rate": 3.854280965813453e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.843521349132061,
      "num_tokens": 167190173.0,
      "step": 138980
    },
    {
      "entropy": 1.9261010527610778,
      "epoch": 0.43085683906571737,
      "grad_norm": 7.206926345825195,
      "learning_rate": 3.8541423091465445e-06,
      "loss": 0.461,
      "mean_token_accuracy": 0.856898583471775,
      "num_tokens": 167201201.0,
      "step": 138990
    },
    {
      "entropy": 1.871525762975216,
      "epoch": 0.4308878381907671,
      "grad_norm": 7.605156898498535,
      "learning_rate": 3.854003667442962e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8476424932479858,
      "num_tokens": 167213579.0,
      "step": 139000
    },
    {
      "entropy": 1.9758980587124824,
      "epoch": 0.43091883731581676,
      "grad_norm": 8.815040588378906,
      "learning_rate": 3.853865040700016e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8458743005990982,
      "num_tokens": 167224666.0,
      "step": 139010
    },
    {
      "entropy": 1.883438839018345,
      "epoch": 0.43094983644086643,
      "grad_norm": 7.926342010498047,
      "learning_rate": 3.853726428915014e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8550115719437599,
      "num_tokens": 167237308.0,
      "step": 139020
    },
    {
      "entropy": 1.984131459891796,
      "epoch": 0.43098083556591615,
      "grad_norm": 7.568789958953857,
      "learning_rate": 3.853587832085266e-06,
      "loss": 0.5039,
      "mean_token_accuracy": 0.8429438158869743,
      "num_tokens": 167248949.0,
      "step": 139030
    },
    {
      "entropy": 1.8531333744525909,
      "epoch": 0.4310118346909658,
      "grad_norm": 4.8112359046936035,
      "learning_rate": 3.8534492502080855e-06,
      "loss": 0.4109,
      "mean_token_accuracy": 0.8603510394692421,
      "num_tokens": 167261627.0,
      "step": 139040
    },
    {
      "entropy": 1.923319575190544,
      "epoch": 0.43104283381601555,
      "grad_norm": 7.872719764709473,
      "learning_rate": 3.853310683280783e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8496645122766495,
      "num_tokens": 167272866.0,
      "step": 139050
    },
    {
      "entropy": 1.9188224956393243,
      "epoch": 0.4310738329410652,
      "grad_norm": 8.1365327835083,
      "learning_rate": 3.853172131300669e-06,
      "loss": 0.4393,
      "mean_token_accuracy": 0.8515357092022896,
      "num_tokens": 167284947.0,
      "step": 139060
    },
    {
      "entropy": 1.8909332856535912,
      "epoch": 0.43110483206611494,
      "grad_norm": 9.189377784729004,
      "learning_rate": 3.8530335942650585e-06,
      "loss": 0.4156,
      "mean_token_accuracy": 0.853959108889103,
      "num_tokens": 167297259.0,
      "step": 139070
    },
    {
      "entropy": 1.8885620832443237,
      "epoch": 0.4311358311911646,
      "grad_norm": 9.13309097290039,
      "learning_rate": 3.852895072171265e-06,
      "loss": 0.4055,
      "mean_token_accuracy": 0.8578508168458938,
      "num_tokens": 167308954.0,
      "step": 139080
    },
    {
      "entropy": 1.8412837103009223,
      "epoch": 0.43116683031621433,
      "grad_norm": 4.5541090965271,
      "learning_rate": 3.8527565650166015e-06,
      "loss": 0.4433,
      "mean_token_accuracy": 0.8527804985642433,
      "num_tokens": 167321410.0,
      "step": 139090
    },
    {
      "entropy": 1.9028522804379464,
      "epoch": 0.431197829441264,
      "grad_norm": 9.414676666259766,
      "learning_rate": 3.852618072798383e-06,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8578800424933434,
      "num_tokens": 167333571.0,
      "step": 139100
    },
    {
      "entropy": 1.8518059805035592,
      "epoch": 0.4312288285663137,
      "grad_norm": 3.911571741104126,
      "learning_rate": 3.852479595513928e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8502576932311058,
      "num_tokens": 167346630.0,
      "step": 139110
    },
    {
      "entropy": 1.9005075827240945,
      "epoch": 0.4312598276913634,
      "grad_norm": 7.447115898132324,
      "learning_rate": 3.852341133160549e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8436521142721176,
      "num_tokens": 167359023.0,
      "step": 139120
    },
    {
      "entropy": 1.955150267481804,
      "epoch": 0.4312908268164131,
      "grad_norm": 8.531542778015137,
      "learning_rate": 3.8522026857355656e-06,
      "loss": 0.4963,
      "mean_token_accuracy": 0.8467582687735558,
      "num_tokens": 167370225.0,
      "step": 139130
    },
    {
      "entropy": 1.912473227083683,
      "epoch": 0.4313218259414628,
      "grad_norm": 7.494203567504883,
      "learning_rate": 3.852064253236293e-06,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8552014127373695,
      "num_tokens": 167382204.0,
      "step": 139140
    },
    {
      "entropy": 1.929689645767212,
      "epoch": 0.4313528250665125,
      "grad_norm": 8.015472412109375,
      "learning_rate": 3.851925835660052e-06,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8602582827210427,
      "num_tokens": 167393606.0,
      "step": 139150
    },
    {
      "entropy": 1.93514022231102,
      "epoch": 0.4313838241915622,
      "grad_norm": 6.807281017303467,
      "learning_rate": 3.851787433004161e-06,
      "loss": 0.456,
      "mean_token_accuracy": 0.855001138150692,
      "num_tokens": 167404837.0,
      "step": 139160
    },
    {
      "entropy": 1.8539175868034363,
      "epoch": 0.4314148233166119,
      "grad_norm": 10.743212699890137,
      "learning_rate": 3.851649045265939e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.8622266680002213,
      "num_tokens": 167416637.0,
      "step": 139170
    },
    {
      "entropy": 1.9064085155725479,
      "epoch": 0.4314458224416616,
      "grad_norm": 8.57366943359375,
      "learning_rate": 3.851510672442708e-06,
      "loss": 0.4703,
      "mean_token_accuracy": 0.8500051259994507,
      "num_tokens": 167428580.0,
      "step": 139180
    },
    {
      "entropy": 1.8125137344002724,
      "epoch": 0.4314768215667113,
      "grad_norm": 8.313339233398438,
      "learning_rate": 3.851372314531787e-06,
      "loss": 0.392,
      "mean_token_accuracy": 0.8584845051169395,
      "num_tokens": 167441286.0,
      "step": 139190
    },
    {
      "entropy": 1.9030530989170074,
      "epoch": 0.43150782069176097,
      "grad_norm": 4.577101707458496,
      "learning_rate": 3.851233971530498e-06,
      "loss": 0.4585,
      "mean_token_accuracy": 0.8593366608023644,
      "num_tokens": 167453514.0,
      "step": 139200
    },
    {
      "entropy": 1.9610344976186753,
      "epoch": 0.4315388198168107,
      "grad_norm": 8.109823226928711,
      "learning_rate": 3.851095643436164e-06,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8386234149336815,
      "num_tokens": 167464843.0,
      "step": 139210
    },
    {
      "entropy": 1.9184462070465087,
      "epoch": 0.43156981894186036,
      "grad_norm": 4.524685859680176,
      "learning_rate": 3.850957330246109e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8548565566539764,
      "num_tokens": 167476607.0,
      "step": 139220
    },
    {
      "entropy": 1.8219572558999062,
      "epoch": 0.4316008180669101,
      "grad_norm": 7.470212936401367,
      "learning_rate": 3.850819031957655e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8599913343787193,
      "num_tokens": 167489375.0,
      "step": 139230
    },
    {
      "entropy": 1.878198716044426,
      "epoch": 0.43163181719195975,
      "grad_norm": 7.494040489196777,
      "learning_rate": 3.850680748568127e-06,
      "loss": 0.4019,
      "mean_token_accuracy": 0.8682243391871453,
      "num_tokens": 167501096.0,
      "step": 139240
    },
    {
      "entropy": 1.882466796040535,
      "epoch": 0.4316628163170095,
      "grad_norm": 3.8041422367095947,
      "learning_rate": 3.85054248007485e-06,
      "loss": 0.4767,
      "mean_token_accuracy": 0.8378754198551178,
      "num_tokens": 167513496.0,
      "step": 139250
    },
    {
      "entropy": 1.9623596340417861,
      "epoch": 0.43169381544205915,
      "grad_norm": 9.18985366821289,
      "learning_rate": 3.850404226475151e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8428544372320175,
      "num_tokens": 167523826.0,
      "step": 139260
    },
    {
      "entropy": 1.8917414352297783,
      "epoch": 0.4317248145671088,
      "grad_norm": 9.102961540222168,
      "learning_rate": 3.850265987766354e-06,
      "loss": 0.437,
      "mean_token_accuracy": 0.8560737207531929,
      "num_tokens": 167535560.0,
      "step": 139270
    },
    {
      "entropy": 1.9341025441884994,
      "epoch": 0.43175581369215854,
      "grad_norm": 9.072357177734375,
      "learning_rate": 3.850127763945788e-06,
      "loss": 0.4758,
      "mean_token_accuracy": 0.8465018764138221,
      "num_tokens": 167546927.0,
      "step": 139280
    },
    {
      "entropy": 1.7552316695451737,
      "epoch": 0.4317868128172082,
      "grad_norm": 6.688268661499023,
      "learning_rate": 3.849989555010781e-06,
      "loss": 0.363,
      "mean_token_accuracy": 0.8682321265339852,
      "num_tokens": 167560862.0,
      "step": 139290
    },
    {
      "entropy": 1.9611527785658835,
      "epoch": 0.43181781194225793,
      "grad_norm": 3.5315842628479004,
      "learning_rate": 3.84985136095866e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8544865190982819,
      "num_tokens": 167572247.0,
      "step": 139300
    },
    {
      "entropy": 1.9224155962467193,
      "epoch": 0.4318488110673076,
      "grad_norm": 8.508109092712402,
      "learning_rate": 3.849713181786755e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8514292508363723,
      "num_tokens": 167583263.0,
      "step": 139310
    },
    {
      "entropy": 1.910231950879097,
      "epoch": 0.4318798101923573,
      "grad_norm": 9.595215797424316,
      "learning_rate": 3.849575017492395e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.8502141878008842,
      "num_tokens": 167595159.0,
      "step": 139320
    },
    {
      "entropy": 1.9836928397417068,
      "epoch": 0.431910809317407,
      "grad_norm": 8.529363632202148,
      "learning_rate": 3.849436868072912e-06,
      "loss": 0.5398,
      "mean_token_accuracy": 0.8343594878911972,
      "num_tokens": 167606457.0,
      "step": 139330
    },
    {
      "entropy": 1.8075464725494386,
      "epoch": 0.4319418084424567,
      "grad_norm": 7.387181282043457,
      "learning_rate": 3.849298733525636e-06,
      "loss": 0.389,
      "mean_token_accuracy": 0.8693550243973732,
      "num_tokens": 167619065.0,
      "step": 139340
    },
    {
      "entropy": 1.8776528060436248,
      "epoch": 0.4319728075675064,
      "grad_norm": 7.418018341064453,
      "learning_rate": 3.849160613847898e-06,
      "loss": 0.4092,
      "mean_token_accuracy": 0.8570902019739151,
      "num_tokens": 167631388.0,
      "step": 139350
    },
    {
      "entropy": 1.8456250533461571,
      "epoch": 0.4320038066925561,
      "grad_norm": 6.901108741760254,
      "learning_rate": 3.849022509037032e-06,
      "loss": 0.4719,
      "mean_token_accuracy": 0.8538346603512764,
      "num_tokens": 167643724.0,
      "step": 139360
    },
    {
      "entropy": 1.8757693514227867,
      "epoch": 0.4320348058176058,
      "grad_norm": 7.9772210121154785,
      "learning_rate": 3.848884419090371e-06,
      "loss": 0.4258,
      "mean_token_accuracy": 0.8559473112225533,
      "num_tokens": 167657019.0,
      "step": 139370
    },
    {
      "entropy": 1.8769934855401516,
      "epoch": 0.4320658049426555,
      "grad_norm": 7.6656413078308105,
      "learning_rate": 3.848746344005249e-06,
      "loss": 0.3837,
      "mean_token_accuracy": 0.8610425606369972,
      "num_tokens": 167669569.0,
      "step": 139380
    },
    {
      "entropy": 1.8971701174974442,
      "epoch": 0.4320968040677052,
      "grad_norm": 8.26727294921875,
      "learning_rate": 3.848608283778998e-06,
      "loss": 0.4906,
      "mean_token_accuracy": 0.8484355866909027,
      "num_tokens": 167681405.0,
      "step": 139390
    },
    {
      "entropy": 1.9876887768507003,
      "epoch": 0.4321278031927549,
      "grad_norm": 7.842116832733154,
      "learning_rate": 3.848470238408956e-06,
      "loss": 0.5748,
      "mean_token_accuracy": 0.8318600550293922,
      "num_tokens": 167692558.0,
      "step": 139400
    },
    {
      "entropy": 1.8590367823839187,
      "epoch": 0.43215880231780457,
      "grad_norm": 8.572957038879395,
      "learning_rate": 3.848332207892458e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8467065960168838,
      "num_tokens": 167705077.0,
      "step": 139410
    },
    {
      "entropy": 1.9348689168691635,
      "epoch": 0.4321898014428543,
      "grad_norm": 7.435234069824219,
      "learning_rate": 3.84819419222684e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8602284207940102,
      "num_tokens": 167716206.0,
      "step": 139420
    },
    {
      "entropy": 1.9003080978989602,
      "epoch": 0.43222080056790396,
      "grad_norm": 3.873739242553711,
      "learning_rate": 3.848056191409439e-06,
      "loss": 0.4818,
      "mean_token_accuracy": 0.8429651230573654,
      "num_tokens": 167728103.0,
      "step": 139430
    },
    {
      "entropy": 1.9454911321401596,
      "epoch": 0.4322517996929537,
      "grad_norm": 7.4276018142700195,
      "learning_rate": 3.847918205437593e-06,
      "loss": 0.5043,
      "mean_token_accuracy": 0.844630342721939,
      "num_tokens": 167738632.0,
      "step": 139440
    },
    {
      "entropy": 1.8453918382525445,
      "epoch": 0.43228279881800336,
      "grad_norm": 8.798940658569336,
      "learning_rate": 3.847780234308641e-06,
      "loss": 0.4311,
      "mean_token_accuracy": 0.8564523667097091,
      "num_tokens": 167750964.0,
      "step": 139450
    },
    {
      "entropy": 1.8049393191933631,
      "epoch": 0.4323137979430531,
      "grad_norm": 9.362977027893066,
      "learning_rate": 3.847642278019923e-06,
      "loss": 0.4083,
      "mean_token_accuracy": 0.8607902884483337,
      "num_tokens": 167763833.0,
      "step": 139460
    },
    {
      "entropy": 1.8966955333948134,
      "epoch": 0.43234479706810275,
      "grad_norm": 7.66609001159668,
      "learning_rate": 3.8475043365687755e-06,
      "loss": 0.4731,
      "mean_token_accuracy": 0.8480794101953506,
      "num_tokens": 167776065.0,
      "step": 139470
    },
    {
      "entropy": 1.8854218780994416,
      "epoch": 0.4323757961931525,
      "grad_norm": 8.28154468536377,
      "learning_rate": 3.847366409952543e-06,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8489232257008552,
      "num_tokens": 167788059.0,
      "step": 139480
    },
    {
      "entropy": 1.8077181428670883,
      "epoch": 0.43240679531820214,
      "grad_norm": 3.7770919799804688,
      "learning_rate": 3.847228498168565e-06,
      "loss": 0.3587,
      "mean_token_accuracy": 0.8708108901977539,
      "num_tokens": 167801348.0,
      "step": 139490
    },
    {
      "entropy": 1.8897404298186302,
      "epoch": 0.43243779444325187,
      "grad_norm": 7.708387851715088,
      "learning_rate": 3.8470906012141825e-06,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8538424223661423,
      "num_tokens": 167814120.0,
      "step": 139500
    },
    {
      "entropy": 1.834130634367466,
      "epoch": 0.43246879356830153,
      "grad_norm": 6.732454299926758,
      "learning_rate": 3.846952719086739e-06,
      "loss": 0.3968,
      "mean_token_accuracy": 0.8624187022447586,
      "num_tokens": 167826074.0,
      "step": 139510
    },
    {
      "entropy": 1.8972522467374802,
      "epoch": 0.4324997926933512,
      "grad_norm": 9.73922061920166,
      "learning_rate": 3.8468148517835766e-06,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8530626803636551,
      "num_tokens": 167836968.0,
      "step": 139520
    },
    {
      "entropy": 1.8679583430290223,
      "epoch": 0.43253079181840093,
      "grad_norm": 6.562689304351807,
      "learning_rate": 3.84667699930204e-06,
      "loss": 0.485,
      "mean_token_accuracy": 0.8574771344661712,
      "num_tokens": 167849043.0,
      "step": 139530
    },
    {
      "entropy": 1.8196332067251206,
      "epoch": 0.4325617909434506,
      "grad_norm": 8.335590362548828,
      "learning_rate": 3.846539161639474e-06,
      "loss": 0.3958,
      "mean_token_accuracy": 0.873447559773922,
      "num_tokens": 167861647.0,
      "step": 139540
    },
    {
      "entropy": 1.8070330306887628,
      "epoch": 0.4325927900685003,
      "grad_norm": 8.824051856994629,
      "learning_rate": 3.846401338793222e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8618940994143486,
      "num_tokens": 167874455.0,
      "step": 139550
    },
    {
      "entropy": 1.9073014467954637,
      "epoch": 0.43262378919355,
      "grad_norm": 8.493125915527344,
      "learning_rate": 3.846263530760633e-06,
      "loss": 0.4563,
      "mean_token_accuracy": 0.8542511522769928,
      "num_tokens": 167885644.0,
      "step": 139560
    },
    {
      "entropy": 1.8156562566757202,
      "epoch": 0.4326547883185997,
      "grad_norm": 9.763708114624023,
      "learning_rate": 3.84612573753905e-06,
      "loss": 0.4308,
      "mean_token_accuracy": 0.8623699083924293,
      "num_tokens": 167898195.0,
      "step": 139570
    },
    {
      "entropy": 1.9067368656396866,
      "epoch": 0.4326857874436494,
      "grad_norm": 8.543749809265137,
      "learning_rate": 3.845987959125823e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.855238850414753,
      "num_tokens": 167908792.0,
      "step": 139580
    },
    {
      "entropy": 1.8284807071089744,
      "epoch": 0.4327167865686991,
      "grad_norm": 7.208462238311768,
      "learning_rate": 3.8458501955182975e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8515859737992286,
      "num_tokens": 167921383.0,
      "step": 139590
    },
    {
      "entropy": 1.872288428246975,
      "epoch": 0.4327477856937488,
      "grad_norm": 7.469978332519531,
      "learning_rate": 3.845712446713823e-06,
      "loss": 0.4476,
      "mean_token_accuracy": 0.8519970834255218,
      "num_tokens": 167932974.0,
      "step": 139600
    },
    {
      "entropy": 1.918510441482067,
      "epoch": 0.4327787848187985,
      "grad_norm": 10.325370788574219,
      "learning_rate": 3.845574712709749e-06,
      "loss": 0.4922,
      "mean_token_accuracy": 0.8455549374222755,
      "num_tokens": 167945139.0,
      "step": 139610
    },
    {
      "entropy": 1.7871095269918442,
      "epoch": 0.43280978394384817,
      "grad_norm": 8.072989463806152,
      "learning_rate": 3.845436993503426e-06,
      "loss": 0.3753,
      "mean_token_accuracy": 0.8632336914539337,
      "num_tokens": 167959774.0,
      "step": 139620
    },
    {
      "entropy": 1.6974161878228187,
      "epoch": 0.4328407830688979,
      "grad_norm": 1.9402376413345337,
      "learning_rate": 3.845299289092203e-06,
      "loss": 0.3739,
      "mean_token_accuracy": 0.8687549382448196,
      "num_tokens": 167974761.0,
      "step": 139630
    },
    {
      "entropy": 1.8783841490745545,
      "epoch": 0.43287178219394756,
      "grad_norm": 8.304696083068848,
      "learning_rate": 3.845161599473431e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8487045183777809,
      "num_tokens": 167987541.0,
      "step": 139640
    },
    {
      "entropy": 1.8873986780643464,
      "epoch": 0.4329027813189973,
      "grad_norm": 7.96927547454834,
      "learning_rate": 3.845023924644462e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8622677832841873,
      "num_tokens": 167999815.0,
      "step": 139650
    },
    {
      "entropy": 1.8961790382862092,
      "epoch": 0.43293378044404696,
      "grad_norm": 9.11035442352295,
      "learning_rate": 3.84488626460265e-06,
      "loss": 0.4601,
      "mean_token_accuracy": 0.8553287327289582,
      "num_tokens": 168011278.0,
      "step": 139660
    },
    {
      "entropy": 1.8724280893802643,
      "epoch": 0.4329647795690967,
      "grad_norm": 7.960391521453857,
      "learning_rate": 3.8447486193453464e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8539118707180023,
      "num_tokens": 168023250.0,
      "step": 139670
    },
    {
      "entropy": 1.9397452518343925,
      "epoch": 0.43299577869414635,
      "grad_norm": 6.704464912414551,
      "learning_rate": 3.8446109888699054e-06,
      "loss": 0.4952,
      "mean_token_accuracy": 0.844795499742031,
      "num_tokens": 168034708.0,
      "step": 139680
    },
    {
      "entropy": 1.9342885583639144,
      "epoch": 0.4330267778191961,
      "grad_norm": 8.719551086425781,
      "learning_rate": 3.8444733731736825e-06,
      "loss": 0.4688,
      "mean_token_accuracy": 0.8504753604531288,
      "num_tokens": 168046323.0,
      "step": 139690
    },
    {
      "entropy": 1.9050405994057655,
      "epoch": 0.43305777694424574,
      "grad_norm": 7.164178371429443,
      "learning_rate": 3.8443357722540315e-06,
      "loss": 0.3677,
      "mean_token_accuracy": 0.87413921803236,
      "num_tokens": 168058199.0,
      "step": 139700
    },
    {
      "entropy": 1.9514371365308762,
      "epoch": 0.43308877606929547,
      "grad_norm": 8.19334602355957,
      "learning_rate": 3.844198186108308e-06,
      "loss": 0.5086,
      "mean_token_accuracy": 0.8353783071041108,
      "num_tokens": 168069154.0,
      "step": 139710
    },
    {
      "entropy": 1.8015297777950763,
      "epoch": 0.43311977519434514,
      "grad_norm": 8.755090713500977,
      "learning_rate": 3.84406061473387e-06,
      "loss": 0.3908,
      "mean_token_accuracy": 0.8574639052152634,
      "num_tokens": 168082496.0,
      "step": 139720
    },
    {
      "entropy": 1.8433510437607765,
      "epoch": 0.43315077431939486,
      "grad_norm": 7.901381492614746,
      "learning_rate": 3.843923058128073e-06,
      "loss": 0.4282,
      "mean_token_accuracy": 0.8601280748844147,
      "num_tokens": 168094337.0,
      "step": 139730
    },
    {
      "entropy": 1.909453672170639,
      "epoch": 0.43318177344444453,
      "grad_norm": 4.069192409515381,
      "learning_rate": 3.8437855162882766e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8465544447302819,
      "num_tokens": 168106420.0,
      "step": 139740
    },
    {
      "entropy": 1.9447955161333084,
      "epoch": 0.43321277256949425,
      "grad_norm": 9.183345794677734,
      "learning_rate": 3.843647989211837e-06,
      "loss": 0.5139,
      "mean_token_accuracy": 0.8410720884799957,
      "num_tokens": 168117779.0,
      "step": 139750
    },
    {
      "entropy": 1.9353636160492897,
      "epoch": 0.4332437716945439,
      "grad_norm": 6.792473793029785,
      "learning_rate": 3.843510476896116e-06,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8577348798513412,
      "num_tokens": 168129340.0,
      "step": 139760
    },
    {
      "entropy": 1.8856726735830307,
      "epoch": 0.4332747708195936,
      "grad_norm": 8.366604804992676,
      "learning_rate": 3.843372979338471e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8540061011910438,
      "num_tokens": 168141683.0,
      "step": 139770
    },
    {
      "entropy": 1.96417216360569,
      "epoch": 0.4333057699446433,
      "grad_norm": 8.365377426147461,
      "learning_rate": 3.843235496536262e-06,
      "loss": 0.513,
      "mean_token_accuracy": 0.8412682712078094,
      "num_tokens": 168152620.0,
      "step": 139780
    },
    {
      "entropy": 1.9343325823545456,
      "epoch": 0.433336769069693,
      "grad_norm": 6.398437023162842,
      "learning_rate": 3.843098028486852e-06,
      "loss": 0.434,
      "mean_token_accuracy": 0.8587636038661003,
      "num_tokens": 168164184.0,
      "step": 139790
    },
    {
      "entropy": 1.8950095444917678,
      "epoch": 0.4333677681947427,
      "grad_norm": 8.576680183410645,
      "learning_rate": 3.842960575187603e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8526839464902878,
      "num_tokens": 168175843.0,
      "step": 139800
    },
    {
      "entropy": 1.8716197207570076,
      "epoch": 0.4333987673197924,
      "grad_norm": 7.882635593414307,
      "learning_rate": 3.842823136635875e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8643146887421608,
      "num_tokens": 168187498.0,
      "step": 139810
    },
    {
      "entropy": 1.8388502165675162,
      "epoch": 0.4334297664448421,
      "grad_norm": 7.46567964553833,
      "learning_rate": 3.842685712829033e-06,
      "loss": 0.4209,
      "mean_token_accuracy": 0.8461194902658462,
      "num_tokens": 168200282.0,
      "step": 139820
    },
    {
      "entropy": 1.8559609532356263,
      "epoch": 0.43346076556989177,
      "grad_norm": 5.5117387771606445,
      "learning_rate": 3.84254830376444e-06,
      "loss": 0.4908,
      "mean_token_accuracy": 0.8383535325527192,
      "num_tokens": 168213312.0,
      "step": 139830
    },
    {
      "entropy": 1.8750181749463082,
      "epoch": 0.4334917646949415,
      "grad_norm": 2.167787790298462,
      "learning_rate": 3.8424109094394605e-06,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8538003221154213,
      "num_tokens": 168226085.0,
      "step": 139840
    },
    {
      "entropy": 1.887896014750004,
      "epoch": 0.43352276381999116,
      "grad_norm": 3.5215682983398438,
      "learning_rate": 3.842273529851461e-06,
      "loss": 0.4376,
      "mean_token_accuracy": 0.8547932624816894,
      "num_tokens": 168237911.0,
      "step": 139850
    },
    {
      "entropy": 1.847142294049263,
      "epoch": 0.4335537629450409,
      "grad_norm": 7.590489864349365,
      "learning_rate": 3.842136164997804e-06,
      "loss": 0.3865,
      "mean_token_accuracy": 0.8609548598527909,
      "num_tokens": 168250763.0,
      "step": 139860
    },
    {
      "entropy": 1.7902900233864785,
      "epoch": 0.43358476207009056,
      "grad_norm": 3.1853187084198,
      "learning_rate": 3.841998814875858e-06,
      "loss": 0.4003,
      "mean_token_accuracy": 0.8613974347710609,
      "num_tokens": 168263670.0,
      "step": 139870
    },
    {
      "entropy": 1.8125322625041007,
      "epoch": 0.4336157611951403,
      "grad_norm": 8.757499694824219,
      "learning_rate": 3.84186147948299e-06,
      "loss": 0.3847,
      "mean_token_accuracy": 0.8673945814371109,
      "num_tokens": 168277183.0,
      "step": 139880
    },
    {
      "entropy": 1.9186308607459068,
      "epoch": 0.43364676032018995,
      "grad_norm": 8.361207962036133,
      "learning_rate": 3.8417241588165675e-06,
      "loss": 0.4835,
      "mean_token_accuracy": 0.8502631932497025,
      "num_tokens": 168288847.0,
      "step": 139890
    },
    {
      "entropy": 1.7792581617832184,
      "epoch": 0.4336777594452397,
      "grad_norm": 3.6714017391204834,
      "learning_rate": 3.841586852873958e-06,
      "loss": 0.4079,
      "mean_token_accuracy": 0.8592787533998489,
      "num_tokens": 168301911.0,
      "step": 139900
    },
    {
      "entropy": 1.9583450973033905,
      "epoch": 0.43370875857028934,
      "grad_norm": 7.83007287979126,
      "learning_rate": 3.841449561652531e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8499109253287316,
      "num_tokens": 168312703.0,
      "step": 139910
    },
    {
      "entropy": 1.83574049025774,
      "epoch": 0.43373975769533907,
      "grad_norm": 8.328694343566895,
      "learning_rate": 3.841312285149657e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8661501377820968,
      "num_tokens": 168325343.0,
      "step": 139920
    },
    {
      "entropy": 1.9416332960128784,
      "epoch": 0.43377075682038874,
      "grad_norm": 7.570043563842773,
      "learning_rate": 3.841175023362706e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8532981321215629,
      "num_tokens": 168337277.0,
      "step": 139930
    },
    {
      "entropy": 1.9153632640838623,
      "epoch": 0.43380175594543846,
      "grad_norm": 6.462372303009033,
      "learning_rate": 3.841037776289048e-06,
      "loss": 0.4737,
      "mean_token_accuracy": 0.8486651003360748,
      "num_tokens": 168348776.0,
      "step": 139940
    },
    {
      "entropy": 1.928571656346321,
      "epoch": 0.43383275507048813,
      "grad_norm": 8.484127044677734,
      "learning_rate": 3.840900543926055e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8461811229586601,
      "num_tokens": 168360653.0,
      "step": 139950
    },
    {
      "entropy": 1.9500855028629303,
      "epoch": 0.43386375419553785,
      "grad_norm": 7.157684803009033,
      "learning_rate": 3.8407633262710995e-06,
      "loss": 0.4805,
      "mean_token_accuracy": 0.8550040423870087,
      "num_tokens": 168371201.0,
      "step": 139960
    },
    {
      "entropy": 1.9280799493193626,
      "epoch": 0.4338947533205875,
      "grad_norm": 6.877387046813965,
      "learning_rate": 3.840626123321555e-06,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8493270486593246,
      "num_tokens": 168382043.0,
      "step": 139970
    },
    {
      "entropy": 1.8996494933962822,
      "epoch": 0.43392575244563725,
      "grad_norm": 7.173794746398926,
      "learning_rate": 3.840488935074794e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.8530132621526718,
      "num_tokens": 168394684.0,
      "step": 139980
    },
    {
      "entropy": 2.0039018139243128,
      "epoch": 0.4339567515706869,
      "grad_norm": 7.884592056274414,
      "learning_rate": 3.840351761528191e-06,
      "loss": 0.5109,
      "mean_token_accuracy": 0.8355337530374527,
      "num_tokens": 168406211.0,
      "step": 139990
    },
    {
      "entropy": 1.8956574127078056,
      "epoch": 0.43398775069573664,
      "grad_norm": 5.687414646148682,
      "learning_rate": 3.840214602679122e-06,
      "loss": 0.4171,
      "mean_token_accuracy": 0.858708880841732,
      "num_tokens": 168418288.0,
      "step": 140000
    },
    {
      "entropy": 1.913138222694397,
      "epoch": 0.4340187498207863,
      "grad_norm": 3.5383858680725098,
      "learning_rate": 3.840077458524961e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8377836272120476,
      "num_tokens": 168430136.0,
      "step": 140010
    },
    {
      "entropy": 1.957027330994606,
      "epoch": 0.434049748945836,
      "grad_norm": 7.953402996063232,
      "learning_rate": 3.839940329063085e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8612416058778762,
      "num_tokens": 168441409.0,
      "step": 140020
    },
    {
      "entropy": 1.8934217691421509,
      "epoch": 0.4340807480708857,
      "grad_norm": 3.684905529022217,
      "learning_rate": 3.839803214290871e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8587592497467995,
      "num_tokens": 168453595.0,
      "step": 140030
    },
    {
      "entropy": 1.9658619672060014,
      "epoch": 0.43411174719593537,
      "grad_norm": 7.9534525871276855,
      "learning_rate": 3.839666114205696e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8584305748343468,
      "num_tokens": 168464254.0,
      "step": 140040
    },
    {
      "entropy": 1.8513989642262458,
      "epoch": 0.4341427463209851,
      "grad_norm": 7.037330150604248,
      "learning_rate": 3.839529028804939e-06,
      "loss": 0.3944,
      "mean_token_accuracy": 0.8667973428964615,
      "num_tokens": 168476419.0,
      "step": 140050
    },
    {
      "entropy": 1.921988594532013,
      "epoch": 0.43417374544603476,
      "grad_norm": 8.005762100219727,
      "learning_rate": 3.839391958085978e-06,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8441802576184273,
      "num_tokens": 168488005.0,
      "step": 140060
    },
    {
      "entropy": 1.874787649512291,
      "epoch": 0.4342047445710845,
      "grad_norm": 9.830438613891602,
      "learning_rate": 3.839254902046193e-06,
      "loss": 0.416,
      "mean_token_accuracy": 0.8617915287613869,
      "num_tokens": 168500064.0,
      "step": 140070
    },
    {
      "entropy": 1.9190911263227464,
      "epoch": 0.43423574369613416,
      "grad_norm": 9.500381469726562,
      "learning_rate": 3.8391178606829646e-06,
      "loss": 0.4676,
      "mean_token_accuracy": 0.8493470534682274,
      "num_tokens": 168511882.0,
      "step": 140080
    },
    {
      "entropy": 1.8970647498965263,
      "epoch": 0.4342667428211839,
      "grad_norm": 6.222733974456787,
      "learning_rate": 3.838980833993672e-06,
      "loss": 0.4962,
      "mean_token_accuracy": 0.845898899435997,
      "num_tokens": 168525185.0,
      "step": 140090
    },
    {
      "entropy": 1.9249570727348329,
      "epoch": 0.43429774194623355,
      "grad_norm": 8.286949157714844,
      "learning_rate": 3.838843821975697e-06,
      "loss": 0.4652,
      "mean_token_accuracy": 0.8432594612240791,
      "num_tokens": 168536800.0,
      "step": 140100
    },
    {
      "entropy": 1.7830603495240211,
      "epoch": 0.4343287410712833,
      "grad_norm": 3.775357723236084,
      "learning_rate": 3.8387068246264244e-06,
      "loss": 0.3601,
      "mean_token_accuracy": 0.8711403638124466,
      "num_tokens": 168550628.0,
      "step": 140110
    },
    {
      "entropy": 1.913707821071148,
      "epoch": 0.43435974019633294,
      "grad_norm": 7.623897075653076,
      "learning_rate": 3.838569841943234e-06,
      "loss": 0.4571,
      "mean_token_accuracy": 0.8579373374581337,
      "num_tokens": 168562584.0,
      "step": 140120
    },
    {
      "entropy": 1.959540669620037,
      "epoch": 0.43439073932138267,
      "grad_norm": 8.677681922912598,
      "learning_rate": 3.8384328739235095e-06,
      "loss": 0.4934,
      "mean_token_accuracy": 0.8414706364274025,
      "num_tokens": 168574108.0,
      "step": 140130
    },
    {
      "entropy": 1.8708548903465272,
      "epoch": 0.43442173844643234,
      "grad_norm": 7.195833683013916,
      "learning_rate": 3.838295920564638e-06,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8590035811066628,
      "num_tokens": 168586835.0,
      "step": 140140
    },
    {
      "entropy": 1.9178714960813523,
      "epoch": 0.43445273757148206,
      "grad_norm": 8.062506675720215,
      "learning_rate": 3.838158981863999e-06,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8575378447771073,
      "num_tokens": 168599034.0,
      "step": 140150
    },
    {
      "entropy": 1.9800096511840821,
      "epoch": 0.43448373669653173,
      "grad_norm": 7.617713451385498,
      "learning_rate": 3.838022057818983e-06,
      "loss": 0.4857,
      "mean_token_accuracy": 0.8560611560940743,
      "num_tokens": 168610043.0,
      "step": 140160
    },
    {
      "entropy": 1.8751267299056054,
      "epoch": 0.43451473582158145,
      "grad_norm": 9.476043701171875,
      "learning_rate": 3.837885148426972e-06,
      "loss": 0.4785,
      "mean_token_accuracy": 0.8532286942005157,
      "num_tokens": 168622646.0,
      "step": 140170
    },
    {
      "entropy": 1.9622350841760636,
      "epoch": 0.4345457349466311,
      "grad_norm": 8.158764839172363,
      "learning_rate": 3.837748253685356e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.852800378203392,
      "num_tokens": 168633945.0,
      "step": 140180
    },
    {
      "entropy": 1.9728653207421303,
      "epoch": 0.43457673407168085,
      "grad_norm": 8.091950416564941,
      "learning_rate": 3.8376113735915195e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8437511593103408,
      "num_tokens": 168645091.0,
      "step": 140190
    },
    {
      "entropy": 1.8134228363633156,
      "epoch": 0.4346077331967305,
      "grad_norm": 3.810476541519165,
      "learning_rate": 3.8374745081428525e-06,
      "loss": 0.3402,
      "mean_token_accuracy": 0.8733655110001564,
      "num_tokens": 168658792.0,
      "step": 140200
    },
    {
      "entropy": 1.9151874303817749,
      "epoch": 0.43463873232178024,
      "grad_norm": 7.474352836608887,
      "learning_rate": 3.837337657336742e-06,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8463295727968216,
      "num_tokens": 168670568.0,
      "step": 140210
    },
    {
      "entropy": 1.9047392055392265,
      "epoch": 0.4346697314468299,
      "grad_norm": 8.334789276123047,
      "learning_rate": 3.8372008211705795e-06,
      "loss": 0.4684,
      "mean_token_accuracy": 0.8502994626760483,
      "num_tokens": 168682642.0,
      "step": 140220
    },
    {
      "entropy": 1.9391117468476295,
      "epoch": 0.43470073057187963,
      "grad_norm": 9.910785675048828,
      "learning_rate": 3.837063999641753e-06,
      "loss": 0.4548,
      "mean_token_accuracy": 0.8517929673194885,
      "num_tokens": 168694378.0,
      "step": 140230
    },
    {
      "entropy": 1.9310035303235054,
      "epoch": 0.4347317296969293,
      "grad_norm": 8.19104290008545,
      "learning_rate": 3.836927192747655e-06,
      "loss": 0.44,
      "mean_token_accuracy": 0.8617075026035309,
      "num_tokens": 168705541.0,
      "step": 140240
    },
    {
      "entropy": 1.8467669546604157,
      "epoch": 0.43476272882197897,
      "grad_norm": 9.081491470336914,
      "learning_rate": 3.8367904004856745e-06,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8642741695046425,
      "num_tokens": 168718141.0,
      "step": 140250
    },
    {
      "entropy": 1.841063352674246,
      "epoch": 0.4347937279470287,
      "grad_norm": 7.375804424285889,
      "learning_rate": 3.836653622853204e-06,
      "loss": 0.3789,
      "mean_token_accuracy": 0.8677456378936768,
      "num_tokens": 168730815.0,
      "step": 140260
    },
    {
      "entropy": 1.8605894804000855,
      "epoch": 0.43482472707207837,
      "grad_norm": 6.377718448638916,
      "learning_rate": 3.836516859847637e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.858807185292244,
      "num_tokens": 168742824.0,
      "step": 140270
    },
    {
      "entropy": 1.824122267216444,
      "epoch": 0.4348557261971281,
      "grad_norm": 2.3955891132354736,
      "learning_rate": 3.836380111466366e-06,
      "loss": 0.3629,
      "mean_token_accuracy": 0.8775916382670402,
      "num_tokens": 168755819.0,
      "step": 140280
    },
    {
      "entropy": 1.9054122567176819,
      "epoch": 0.43488672532217776,
      "grad_norm": 10.010201454162598,
      "learning_rate": 3.836243377706786e-06,
      "loss": 0.4649,
      "mean_token_accuracy": 0.8460003793239593,
      "num_tokens": 168767525.0,
      "step": 140290
    },
    {
      "entropy": 1.9120841890573501,
      "epoch": 0.4349177244472275,
      "grad_norm": 8.370523452758789,
      "learning_rate": 3.83610665856629e-06,
      "loss": 0.4713,
      "mean_token_accuracy": 0.8461458712816239,
      "num_tokens": 168779091.0,
      "step": 140300
    },
    {
      "entropy": 1.9416150897741318,
      "epoch": 0.43494872357227715,
      "grad_norm": 8.070106506347656,
      "learning_rate": 3.835969954042273e-06,
      "loss": 0.4725,
      "mean_token_accuracy": 0.8444668143987656,
      "num_tokens": 168790525.0,
      "step": 140310
    },
    {
      "entropy": 1.873001678287983,
      "epoch": 0.4349797226973269,
      "grad_norm": 6.8141584396362305,
      "learning_rate": 3.835833264132131e-06,
      "loss": 0.3869,
      "mean_token_accuracy": 0.8707495495676995,
      "num_tokens": 168801983.0,
      "step": 140320
    },
    {
      "entropy": 1.9271577775478363,
      "epoch": 0.43501072182237654,
      "grad_norm": 6.567296504974365,
      "learning_rate": 3.835696588833263e-06,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8497977092862129,
      "num_tokens": 168814329.0,
      "step": 140330
    },
    {
      "entropy": 1.913456965982914,
      "epoch": 0.43504172094742627,
      "grad_norm": 7.978017330169678,
      "learning_rate": 3.8355599281430635e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.8430664092302322,
      "num_tokens": 168826198.0,
      "step": 140340
    },
    {
      "entropy": 1.882510894536972,
      "epoch": 0.43507272007247594,
      "grad_norm": 9.187688827514648,
      "learning_rate": 3.83542328205893e-06,
      "loss": 0.4491,
      "mean_token_accuracy": 0.8393991827964783,
      "num_tokens": 168838743.0,
      "step": 140350
    },
    {
      "entropy": 1.8549660280346871,
      "epoch": 0.43510371919752566,
      "grad_norm": 7.461533069610596,
      "learning_rate": 3.835286650578262e-06,
      "loss": 0.4404,
      "mean_token_accuracy": 0.8586645647883415,
      "num_tokens": 168851245.0,
      "step": 140360
    },
    {
      "entropy": 1.8594406992197037,
      "epoch": 0.43513471832257533,
      "grad_norm": 3.6771328449249268,
      "learning_rate": 3.835150033698458e-06,
      "loss": 0.3623,
      "mean_token_accuracy": 0.8606583446264267,
      "num_tokens": 168864222.0,
      "step": 140370
    },
    {
      "entropy": 1.9253443703055382,
      "epoch": 0.43516571744762506,
      "grad_norm": 7.7639241218566895,
      "learning_rate": 3.835013431416919e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8500281646847725,
      "num_tokens": 168876003.0,
      "step": 140380
    },
    {
      "entropy": 1.9680373221635818,
      "epoch": 0.4351967165726747,
      "grad_norm": 6.00819206237793,
      "learning_rate": 3.834876843731043e-06,
      "loss": 0.4793,
      "mean_token_accuracy": 0.848180590569973,
      "num_tokens": 168886871.0,
      "step": 140390
    },
    {
      "entropy": 1.9534240260720253,
      "epoch": 0.43522771569772445,
      "grad_norm": 7.778793811798096,
      "learning_rate": 3.834740270638232e-06,
      "loss": 0.476,
      "mean_token_accuracy": 0.8536478534340859,
      "num_tokens": 168898090.0,
      "step": 140400
    },
    {
      "entropy": 1.9209111735224724,
      "epoch": 0.4352587148227741,
      "grad_norm": 6.560374736785889,
      "learning_rate": 3.834603712135889e-06,
      "loss": 0.4202,
      "mean_token_accuracy": 0.8524667829275131,
      "num_tokens": 168909996.0,
      "step": 140410
    },
    {
      "entropy": 1.915994156897068,
      "epoch": 0.43528971394782384,
      "grad_norm": 4.213437080383301,
      "learning_rate": 3.834467168221415e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8442320972681046,
      "num_tokens": 168921960.0,
      "step": 140420
    },
    {
      "entropy": 1.9615544810891152,
      "epoch": 0.4353207130728735,
      "grad_norm": 8.39879035949707,
      "learning_rate": 3.834330638892213e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8525176584720612,
      "num_tokens": 168933811.0,
      "step": 140430
    },
    {
      "entropy": 1.8864429205656053,
      "epoch": 0.43535171219792324,
      "grad_norm": 7.037664413452148,
      "learning_rate": 3.834194124145686e-06,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8612302899360657,
      "num_tokens": 168946667.0,
      "step": 140440
    },
    {
      "entropy": 2.0172328650951385,
      "epoch": 0.4353827113229729,
      "grad_norm": 4.6347975730896,
      "learning_rate": 3.834057623979241e-06,
      "loss": 0.5135,
      "mean_token_accuracy": 0.842670176923275,
      "num_tokens": 168958082.0,
      "step": 140450
    },
    {
      "entropy": 2.007206231355667,
      "epoch": 0.43541371044802263,
      "grad_norm": 8.676667213439941,
      "learning_rate": 3.833921138390279e-06,
      "loss": 0.506,
      "mean_token_accuracy": 0.8318960726261139,
      "num_tokens": 168969385.0,
      "step": 140460
    },
    {
      "entropy": 1.9538621738553048,
      "epoch": 0.4354447095730723,
      "grad_norm": 7.149422645568848,
      "learning_rate": 3.833784667376208e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8514281839132309,
      "num_tokens": 168981528.0,
      "step": 140470
    },
    {
      "entropy": 1.9269013822078704,
      "epoch": 0.435475708698122,
      "grad_norm": 7.5539350509643555,
      "learning_rate": 3.833648210934433e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8525858089327812,
      "num_tokens": 168994111.0,
      "step": 140480
    },
    {
      "entropy": 2.015850293636322,
      "epoch": 0.4355067078231717,
      "grad_norm": 7.674806118011475,
      "learning_rate": 3.8335117690623616e-06,
      "loss": 0.5322,
      "mean_token_accuracy": 0.8376188382506371,
      "num_tokens": 169004802.0,
      "step": 140490
    },
    {
      "entropy": 1.868292573094368,
      "epoch": 0.43553770694822136,
      "grad_norm": 3.716049909591675,
      "learning_rate": 3.8333753417574014e-06,
      "loss": 0.4029,
      "mean_token_accuracy": 0.8568149447441101,
      "num_tokens": 169017161.0,
      "step": 140500
    },
    {
      "entropy": 1.999607202410698,
      "epoch": 0.4355687060732711,
      "grad_norm": 7.225944519042969,
      "learning_rate": 3.833238929016959e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8523830324411392,
      "num_tokens": 169027960.0,
      "step": 140510
    },
    {
      "entropy": 1.9995219513773919,
      "epoch": 0.43559970519832075,
      "grad_norm": 10.309820175170898,
      "learning_rate": 3.833102530838446e-06,
      "loss": 0.5616,
      "mean_token_accuracy": 0.8337021216750145,
      "num_tokens": 169039707.0,
      "step": 140520
    },
    {
      "entropy": 1.8859604820609093,
      "epoch": 0.4356307043233705,
      "grad_norm": 5.491618633270264,
      "learning_rate": 3.8329661472192686e-06,
      "loss": 0.4348,
      "mean_token_accuracy": 0.8571705892682076,
      "num_tokens": 169052242.0,
      "step": 140530
    },
    {
      "entropy": 1.8706535249948502,
      "epoch": 0.43566170344842015,
      "grad_norm": 4.212784767150879,
      "learning_rate": 3.832829778156839e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8539310187101364,
      "num_tokens": 169065057.0,
      "step": 140540
    },
    {
      "entropy": 1.9854949221014977,
      "epoch": 0.43569270257346987,
      "grad_norm": 7.0491461753845215,
      "learning_rate": 3.8326934236485665e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.855704678595066,
      "num_tokens": 169076702.0,
      "step": 140550
    },
    {
      "entropy": 1.8700939059257506,
      "epoch": 0.43572370169851954,
      "grad_norm": 7.940951824188232,
      "learning_rate": 3.832557083691864e-06,
      "loss": 0.4005,
      "mean_token_accuracy": 0.8501318633556366,
      "num_tokens": 169089351.0,
      "step": 140560
    },
    {
      "entropy": 1.975642454624176,
      "epoch": 0.43575470082356926,
      "grad_norm": 8.858991622924805,
      "learning_rate": 3.832420758284142e-06,
      "loss": 0.5046,
      "mean_token_accuracy": 0.8454146534204483,
      "num_tokens": 169100724.0,
      "step": 140570
    },
    {
      "entropy": 1.8851625487208366,
      "epoch": 0.43578569994861893,
      "grad_norm": 7.867511749267578,
      "learning_rate": 3.832284447422814e-06,
      "loss": 0.4544,
      "mean_token_accuracy": 0.8477107077836991,
      "num_tokens": 169112999.0,
      "step": 140580
    },
    {
      "entropy": 1.8516397640109061,
      "epoch": 0.43581669907366866,
      "grad_norm": 3.843761682510376,
      "learning_rate": 3.832148151105293e-06,
      "loss": 0.4388,
      "mean_token_accuracy": 0.8512039586901665,
      "num_tokens": 169126495.0,
      "step": 140590
    },
    {
      "entropy": 2.005415938794613,
      "epoch": 0.4358476981987183,
      "grad_norm": 7.267820835113525,
      "learning_rate": 3.832011869328994e-06,
      "loss": 0.5148,
      "mean_token_accuracy": 0.8331288233399391,
      "num_tokens": 169137715.0,
      "step": 140600
    },
    {
      "entropy": 1.9655996575951575,
      "epoch": 0.43587869732376805,
      "grad_norm": 7.318187236785889,
      "learning_rate": 3.831875602091329e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8459896892309189,
      "num_tokens": 169149332.0,
      "step": 140610
    },
    {
      "entropy": 1.9661901488900184,
      "epoch": 0.4359096964488177,
      "grad_norm": 7.033077716827393,
      "learning_rate": 3.831739349389714e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8531352400779724,
      "num_tokens": 169160520.0,
      "step": 140620
    },
    {
      "entropy": 1.859518001973629,
      "epoch": 0.43594069557386744,
      "grad_norm": 8.226398468017578,
      "learning_rate": 3.8316031112215676e-06,
      "loss": 0.371,
      "mean_token_accuracy": 0.866145646572113,
      "num_tokens": 169173970.0,
      "step": 140630
    },
    {
      "entropy": 1.967048665881157,
      "epoch": 0.4359716946989171,
      "grad_norm": 7.1692962646484375,
      "learning_rate": 3.831466887584302e-06,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8470461040735244,
      "num_tokens": 169184890.0,
      "step": 140640
    },
    {
      "entropy": 1.901542803645134,
      "epoch": 0.43600269382396684,
      "grad_norm": 7.0217156410217285,
      "learning_rate": 3.831330678475338e-06,
      "loss": 0.4176,
      "mean_token_accuracy": 0.8561606049537659,
      "num_tokens": 169197071.0,
      "step": 140650
    },
    {
      "entropy": 1.7790475860238075,
      "epoch": 0.4360336929490165,
      "grad_norm": 10.4070405960083,
      "learning_rate": 3.831194483892091e-06,
      "loss": 0.3602,
      "mean_token_accuracy": 0.8745766609907151,
      "num_tokens": 169210648.0,
      "step": 140660
    },
    {
      "entropy": 1.8767062574625015,
      "epoch": 0.43606469207406623,
      "grad_norm": 6.854859352111816,
      "learning_rate": 3.831058303831981e-06,
      "loss": 0.4237,
      "mean_token_accuracy": 0.8566436514258384,
      "num_tokens": 169222860.0,
      "step": 140670
    },
    {
      "entropy": 1.956781129539013,
      "epoch": 0.4360956911991159,
      "grad_norm": 8.431173324584961,
      "learning_rate": 3.830922138292426e-06,
      "loss": 0.4847,
      "mean_token_accuracy": 0.8472963184118271,
      "num_tokens": 169234554.0,
      "step": 140680
    },
    {
      "entropy": 1.9454703062772751,
      "epoch": 0.4361266903241656,
      "grad_norm": 9.45147705078125,
      "learning_rate": 3.8307859872708455e-06,
      "loss": 0.5006,
      "mean_token_accuracy": 0.848207201063633,
      "num_tokens": 169245918.0,
      "step": 140690
    },
    {
      "entropy": 1.9917195424437524,
      "epoch": 0.4361576894492153,
      "grad_norm": 8.706761360168457,
      "learning_rate": 3.830649850764661e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8496665880084038,
      "num_tokens": 169257011.0,
      "step": 140700
    },
    {
      "entropy": 1.9021431505680084,
      "epoch": 0.436188688574265,
      "grad_norm": 8.302827835083008,
      "learning_rate": 3.830513728771293e-06,
      "loss": 0.4093,
      "mean_token_accuracy": 0.8685775399208069,
      "num_tokens": 169268523.0,
      "step": 140710
    },
    {
      "entropy": 1.8747971653938293,
      "epoch": 0.4362196876993147,
      "grad_norm": 10.040637969970703,
      "learning_rate": 3.830377621288163e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8577211305499077,
      "num_tokens": 169281126.0,
      "step": 140720
    },
    {
      "entropy": 2.011331743001938,
      "epoch": 0.4362506868243644,
      "grad_norm": 9.337910652160645,
      "learning_rate": 3.8302415283126924e-06,
      "loss": 0.5219,
      "mean_token_accuracy": 0.8415524423122406,
      "num_tokens": 169291751.0,
      "step": 140730
    },
    {
      "entropy": 2.007808841764927,
      "epoch": 0.4362816859494141,
      "grad_norm": 7.370612144470215,
      "learning_rate": 3.830105449842306e-06,
      "loss": 0.5172,
      "mean_token_accuracy": 0.83756934851408,
      "num_tokens": 169303260.0,
      "step": 140740
    },
    {
      "entropy": 1.9281100079417228,
      "epoch": 0.43631268507446375,
      "grad_norm": 7.993682861328125,
      "learning_rate": 3.829969385874425e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8419538453221321,
      "num_tokens": 169315175.0,
      "step": 140750
    },
    {
      "entropy": 1.8865912228822708,
      "epoch": 0.43634368419951347,
      "grad_norm": 5.286984443664551,
      "learning_rate": 3.829833336406477e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8586981207132339,
      "num_tokens": 169328237.0,
      "step": 140760
    },
    {
      "entropy": 1.9767191842198373,
      "epoch": 0.43637468332456314,
      "grad_norm": 7.798305034637451,
      "learning_rate": 3.8296973014358825e-06,
      "loss": 0.5088,
      "mean_token_accuracy": 0.8429146945476532,
      "num_tokens": 169339917.0,
      "step": 140770
    },
    {
      "entropy": 1.928345987200737,
      "epoch": 0.43640568244961286,
      "grad_norm": 3.7796151638031006,
      "learning_rate": 3.829561280960071e-06,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8631057515740395,
      "num_tokens": 169351691.0,
      "step": 140780
    },
    {
      "entropy": 1.9100732266902924,
      "epoch": 0.43643668157466253,
      "grad_norm": 8.390362739562988,
      "learning_rate": 3.829425274976465e-06,
      "loss": 0.4309,
      "mean_token_accuracy": 0.8546838641166687,
      "num_tokens": 169364207.0,
      "step": 140790
    },
    {
      "entropy": 1.8982257694005966,
      "epoch": 0.43646768069971226,
      "grad_norm": 7.438647747039795,
      "learning_rate": 3.829289283482494e-06,
      "loss": 0.4037,
      "mean_token_accuracy": 0.8580241903662682,
      "num_tokens": 169376638.0,
      "step": 140800
    },
    {
      "entropy": 1.951183719933033,
      "epoch": 0.4364986798247619,
      "grad_norm": 3.837440252304077,
      "learning_rate": 3.829153306475584e-06,
      "loss": 0.4269,
      "mean_token_accuracy": 0.8538202360272408,
      "num_tokens": 169388468.0,
      "step": 140810
    },
    {
      "entropy": 1.9201230362057686,
      "epoch": 0.43652967894981165,
      "grad_norm": 8.309687614440918,
      "learning_rate": 3.829017343953164e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8367157101631164,
      "num_tokens": 169401878.0,
      "step": 140820
    },
    {
      "entropy": 1.9850788608193397,
      "epoch": 0.4365606780748613,
      "grad_norm": 8.711220741271973,
      "learning_rate": 3.828881395912661e-06,
      "loss": 0.4953,
      "mean_token_accuracy": 0.8478342741727829,
      "num_tokens": 169413486.0,
      "step": 140830
    },
    {
      "entropy": 1.9063683211803437,
      "epoch": 0.43659167719991104,
      "grad_norm": 10.316767692565918,
      "learning_rate": 3.828745462351506e-06,
      "loss": 0.4484,
      "mean_token_accuracy": 0.8501601129770279,
      "num_tokens": 169426451.0,
      "step": 140840
    },
    {
      "entropy": 2.031948208808899,
      "epoch": 0.4366226763249607,
      "grad_norm": 8.402220726013184,
      "learning_rate": 3.828609543267129e-06,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8418742671608925,
      "num_tokens": 169437295.0,
      "step": 140850
    },
    {
      "entropy": 1.7779492631554603,
      "epoch": 0.43665367545001044,
      "grad_norm": 8.303640365600586,
      "learning_rate": 3.828473638656959e-06,
      "loss": 0.3924,
      "mean_token_accuracy": 0.8546249598264695,
      "num_tokens": 169451864.0,
      "step": 140860
    },
    {
      "entropy": 1.8658212095499038,
      "epoch": 0.4366846745750601,
      "grad_norm": 8.145844459533691,
      "learning_rate": 3.828337748518429e-06,
      "loss": 0.4204,
      "mean_token_accuracy": 0.8555705919861794,
      "num_tokens": 169465061.0,
      "step": 140870
    },
    {
      "entropy": 1.9951273322105407,
      "epoch": 0.43671567370010983,
      "grad_norm": 7.6853461265563965,
      "learning_rate": 3.8282018728489685e-06,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8458278581500054,
      "num_tokens": 169476203.0,
      "step": 140880
    },
    {
      "entropy": 1.907281294465065,
      "epoch": 0.4367466728251595,
      "grad_norm": 4.954738616943359,
      "learning_rate": 3.828066011646013e-06,
      "loss": 0.4118,
      "mean_token_accuracy": 0.84654021859169,
      "num_tokens": 169488896.0,
      "step": 140890
    },
    {
      "entropy": 2.035962425172329,
      "epoch": 0.4367776719502092,
      "grad_norm": 9.046393394470215,
      "learning_rate": 3.8279301649069935e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8471018671989441,
      "num_tokens": 169500116.0,
      "step": 140900
    },
    {
      "entropy": 1.9954482674598695,
      "epoch": 0.4368086710752589,
      "grad_norm": 8.976217269897461,
      "learning_rate": 3.827794332629344e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8554398879408837,
      "num_tokens": 169510982.0,
      "step": 140910
    },
    {
      "entropy": 2.049514207243919,
      "epoch": 0.4368396702003086,
      "grad_norm": 9.787312507629395,
      "learning_rate": 3.8276585148105e-06,
      "loss": 0.5093,
      "mean_token_accuracy": 0.8407163426280022,
      "num_tokens": 169521572.0,
      "step": 140920
    },
    {
      "entropy": 1.923102656006813,
      "epoch": 0.4368706693253583,
      "grad_norm": 8.028412818908691,
      "learning_rate": 3.827522711447895e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8506998717784882,
      "num_tokens": 169533699.0,
      "step": 140930
    },
    {
      "entropy": 2.0024877056479453,
      "epoch": 0.436901668450408,
      "grad_norm": 4.008644104003906,
      "learning_rate": 3.827386922538967e-06,
      "loss": 0.4489,
      "mean_token_accuracy": 0.8583690106868744,
      "num_tokens": 169544829.0,
      "step": 140940
    },
    {
      "entropy": 1.84952729716897,
      "epoch": 0.4369326675754577,
      "grad_norm": 9.895330429077148,
      "learning_rate": 3.827251148081149e-06,
      "loss": 0.3781,
      "mean_token_accuracy": 0.8501643449068069,
      "num_tokens": 169558584.0,
      "step": 140950
    },
    {
      "entropy": 1.9892782807350158,
      "epoch": 0.4369636667005074,
      "grad_norm": 7.615428447723389,
      "learning_rate": 3.827115388071881e-06,
      "loss": 0.5019,
      "mean_token_accuracy": 0.844859579205513,
      "num_tokens": 169569716.0,
      "step": 140960
    },
    {
      "entropy": 2.0157076716423035,
      "epoch": 0.43699466582555707,
      "grad_norm": 8.59665298461914,
      "learning_rate": 3.8269796425086e-06,
      "loss": 0.5129,
      "mean_token_accuracy": 0.8417162656784057,
      "num_tokens": 169580522.0,
      "step": 140970
    },
    {
      "entropy": 1.925472079217434,
      "epoch": 0.4370256649506068,
      "grad_norm": 3.7771337032318115,
      "learning_rate": 3.826843911388742e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8646034061908722,
      "num_tokens": 169592175.0,
      "step": 140980
    },
    {
      "entropy": 1.8540797725319862,
      "epoch": 0.43705666407565646,
      "grad_norm": 7.4315056800842285,
      "learning_rate": 3.826708194709748e-06,
      "loss": 0.3951,
      "mean_token_accuracy": 0.8597232013940811,
      "num_tokens": 169604782.0,
      "step": 140990
    },
    {
      "entropy": 1.9015352100133895,
      "epoch": 0.43708766320070613,
      "grad_norm": 4.213851451873779,
      "learning_rate": 3.826572492469057e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8474423259496688,
      "num_tokens": 169617522.0,
      "step": 141000
    },
    {
      "entropy": 2.016425573825836,
      "epoch": 0.43711866232575586,
      "grad_norm": 8.818294525146484,
      "learning_rate": 3.826436804664109e-06,
      "loss": 0.5032,
      "mean_token_accuracy": 0.8433761283755302,
      "num_tokens": 169628341.0,
      "step": 141010
    },
    {
      "entropy": 1.9073638312518597,
      "epoch": 0.4371496614508055,
      "grad_norm": 7.4958014488220215,
      "learning_rate": 3.826301131292346e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8657842293381691,
      "num_tokens": 169641206.0,
      "step": 141020
    },
    {
      "entropy": 1.8773157000541687,
      "epoch": 0.43718066057585525,
      "grad_norm": 9.210996627807617,
      "learning_rate": 3.826165472351208e-06,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8486377000808716,
      "num_tokens": 169652953.0,
      "step": 141030
    },
    {
      "entropy": 1.8629372149705887,
      "epoch": 0.4372116597009049,
      "grad_norm": 8.852280616760254,
      "learning_rate": 3.826029827838136e-06,
      "loss": 0.436,
      "mean_token_accuracy": 0.8569909855723381,
      "num_tokens": 169665732.0,
      "step": 141040
    },
    {
      "entropy": 1.8929132372140884,
      "epoch": 0.43724265882595464,
      "grad_norm": 7.813331127166748,
      "learning_rate": 3.825894197750575e-06,
      "loss": 0.4269,
      "mean_token_accuracy": 0.8591722935438156,
      "num_tokens": 169677613.0,
      "step": 141050
    },
    {
      "entropy": 1.9381250068545341,
      "epoch": 0.4372736579510043,
      "grad_norm": 8.72669792175293,
      "learning_rate": 3.825758582085967e-06,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8587992191314697,
      "num_tokens": 169689244.0,
      "step": 141060
    },
    {
      "entropy": 1.7425117641687393,
      "epoch": 0.43730465707605404,
      "grad_norm": 8.140493392944336,
      "learning_rate": 3.825622980841757e-06,
      "loss": 0.3859,
      "mean_token_accuracy": 0.8660785302519798,
      "num_tokens": 169703991.0,
      "step": 141070
    },
    {
      "entropy": 1.8813914477825164,
      "epoch": 0.4373356562011037,
      "grad_norm": 9.09639835357666,
      "learning_rate": 3.825487394015388e-06,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8516086161136627,
      "num_tokens": 169716782.0,
      "step": 141080
    },
    {
      "entropy": 1.8726382121443748,
      "epoch": 0.43736665532615343,
      "grad_norm": 4.414510250091553,
      "learning_rate": 3.825351821604306e-06,
      "loss": 0.455,
      "mean_token_accuracy": 0.8511142015457154,
      "num_tokens": 169729151.0,
      "step": 141090
    },
    {
      "entropy": 1.905012820661068,
      "epoch": 0.4373976544512031,
      "grad_norm": 8.519266128540039,
      "learning_rate": 3.825216263605957e-06,
      "loss": 0.477,
      "mean_token_accuracy": 0.8418341457843781,
      "num_tokens": 169741404.0,
      "step": 141100
    },
    {
      "entropy": 1.9148242220282554,
      "epoch": 0.4374286535762528,
      "grad_norm": 8.985989570617676,
      "learning_rate": 3.825080720017787e-06,
      "loss": 0.4941,
      "mean_token_accuracy": 0.842382799088955,
      "num_tokens": 169753529.0,
      "step": 141110
    },
    {
      "entropy": 1.8772297248244285,
      "epoch": 0.4374596527013025,
      "grad_norm": 3.817615509033203,
      "learning_rate": 3.824945190837244e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8559065282344818,
      "num_tokens": 169765965.0,
      "step": 141120
    },
    {
      "entropy": 1.9383199408650398,
      "epoch": 0.4374906518263522,
      "grad_norm": 3.4553632736206055,
      "learning_rate": 3.824809676061776e-06,
      "loss": 0.468,
      "mean_token_accuracy": 0.8500086024403573,
      "num_tokens": 169777683.0,
      "step": 141130
    },
    {
      "entropy": 1.9334457024931908,
      "epoch": 0.4375216509514019,
      "grad_norm": 4.0566182136535645,
      "learning_rate": 3.82467417568883e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8465969949960709,
      "num_tokens": 169789469.0,
      "step": 141140
    },
    {
      "entropy": 1.9533229991793633,
      "epoch": 0.4375526500764516,
      "grad_norm": 7.992478847503662,
      "learning_rate": 3.824538689715855e-06,
      "loss": 0.4447,
      "mean_token_accuracy": 0.8505326002836228,
      "num_tokens": 169800995.0,
      "step": 141150
    },
    {
      "entropy": 1.9198670163750648,
      "epoch": 0.4375836492015013,
      "grad_norm": 4.070978164672852,
      "learning_rate": 3.8244032181403015e-06,
      "loss": 0.5319,
      "mean_token_accuracy": 0.8470968812704086,
      "num_tokens": 169813330.0,
      "step": 141160
    },
    {
      "entropy": 1.8759153246879579,
      "epoch": 0.437614648326551,
      "grad_norm": 10.018072128295898,
      "learning_rate": 3.8242677609596205e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8579384982585907,
      "num_tokens": 169826267.0,
      "step": 141170
    },
    {
      "entropy": 1.839848317205906,
      "epoch": 0.4376456474516007,
      "grad_norm": 8.282307624816895,
      "learning_rate": 3.824132318171262e-06,
      "loss": 0.4227,
      "mean_token_accuracy": 0.8613649949431419,
      "num_tokens": 169840238.0,
      "step": 141180
    },
    {
      "entropy": 1.9160371258854867,
      "epoch": 0.4376766465766504,
      "grad_norm": 8.569047927856445,
      "learning_rate": 3.8239968897726755e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.8511138498783112,
      "num_tokens": 169852089.0,
      "step": 141190
    },
    {
      "entropy": 1.9733612596988679,
      "epoch": 0.43770764570170007,
      "grad_norm": 7.161515712738037,
      "learning_rate": 3.823861475761317e-06,
      "loss": 0.4506,
      "mean_token_accuracy": 0.8522224485874176,
      "num_tokens": 169863089.0,
      "step": 141200
    },
    {
      "entropy": 1.9556746795773505,
      "epoch": 0.4377386448267498,
      "grad_norm": 8.775317192077637,
      "learning_rate": 3.823726076134636e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8549538642168045,
      "num_tokens": 169875115.0,
      "step": 141210
    },
    {
      "entropy": 1.9240303069353104,
      "epoch": 0.43776964395179946,
      "grad_norm": 8.23504638671875,
      "learning_rate": 3.823590690890089e-06,
      "loss": 0.469,
      "mean_token_accuracy": 0.853808656334877,
      "num_tokens": 169887042.0,
      "step": 141220
    },
    {
      "entropy": 1.8046733349561692,
      "epoch": 0.4378006430768492,
      "grad_norm": 3.5888330936431885,
      "learning_rate": 3.823455320025128e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8697149589657783,
      "num_tokens": 169900449.0,
      "step": 141230
    },
    {
      "entropy": 1.8609862461686135,
      "epoch": 0.43783164220189885,
      "grad_norm": 3.756772041320801,
      "learning_rate": 3.823319963537208e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8570352002978325,
      "num_tokens": 169912988.0,
      "step": 141240
    },
    {
      "entropy": 1.8794613614678384,
      "epoch": 0.4378626413269485,
      "grad_norm": 3.9449973106384277,
      "learning_rate": 3.823184621423784e-06,
      "loss": 0.4047,
      "mean_token_accuracy": 0.8526434049010276,
      "num_tokens": 169926018.0,
      "step": 141250
    },
    {
      "entropy": 1.7661522284150124,
      "epoch": 0.43789364045199825,
      "grad_norm": 9.034821510314941,
      "learning_rate": 3.8230492936823135e-06,
      "loss": 0.355,
      "mean_token_accuracy": 0.8720221489667892,
      "num_tokens": 169939851.0,
      "step": 141260
    },
    {
      "entropy": 1.978477604687214,
      "epoch": 0.4379246395770479,
      "grad_norm": 7.994080543518066,
      "learning_rate": 3.822913980310252e-06,
      "loss": 0.4984,
      "mean_token_accuracy": 0.8487569943070412,
      "num_tokens": 169951672.0,
      "step": 141270
    },
    {
      "entropy": 1.9159216687083245,
      "epoch": 0.43795563870209764,
      "grad_norm": 7.902944564819336,
      "learning_rate": 3.822778681305056e-06,
      "loss": 0.4246,
      "mean_token_accuracy": 0.8606354176998139,
      "num_tokens": 169963211.0,
      "step": 141280
    },
    {
      "entropy": 1.8818385779857636,
      "epoch": 0.4379866378271473,
      "grad_norm": 4.299989700317383,
      "learning_rate": 3.822643396664184e-06,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8490740895271301,
      "num_tokens": 169975835.0,
      "step": 141290
    },
    {
      "entropy": 1.9505405515432357,
      "epoch": 0.43801763695219703,
      "grad_norm": 7.912982940673828,
      "learning_rate": 3.822508126385095e-06,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8519005611538887,
      "num_tokens": 169987046.0,
      "step": 141300
    },
    {
      "entropy": 1.8943722754716874,
      "epoch": 0.4380486360772467,
      "grad_norm": 9.575815200805664,
      "learning_rate": 3.822372870465247e-06,
      "loss": 0.424,
      "mean_token_accuracy": 0.8552343189716339,
      "num_tokens": 169999313.0,
      "step": 141310
    },
    {
      "entropy": 2.0350208193063737,
      "epoch": 0.4380796352022964,
      "grad_norm": 9.12436580657959,
      "learning_rate": 3.822237628902101e-06,
      "loss": 0.5847,
      "mean_token_accuracy": 0.8212827384471894,
      "num_tokens": 170010308.0,
      "step": 141320
    },
    {
      "entropy": 1.9424827635288238,
      "epoch": 0.4381106343273461,
      "grad_norm": 7.913768291473389,
      "learning_rate": 3.822102401693118e-06,
      "loss": 0.4541,
      "mean_token_accuracy": 0.8558102324604988,
      "num_tokens": 170021453.0,
      "step": 141330
    },
    {
      "entropy": 1.9400149762630463,
      "epoch": 0.4381416334523958,
      "grad_norm": 7.435611724853516,
      "learning_rate": 3.821967188835756e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8472892969846726,
      "num_tokens": 170033154.0,
      "step": 141340
    },
    {
      "entropy": 1.9589600950479507,
      "epoch": 0.4381726325774455,
      "grad_norm": 4.710630416870117,
      "learning_rate": 3.82183199032748e-06,
      "loss": 0.4518,
      "mean_token_accuracy": 0.8444600135087967,
      "num_tokens": 170044448.0,
      "step": 141350
    },
    {
      "entropy": 2.002592481672764,
      "epoch": 0.4382036317024952,
      "grad_norm": 3.6701736450195312,
      "learning_rate": 3.821696806165749e-06,
      "loss": 0.4958,
      "mean_token_accuracy": 0.8414113700389863,
      "num_tokens": 170055393.0,
      "step": 141360
    },
    {
      "entropy": 1.9110810875892639,
      "epoch": 0.4382346308275449,
      "grad_norm": 8.964020729064941,
      "learning_rate": 3.821561636348028e-06,
      "loss": 0.4121,
      "mean_token_accuracy": 0.8573357686400414,
      "num_tokens": 170067640.0,
      "step": 141370
    },
    {
      "entropy": 1.9764687582850455,
      "epoch": 0.4382656299525946,
      "grad_norm": 8.130205154418945,
      "learning_rate": 3.8214264808717814e-06,
      "loss": 0.5127,
      "mean_token_accuracy": 0.8418732464313508,
      "num_tokens": 170079057.0,
      "step": 141380
    },
    {
      "entropy": 1.8818054497241974,
      "epoch": 0.4382966290776443,
      "grad_norm": 4.381757736206055,
      "learning_rate": 3.82129133973447e-06,
      "loss": 0.4403,
      "mean_token_accuracy": 0.8557385444641114,
      "num_tokens": 170091217.0,
      "step": 141390
    },
    {
      "entropy": 1.999481311440468,
      "epoch": 0.438327628202694,
      "grad_norm": 8.559086799621582,
      "learning_rate": 3.821156212933562e-06,
      "loss": 0.5061,
      "mean_token_accuracy": 0.8520580142736435,
      "num_tokens": 170102107.0,
      "step": 141400
    },
    {
      "entropy": 1.877190275490284,
      "epoch": 0.43835862732774367,
      "grad_norm": 7.804668426513672,
      "learning_rate": 3.8210211004665206e-06,
      "loss": 0.4605,
      "mean_token_accuracy": 0.8549025520682335,
      "num_tokens": 170114007.0,
      "step": 141410
    },
    {
      "entropy": 1.7784633502364158,
      "epoch": 0.4383896264527934,
      "grad_norm": 7.331371784210205,
      "learning_rate": 3.820886002330814e-06,
      "loss": 0.3455,
      "mean_token_accuracy": 0.8712356150150299,
      "num_tokens": 170127600.0,
      "step": 141420
    },
    {
      "entropy": 1.948923571407795,
      "epoch": 0.43842062557784306,
      "grad_norm": 9.89287281036377,
      "learning_rate": 3.820750918523906e-06,
      "loss": 0.4901,
      "mean_token_accuracy": 0.8478708654642105,
      "num_tokens": 170138776.0,
      "step": 141430
    },
    {
      "entropy": 1.9679486066102982,
      "epoch": 0.4384516247028928,
      "grad_norm": 9.170439720153809,
      "learning_rate": 3.820615849043266e-06,
      "loss": 0.4568,
      "mean_token_accuracy": 0.8551394686102867,
      "num_tokens": 170150407.0,
      "step": 141440
    },
    {
      "entropy": 1.9057482168078423,
      "epoch": 0.43848262382794245,
      "grad_norm": 9.319578170776367,
      "learning_rate": 3.820480793886361e-06,
      "loss": 0.4794,
      "mean_token_accuracy": 0.8556445702910424,
      "num_tokens": 170162841.0,
      "step": 141450
    },
    {
      "entropy": 1.872725434601307,
      "epoch": 0.4385136229529922,
      "grad_norm": 8.340863227844238,
      "learning_rate": 3.820345753050659e-06,
      "loss": 0.4323,
      "mean_token_accuracy": 0.8682135134935379,
      "num_tokens": 170175371.0,
      "step": 141460
    },
    {
      "entropy": 1.8802336245775222,
      "epoch": 0.43854462207804185,
      "grad_norm": 8.725083351135254,
      "learning_rate": 3.82021072653363e-06,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8557636126875877,
      "num_tokens": 170187618.0,
      "step": 141470
    },
    {
      "entropy": 1.8927005499601364,
      "epoch": 0.43857562120309157,
      "grad_norm": 3.3288369178771973,
      "learning_rate": 3.820075714332744e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8532668456435204,
      "num_tokens": 170199475.0,
      "step": 141480
    },
    {
      "entropy": 1.996702989935875,
      "epoch": 0.43860662032814124,
      "grad_norm": 7.832830429077148,
      "learning_rate": 3.819940716445472e-06,
      "loss": 0.5113,
      "mean_token_accuracy": 0.8426687330007553,
      "num_tokens": 170210531.0,
      "step": 141490
    },
    {
      "entropy": 1.9969799607992171,
      "epoch": 0.4386376194531909,
      "grad_norm": 8.912454605102539,
      "learning_rate": 3.819805732869283e-06,
      "loss": 0.5241,
      "mean_token_accuracy": 0.8406167760491371,
      "num_tokens": 170221403.0,
      "step": 141500
    },
    {
      "entropy": 1.9903255164623261,
      "epoch": 0.43866861857824063,
      "grad_norm": 8.586922645568848,
      "learning_rate": 3.81967076360165e-06,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8494381055235862,
      "num_tokens": 170232697.0,
      "step": 141510
    },
    {
      "entropy": 1.893722152709961,
      "epoch": 0.4386996177032903,
      "grad_norm": 7.579135894775391,
      "learning_rate": 3.819535808640045e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8595759615302085,
      "num_tokens": 170244794.0,
      "step": 141520
    },
    {
      "entropy": 1.9859010726213455,
      "epoch": 0.43873061682834,
      "grad_norm": 8.621550559997559,
      "learning_rate": 3.819400867981941e-06,
      "loss": 0.5332,
      "mean_token_accuracy": 0.8447252556681633,
      "num_tokens": 170255961.0,
      "step": 141530
    },
    {
      "entropy": 1.945717729628086,
      "epoch": 0.4387616159533897,
      "grad_norm": 8.728808403015137,
      "learning_rate": 3.819265941624811e-06,
      "loss": 0.4971,
      "mean_token_accuracy": 0.8427330657839776,
      "num_tokens": 170268226.0,
      "step": 141540
    },
    {
      "entropy": 2.0041126042604445,
      "epoch": 0.4387926150784394,
      "grad_norm": 10.106131553649902,
      "learning_rate": 3.819131029566131e-06,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8371750578284264,
      "num_tokens": 170278720.0,
      "step": 141550
    },
    {
      "entropy": 1.9148709684610368,
      "epoch": 0.4388236142034891,
      "grad_norm": 9.913314819335938,
      "learning_rate": 3.818996131803373e-06,
      "loss": 0.4727,
      "mean_token_accuracy": 0.840086530148983,
      "num_tokens": 170291028.0,
      "step": 141560
    },
    {
      "entropy": 1.9021180346608162,
      "epoch": 0.4388546133285388,
      "grad_norm": 7.955324649810791,
      "learning_rate": 3.818861248334014e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8594782516360283,
      "num_tokens": 170303129.0,
      "step": 141570
    },
    {
      "entropy": 1.8553086191415786,
      "epoch": 0.4388856124535885,
      "grad_norm": 4.323651313781738,
      "learning_rate": 3.81872637915553e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.854140630364418,
      "num_tokens": 170316049.0,
      "step": 141580
    },
    {
      "entropy": 1.9587162017822266,
      "epoch": 0.4389166115786382,
      "grad_norm": 8.201522827148438,
      "learning_rate": 3.818591524265398e-06,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8508410602807999,
      "num_tokens": 170327674.0,
      "step": 141590
    },
    {
      "entropy": 1.986707004904747,
      "epoch": 0.4389476107036879,
      "grad_norm": 7.647431373596191,
      "learning_rate": 3.8184566836610944e-06,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8480009466409684,
      "num_tokens": 170339108.0,
      "step": 141600
    },
    {
      "entropy": 1.948198501765728,
      "epoch": 0.4389786098287376,
      "grad_norm": 7.071403503417969,
      "learning_rate": 3.818321857340097e-06,
      "loss": 0.509,
      "mean_token_accuracy": 0.8380307227373123,
      "num_tokens": 170350766.0,
      "step": 141610
    },
    {
      "entropy": 1.9600090399384498,
      "epoch": 0.43900960895378727,
      "grad_norm": 9.295860290527344,
      "learning_rate": 3.818187045299886e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8459015518426896,
      "num_tokens": 170362080.0,
      "step": 141620
    },
    {
      "entropy": 1.9097954377532005,
      "epoch": 0.439040608078837,
      "grad_norm": 4.016225337982178,
      "learning_rate": 3.818052247537938e-06,
      "loss": 0.4367,
      "mean_token_accuracy": 0.8569407507777214,
      "num_tokens": 170374022.0,
      "step": 141630
    },
    {
      "entropy": 1.8951156988739968,
      "epoch": 0.43907160720388666,
      "grad_norm": 7.259555339813232,
      "learning_rate": 3.817917464051734e-06,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8556249186396598,
      "num_tokens": 170386951.0,
      "step": 141640
    },
    {
      "entropy": 1.9594153746962548,
      "epoch": 0.4391026063289364,
      "grad_norm": 7.583054065704346,
      "learning_rate": 3.817782694838756e-06,
      "loss": 0.4526,
      "mean_token_accuracy": 0.856129738688469,
      "num_tokens": 170398778.0,
      "step": 141650
    },
    {
      "entropy": 1.8969098508358002,
      "epoch": 0.43913360545398605,
      "grad_norm": 9.536150932312012,
      "learning_rate": 3.817647939896483e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8510769203305244,
      "num_tokens": 170411665.0,
      "step": 141660
    },
    {
      "entropy": 1.9632080033421517,
      "epoch": 0.4391646045790358,
      "grad_norm": 7.881152153015137,
      "learning_rate": 3.8175131992223965e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8432640329003334,
      "num_tokens": 170423508.0,
      "step": 141670
    },
    {
      "entropy": 1.8790015771985054,
      "epoch": 0.43919560370408545,
      "grad_norm": 9.358724594116211,
      "learning_rate": 3.81737847281398e-06,
      "loss": 0.4405,
      "mean_token_accuracy": 0.8637870714068413,
      "num_tokens": 170435638.0,
      "step": 141680
    },
    {
      "entropy": 1.8561220914125443,
      "epoch": 0.43922660282913517,
      "grad_norm": 8.335770606994629,
      "learning_rate": 3.8172437606687156e-06,
      "loss": 0.473,
      "mean_token_accuracy": 0.8470362722873688,
      "num_tokens": 170449929.0,
      "step": 141690
    },
    {
      "entropy": 1.9746687322854997,
      "epoch": 0.43925760195418484,
      "grad_norm": 7.2637248039245605,
      "learning_rate": 3.817109062784087e-06,
      "loss": 0.5127,
      "mean_token_accuracy": 0.842483501136303,
      "num_tokens": 170461051.0,
      "step": 141700
    },
    {
      "entropy": 1.9206561237573623,
      "epoch": 0.43928860107923456,
      "grad_norm": 9.621170997619629,
      "learning_rate": 3.816974379157578e-06,
      "loss": 0.4337,
      "mean_token_accuracy": 0.8608662933111191,
      "num_tokens": 170472540.0,
      "step": 141710
    },
    {
      "entropy": 1.9312569439411162,
      "epoch": 0.43931960020428423,
      "grad_norm": 9.245631217956543,
      "learning_rate": 3.816839709786675e-06,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8509629473090172,
      "num_tokens": 170484584.0,
      "step": 141720
    },
    {
      "entropy": 1.9452128455042839,
      "epoch": 0.4393505993293339,
      "grad_norm": 8.906986236572266,
      "learning_rate": 3.81670505466886e-06,
      "loss": 0.4482,
      "mean_token_accuracy": 0.8519314289093017,
      "num_tokens": 170496707.0,
      "step": 141730
    },
    {
      "entropy": 1.957577820122242,
      "epoch": 0.4393815984543836,
      "grad_norm": 7.901788234710693,
      "learning_rate": 3.816570413801622e-06,
      "loss": 0.4578,
      "mean_token_accuracy": 0.8558757767081261,
      "num_tokens": 170507715.0,
      "step": 141740
    },
    {
      "entropy": 1.9821037575602531,
      "epoch": 0.4394125975794333,
      "grad_norm": 7.910948276519775,
      "learning_rate": 3.8164357871824466e-06,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8592849254608155,
      "num_tokens": 170518782.0,
      "step": 141750
    },
    {
      "entropy": 1.9237025648355484,
      "epoch": 0.439443596704483,
      "grad_norm": 7.395272731781006,
      "learning_rate": 3.816301174808821e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.8577270656824112,
      "num_tokens": 170531656.0,
      "step": 141760
    },
    {
      "entropy": 1.9360353201627731,
      "epoch": 0.4394745958295327,
      "grad_norm": 4.2066426277160645,
      "learning_rate": 3.8161665766782335e-06,
      "loss": 0.4523,
      "mean_token_accuracy": 0.8494734823703766,
      "num_tokens": 170543484.0,
      "step": 141770
    },
    {
      "entropy": 1.969169418513775,
      "epoch": 0.4395055949545824,
      "grad_norm": 4.006796360015869,
      "learning_rate": 3.816031992788171e-06,
      "loss": 0.4693,
      "mean_token_accuracy": 0.8522426009178161,
      "num_tokens": 170555226.0,
      "step": 141780
    },
    {
      "entropy": 1.7786556363105774,
      "epoch": 0.4395365940796321,
      "grad_norm": 8.278726577758789,
      "learning_rate": 3.815897423136125e-06,
      "loss": 0.3426,
      "mean_token_accuracy": 0.8771900832653046,
      "num_tokens": 170568645.0,
      "step": 141790
    },
    {
      "entropy": 1.9889621213078499,
      "epoch": 0.4395675932046818,
      "grad_norm": 8.236480712890625,
      "learning_rate": 3.815762867719584e-06,
      "loss": 0.4572,
      "mean_token_accuracy": 0.8558431312441825,
      "num_tokens": 170580292.0,
      "step": 141800
    },
    {
      "entropy": 1.8943669281899929,
      "epoch": 0.4395985923297315,
      "grad_norm": 8.767983436584473,
      "learning_rate": 3.815628326536037e-06,
      "loss": 0.462,
      "mean_token_accuracy": 0.8443626776337624,
      "num_tokens": 170593285.0,
      "step": 141810
    },
    {
      "entropy": 1.9128671914339066,
      "epoch": 0.4396295914547812,
      "grad_norm": 3.5375518798828125,
      "learning_rate": 3.815493799582977e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8669633626937866,
      "num_tokens": 170605706.0,
      "step": 141820
    },
    {
      "entropy": 1.985709111392498,
      "epoch": 0.43966059057983087,
      "grad_norm": 7.396274566650391,
      "learning_rate": 3.815359286857895e-06,
      "loss": 0.4849,
      "mean_token_accuracy": 0.8488274723291397,
      "num_tokens": 170617185.0,
      "step": 141830
    },
    {
      "entropy": 1.9252109676599503,
      "epoch": 0.4396915897048806,
      "grad_norm": 5.221693515777588,
      "learning_rate": 3.815224788358284e-06,
      "loss": 0.4519,
      "mean_token_accuracy": 0.8559987634420395,
      "num_tokens": 170629091.0,
      "step": 141840
    },
    {
      "entropy": 1.8718436300754546,
      "epoch": 0.43972258882993026,
      "grad_norm": 4.875570297241211,
      "learning_rate": 3.815090304081635e-06,
      "loss": 0.4358,
      "mean_token_accuracy": 0.8549114823341369,
      "num_tokens": 170642049.0,
      "step": 141850
    },
    {
      "entropy": 1.8797766268253326,
      "epoch": 0.43975358795498,
      "grad_norm": 7.365478515625,
      "learning_rate": 3.814955834025442e-06,
      "loss": 0.4081,
      "mean_token_accuracy": 0.8585930600762367,
      "num_tokens": 170654668.0,
      "step": 141860
    },
    {
      "entropy": 1.9316943183541297,
      "epoch": 0.43978458708002965,
      "grad_norm": 4.897908687591553,
      "learning_rate": 3.814821378187199e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8544142752885818,
      "num_tokens": 170666437.0,
      "step": 141870
    },
    {
      "entropy": 1.799608789384365,
      "epoch": 0.4398155862050794,
      "grad_norm": 4.87161922454834,
      "learning_rate": 3.814686936564401e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8723533883690834,
      "num_tokens": 170680355.0,
      "step": 141880
    },
    {
      "entropy": 1.9116061985492707,
      "epoch": 0.43984658533012905,
      "grad_norm": 7.900229454040527,
      "learning_rate": 3.814552509154544e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8542070344090462,
      "num_tokens": 170691960.0,
      "step": 141890
    },
    {
      "entropy": 1.9224073961377144,
      "epoch": 0.43987758445517877,
      "grad_norm": 7.187701225280762,
      "learning_rate": 3.8144180959551223e-06,
      "loss": 0.4004,
      "mean_token_accuracy": 0.8590036764740944,
      "num_tokens": 170703505.0,
      "step": 141900
    },
    {
      "entropy": 1.8964632406830788,
      "epoch": 0.43990858358022844,
      "grad_norm": 7.443815231323242,
      "learning_rate": 3.8142836969636336e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8544583544135094,
      "num_tokens": 170715715.0,
      "step": 141910
    },
    {
      "entropy": 1.8948586672544478,
      "epoch": 0.43993958270527816,
      "grad_norm": 8.835453987121582,
      "learning_rate": 3.8141493121775747e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8569506093859672,
      "num_tokens": 170728395.0,
      "step": 141920
    },
    {
      "entropy": 1.925899577140808,
      "epoch": 0.43997058183032783,
      "grad_norm": 7.583258628845215,
      "learning_rate": 3.814014941594443e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.850526462495327,
      "num_tokens": 170740412.0,
      "step": 141930
    },
    {
      "entropy": 1.9305981874465943,
      "epoch": 0.44000158095537756,
      "grad_norm": 3.2110064029693604,
      "learning_rate": 3.8138805852117377e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8527848124504089,
      "num_tokens": 170752268.0,
      "step": 141940
    },
    {
      "entropy": 1.8623799979686737,
      "epoch": 0.4400325800804272,
      "grad_norm": 7.726592540740967,
      "learning_rate": 3.8137462430269565e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8616868555545807,
      "num_tokens": 170764922.0,
      "step": 141950
    },
    {
      "entropy": 1.9484892964363099,
      "epoch": 0.44006357920547695,
      "grad_norm": 8.359121322631836,
      "learning_rate": 3.8136119150376006e-06,
      "loss": 0.4418,
      "mean_token_accuracy": 0.8625785291194916,
      "num_tokens": 170776660.0,
      "step": 141960
    },
    {
      "entropy": 1.898753324151039,
      "epoch": 0.4400945783305266,
      "grad_norm": 3.5382606983184814,
      "learning_rate": 3.8134776012411693e-06,
      "loss": 0.4573,
      "mean_token_accuracy": 0.8586586147546769,
      "num_tokens": 170789396.0,
      "step": 141970
    },
    {
      "entropy": 1.8218749940395356,
      "epoch": 0.4401255774555763,
      "grad_norm": 7.84831428527832,
      "learning_rate": 3.8133433016351636e-06,
      "loss": 0.3302,
      "mean_token_accuracy": 0.8637544587254524,
      "num_tokens": 170802890.0,
      "step": 141980
    },
    {
      "entropy": 1.9783765748143196,
      "epoch": 0.440156576580626,
      "grad_norm": 7.985630512237549,
      "learning_rate": 3.813209016217084e-06,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8482358425855636,
      "num_tokens": 170814592.0,
      "step": 141990
    },
    {
      "entropy": 1.8626375451683999,
      "epoch": 0.4401875757056757,
      "grad_norm": 4.535806179046631,
      "learning_rate": 3.8130747449844342e-06,
      "loss": 0.3822,
      "mean_token_accuracy": 0.8576488435268402,
      "num_tokens": 170828197.0,
      "step": 142000
    },
    {
      "entropy": 1.9288464725017547,
      "epoch": 0.4402185748307254,
      "grad_norm": 4.06549596786499,
      "learning_rate": 3.812940487934716e-06,
      "loss": 0.4301,
      "mean_token_accuracy": 0.8626020058989525,
      "num_tokens": 170840003.0,
      "step": 142010
    },
    {
      "entropy": 1.9120705425739288,
      "epoch": 0.4402495739557751,
      "grad_norm": 7.126809597015381,
      "learning_rate": 3.8128062450654323e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.8442542180418968,
      "num_tokens": 170852921.0,
      "step": 142020
    },
    {
      "entropy": 1.9289035245776176,
      "epoch": 0.4402805730808248,
      "grad_norm": 7.8247246742248535,
      "learning_rate": 3.8126720163740884e-06,
      "loss": 0.4411,
      "mean_token_accuracy": 0.846360231935978,
      "num_tokens": 170866003.0,
      "step": 142030
    },
    {
      "entropy": 1.9721984177827836,
      "epoch": 0.44031157220587447,
      "grad_norm": 8.725114822387695,
      "learning_rate": 3.8125378018581864e-06,
      "loss": 0.4929,
      "mean_token_accuracy": 0.85105609446764,
      "num_tokens": 170877690.0,
      "step": 142040
    },
    {
      "entropy": 1.8866790056228637,
      "epoch": 0.4403425713309242,
      "grad_norm": 9.154146194458008,
      "learning_rate": 3.8124036015152344e-06,
      "loss": 0.4527,
      "mean_token_accuracy": 0.8527222231030465,
      "num_tokens": 170890766.0,
      "step": 142050
    },
    {
      "entropy": 1.9038648203015327,
      "epoch": 0.44037357045597386,
      "grad_norm": 9.711874008178711,
      "learning_rate": 3.8122694153427362e-06,
      "loss": 0.4257,
      "mean_token_accuracy": 0.8540343955159188,
      "num_tokens": 170902831.0,
      "step": 142060
    },
    {
      "entropy": 1.946737252175808,
      "epoch": 0.4404045695810236,
      "grad_norm": 6.8612380027771,
      "learning_rate": 3.8121352433381986e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8524143502116204,
      "num_tokens": 170914196.0,
      "step": 142070
    },
    {
      "entropy": 1.9659889578819274,
      "epoch": 0.44043556870607325,
      "grad_norm": 7.955692768096924,
      "learning_rate": 3.812001085499129e-06,
      "loss": 0.556,
      "mean_token_accuracy": 0.8456687927246094,
      "num_tokens": 170925736.0,
      "step": 142080
    },
    {
      "entropy": 1.918195514380932,
      "epoch": 0.440466567831123,
      "grad_norm": 4.864497184753418,
      "learning_rate": 3.811866941823035e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8531820371747016,
      "num_tokens": 170937320.0,
      "step": 142090
    },
    {
      "entropy": 1.924280734360218,
      "epoch": 0.44049756695617265,
      "grad_norm": 8.45577335357666,
      "learning_rate": 3.8117328123074237e-06,
      "loss": 0.4733,
      "mean_token_accuracy": 0.8453326195478439,
      "num_tokens": 170949120.0,
      "step": 142100
    },
    {
      "entropy": 1.8904883772134782,
      "epoch": 0.4405285660812224,
      "grad_norm": 3.6915700435638428,
      "learning_rate": 3.8115986969498047e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8524854764342308,
      "num_tokens": 170960950.0,
      "step": 142110
    },
    {
      "entropy": 1.787447765469551,
      "epoch": 0.44055956520627204,
      "grad_norm": 2.5302090644836426,
      "learning_rate": 3.811464595747688e-06,
      "loss": 0.376,
      "mean_token_accuracy": 0.8622151434421539,
      "num_tokens": 170974946.0,
      "step": 142120
    },
    {
      "entropy": 1.7317695260047912,
      "epoch": 0.44059056433132177,
      "grad_norm": 4.011857509613037,
      "learning_rate": 3.811330508698583e-06,
      "loss": 0.3431,
      "mean_token_accuracy": 0.8648561984300613,
      "num_tokens": 170989419.0,
      "step": 142130
    },
    {
      "entropy": 1.9078683421015739,
      "epoch": 0.44062156345637143,
      "grad_norm": 4.096642971038818,
      "learning_rate": 3.8111964358000005e-06,
      "loss": 0.4389,
      "mean_token_accuracy": 0.8534543007612229,
      "num_tokens": 171001053.0,
      "step": 142140
    },
    {
      "entropy": 1.938607893884182,
      "epoch": 0.44065256258142116,
      "grad_norm": 8.272358894348145,
      "learning_rate": 3.8110623770494515e-06,
      "loss": 0.4844,
      "mean_token_accuracy": 0.8414226740598678,
      "num_tokens": 171012985.0,
      "step": 142150
    },
    {
      "entropy": 1.9575382590293884,
      "epoch": 0.4406835617064708,
      "grad_norm": 8.397919654846191,
      "learning_rate": 3.8109283324444484e-06,
      "loss": 0.4773,
      "mean_token_accuracy": 0.8492600947618485,
      "num_tokens": 171023779.0,
      "step": 142160
    },
    {
      "entropy": 1.935109880566597,
      "epoch": 0.44071456083152055,
      "grad_norm": 6.095836162567139,
      "learning_rate": 3.8107943019825027e-06,
      "loss": 0.4592,
      "mean_token_accuracy": 0.8479932472109795,
      "num_tokens": 171035289.0,
      "step": 142170
    },
    {
      "entropy": 1.952261172235012,
      "epoch": 0.4407455599565702,
      "grad_norm": 7.735157489776611,
      "learning_rate": 3.8106602856611296e-06,
      "loss": 0.579,
      "mean_token_accuracy": 0.8471168518066406,
      "num_tokens": 171047459.0,
      "step": 142180
    },
    {
      "entropy": 1.929404976963997,
      "epoch": 0.44077655908161995,
      "grad_norm": 3.8500773906707764,
      "learning_rate": 3.8105262834778404e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8393190920352935,
      "num_tokens": 171059462.0,
      "step": 142190
    },
    {
      "entropy": 1.8857171356678009,
      "epoch": 0.4408075582066696,
      "grad_norm": 8.527490615844727,
      "learning_rate": 3.8103922954301514e-06,
      "loss": 0.4077,
      "mean_token_accuracy": 0.8652898028492928,
      "num_tokens": 171071937.0,
      "step": 142200
    },
    {
      "entropy": 1.9904403433203697,
      "epoch": 0.44083855733171934,
      "grad_norm": 9.216706275939941,
      "learning_rate": 3.810258321515576e-06,
      "loss": 0.5001,
      "mean_token_accuracy": 0.8476386234164238,
      "num_tokens": 171083344.0,
      "step": 142210
    },
    {
      "entropy": 1.8266815394163132,
      "epoch": 0.440869556456769,
      "grad_norm": 9.037412643432617,
      "learning_rate": 3.81012436173163e-06,
      "loss": 0.4039,
      "mean_token_accuracy": 0.855592280626297,
      "num_tokens": 171097604.0,
      "step": 142220
    },
    {
      "entropy": 1.963519898056984,
      "epoch": 0.4409005555818187,
      "grad_norm": 7.543654441833496,
      "learning_rate": 3.809990416075832e-06,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8579617440700531,
      "num_tokens": 171108370.0,
      "step": 142230
    },
    {
      "entropy": 1.9259283766150475,
      "epoch": 0.4409315547068684,
      "grad_norm": 7.956765174865723,
      "learning_rate": 3.809856484545695e-06,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8434132248163223,
      "num_tokens": 171120766.0,
      "step": 142240
    },
    {
      "entropy": 1.9885295629501343,
      "epoch": 0.44096255383191807,
      "grad_norm": 8.011740684509277,
      "learning_rate": 3.8097225671387384e-06,
      "loss": 0.4909,
      "mean_token_accuracy": 0.8473825648427009,
      "num_tokens": 171131762.0,
      "step": 142250
    },
    {
      "entropy": 1.993890830874443,
      "epoch": 0.4409935529569678,
      "grad_norm": 8.500561714172363,
      "learning_rate": 3.80958866385248e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8485628053545952,
      "num_tokens": 171142751.0,
      "step": 142260
    },
    {
      "entropy": 1.937794804573059,
      "epoch": 0.44102455208201746,
      "grad_norm": 4.054864883422852,
      "learning_rate": 3.8094547746844392e-06,
      "loss": 0.4594,
      "mean_token_accuracy": 0.8569764137268067,
      "num_tokens": 171155116.0,
      "step": 142270
    },
    {
      "entropy": 1.9017036750912666,
      "epoch": 0.4410555512070672,
      "grad_norm": 7.951387882232666,
      "learning_rate": 3.809320899632134e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8537882640957832,
      "num_tokens": 171167623.0,
      "step": 142280
    },
    {
      "entropy": 1.9774706676602363,
      "epoch": 0.44108655033211686,
      "grad_norm": 9.138680458068848,
      "learning_rate": 3.809187038693085e-06,
      "loss": 0.5366,
      "mean_token_accuracy": 0.8372975870966911,
      "num_tokens": 171179230.0,
      "step": 142290
    },
    {
      "entropy": 1.9862707808613778,
      "epoch": 0.4411175494571666,
      "grad_norm": 7.265309810638428,
      "learning_rate": 3.809053191864811e-06,
      "loss": 0.4819,
      "mean_token_accuracy": 0.8446347713470459,
      "num_tokens": 171190091.0,
      "step": 142300
    },
    {
      "entropy": 1.9375395745038986,
      "epoch": 0.44114854858221625,
      "grad_norm": 8.636645317077637,
      "learning_rate": 3.808919359144836e-06,
      "loss": 0.4197,
      "mean_token_accuracy": 0.8591432020068168,
      "num_tokens": 171201834.0,
      "step": 142310
    },
    {
      "entropy": 1.905143465101719,
      "epoch": 0.441179547707266,
      "grad_norm": 4.498809814453125,
      "learning_rate": 3.8087855405306796e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8690479651093483,
      "num_tokens": 171213960.0,
      "step": 142320
    },
    {
      "entropy": 1.9409716859459878,
      "epoch": 0.44121054683231564,
      "grad_norm": 8.317008972167969,
      "learning_rate": 3.8086517360198645e-06,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8425351768732071,
      "num_tokens": 171225664.0,
      "step": 142330
    },
    {
      "entropy": 1.9071755453944206,
      "epoch": 0.44124154595736537,
      "grad_norm": 7.693809986114502,
      "learning_rate": 3.8085179456099135e-06,
      "loss": 0.4228,
      "mean_token_accuracy": 0.8560735136270523,
      "num_tokens": 171238142.0,
      "step": 142340
    },
    {
      "entropy": 1.8364718183875084,
      "epoch": 0.44127254508241504,
      "grad_norm": 5.97056245803833,
      "learning_rate": 3.80838416929835e-06,
      "loss": 0.395,
      "mean_token_accuracy": 0.8731970608234405,
      "num_tokens": 171251102.0,
      "step": 142350
    },
    {
      "entropy": 1.8636333227157593,
      "epoch": 0.44130354420746476,
      "grad_norm": 4.0348920822143555,
      "learning_rate": 3.8082504070826986e-06,
      "loss": 0.4076,
      "mean_token_accuracy": 0.863330303132534,
      "num_tokens": 171263656.0,
      "step": 142360
    },
    {
      "entropy": 1.8688743382692337,
      "epoch": 0.44133454333251443,
      "grad_norm": 8.717557907104492,
      "learning_rate": 3.8081166589604846e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.8605389028787613,
      "num_tokens": 171276463.0,
      "step": 142370
    },
    {
      "entropy": 1.851096446812153,
      "epoch": 0.44136554245756415,
      "grad_norm": 7.275476455688477,
      "learning_rate": 3.8079829249292312e-06,
      "loss": 0.4234,
      "mean_token_accuracy": 0.8582789734005928,
      "num_tokens": 171289605.0,
      "step": 142380
    },
    {
      "entropy": 1.966257530450821,
      "epoch": 0.4413965415826138,
      "grad_norm": 9.199238777160645,
      "learning_rate": 3.8078492049864664e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8538633704185485,
      "num_tokens": 171300168.0,
      "step": 142390
    },
    {
      "entropy": 1.7878931298851968,
      "epoch": 0.44142754070766355,
      "grad_norm": 3.798326253890991,
      "learning_rate": 3.8077154991297155e-06,
      "loss": 0.3512,
      "mean_token_accuracy": 0.863753379881382,
      "num_tokens": 171314188.0,
      "step": 142400
    },
    {
      "entropy": 1.9457097873091698,
      "epoch": 0.4414585398327132,
      "grad_norm": 7.9168853759765625,
      "learning_rate": 3.8075818073565064e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8465751528739929,
      "num_tokens": 171326103.0,
      "step": 142410
    },
    {
      "entropy": 1.9705891758203506,
      "epoch": 0.44148953895776294,
      "grad_norm": 9.16375732421875,
      "learning_rate": 3.8074481296643662e-06,
      "loss": 0.4359,
      "mean_token_accuracy": 0.860154564678669,
      "num_tokens": 171337413.0,
      "step": 142420
    },
    {
      "entropy": 1.942754976451397,
      "epoch": 0.4415205380828126,
      "grad_norm": 8.25278091430664,
      "learning_rate": 3.807314466050824e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8554434284567833,
      "num_tokens": 171349655.0,
      "step": 142430
    },
    {
      "entropy": 1.9804495930671693,
      "epoch": 0.44155153720786233,
      "grad_norm": 7.4566144943237305,
      "learning_rate": 3.8071808165134087e-06,
      "loss": 0.5058,
      "mean_token_accuracy": 0.848778210580349,
      "num_tokens": 171360943.0,
      "step": 142440
    },
    {
      "entropy": 1.8187919482588768,
      "epoch": 0.441582536332912,
      "grad_norm": 7.4357686042785645,
      "learning_rate": 3.807047181049649e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.8577719181776047,
      "num_tokens": 171374928.0,
      "step": 142450
    },
    {
      "entropy": 1.9486502900719642,
      "epoch": 0.4416135354579617,
      "grad_norm": 8.6198148727417,
      "learning_rate": 3.8069135596570757e-06,
      "loss": 0.4717,
      "mean_token_accuracy": 0.8480415970087052,
      "num_tokens": 171387021.0,
      "step": 142460
    },
    {
      "entropy": 1.7527381375432014,
      "epoch": 0.4416445345830114,
      "grad_norm": 7.5862135887146,
      "learning_rate": 3.806779952333219e-06,
      "loss": 0.3672,
      "mean_token_accuracy": 0.8725428983569146,
      "num_tokens": 171401592.0,
      "step": 142470
    },
    {
      "entropy": 1.9407009825110435,
      "epoch": 0.44167553370806106,
      "grad_norm": 7.653254985809326,
      "learning_rate": 3.806646359075612e-06,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8616171032190323,
      "num_tokens": 171413307.0,
      "step": 142480
    },
    {
      "entropy": 1.8926226265728474,
      "epoch": 0.4417065328331108,
      "grad_norm": 7.93331241607666,
      "learning_rate": 3.806512779881785e-06,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8604073897004128,
      "num_tokens": 171427027.0,
      "step": 142490
    },
    {
      "entropy": 1.9086723767220974,
      "epoch": 0.44173753195816046,
      "grad_norm": 8.239239692687988,
      "learning_rate": 3.806379214749271e-06,
      "loss": 0.5015,
      "mean_token_accuracy": 0.8534449085593223,
      "num_tokens": 171440127.0,
      "step": 142500
    },
    {
      "entropy": 1.888797688484192,
      "epoch": 0.4417685310832102,
      "grad_norm": 8.310606956481934,
      "learning_rate": 3.8062456636756035e-06,
      "loss": 0.3774,
      "mean_token_accuracy": 0.871643802523613,
      "num_tokens": 171452073.0,
      "step": 142510
    },
    {
      "entropy": 1.8726611629128456,
      "epoch": 0.44179953020825985,
      "grad_norm": 8.1859769821167,
      "learning_rate": 3.8061121266583157e-06,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8519933164119721,
      "num_tokens": 171465272.0,
      "step": 142520
    },
    {
      "entropy": 1.9565669789910316,
      "epoch": 0.4418305293333096,
      "grad_norm": 7.003886699676514,
      "learning_rate": 3.805978603694943e-06,
      "loss": 0.4643,
      "mean_token_accuracy": 0.8577442169189453,
      "num_tokens": 171476604.0,
      "step": 142530
    },
    {
      "entropy": 1.969398957490921,
      "epoch": 0.44186152845835924,
      "grad_norm": 7.602597713470459,
      "learning_rate": 3.80584509478302e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8602959275245666,
      "num_tokens": 171487864.0,
      "step": 142540
    },
    {
      "entropy": 1.9313815072178842,
      "epoch": 0.44189252758340897,
      "grad_norm": 7.627027988433838,
      "learning_rate": 3.805711599920083e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8534037470817566,
      "num_tokens": 171500018.0,
      "step": 142550
    },
    {
      "entropy": 1.9397203966975212,
      "epoch": 0.44192352670845864,
      "grad_norm": 8.17835521697998,
      "learning_rate": 3.805578119103666e-06,
      "loss": 0.4608,
      "mean_token_accuracy": 0.8493491873145104,
      "num_tokens": 171511356.0,
      "step": 142560
    },
    {
      "entropy": 1.8530723094940185,
      "epoch": 0.44195452583350836,
      "grad_norm": 7.996511459350586,
      "learning_rate": 3.8054446523313083e-06,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8453282520174981,
      "num_tokens": 171524204.0,
      "step": 142570
    },
    {
      "entropy": 1.839161404967308,
      "epoch": 0.44198552495855803,
      "grad_norm": 4.661229133605957,
      "learning_rate": 3.8053111996005454e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.859870445728302,
      "num_tokens": 171537124.0,
      "step": 142580
    },
    {
      "entropy": 1.8566135600209237,
      "epoch": 0.44201652408360775,
      "grad_norm": 5.89265775680542,
      "learning_rate": 3.8051777609089174e-06,
      "loss": 0.4537,
      "mean_token_accuracy": 0.8573786094784737,
      "num_tokens": 171550107.0,
      "step": 142590
    },
    {
      "entropy": 1.8885900154709816,
      "epoch": 0.4420475232086574,
      "grad_norm": 7.972777366638184,
      "learning_rate": 3.805044336253962e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8563668861985206,
      "num_tokens": 171562634.0,
      "step": 142600
    },
    {
      "entropy": 1.9973913908004761,
      "epoch": 0.44207852233370715,
      "grad_norm": 7.949423789978027,
      "learning_rate": 3.8049109256332184e-06,
      "loss": 0.5469,
      "mean_token_accuracy": 0.8333679780364036,
      "num_tokens": 171574739.0,
      "step": 142610
    },
    {
      "entropy": 1.9654783993959426,
      "epoch": 0.4421095214587568,
      "grad_norm": 8.60360050201416,
      "learning_rate": 3.804777529044226e-06,
      "loss": 0.4183,
      "mean_token_accuracy": 0.8631428495049477,
      "num_tokens": 171585950.0,
      "step": 142620
    },
    {
      "entropy": 1.9139345183968544,
      "epoch": 0.44214052058380654,
      "grad_norm": 8.448734283447266,
      "learning_rate": 3.804644146484526e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.850536921620369,
      "num_tokens": 171597859.0,
      "step": 142630
    },
    {
      "entropy": 1.8951680943369866,
      "epoch": 0.4421715197088562,
      "grad_norm": 7.874819278717041,
      "learning_rate": 3.8045107779516586e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8422896340489388,
      "num_tokens": 171610294.0,
      "step": 142640
    },
    {
      "entropy": 1.9419591963291167,
      "epoch": 0.44220251883390593,
      "grad_norm": 3.6570918560028076,
      "learning_rate": 3.8043774234431667e-06,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8482140868902206,
      "num_tokens": 171621850.0,
      "step": 142650
    },
    {
      "entropy": 1.8761625073850154,
      "epoch": 0.4422335179589556,
      "grad_norm": 8.101895332336426,
      "learning_rate": 3.804244082956592e-06,
      "loss": 0.4294,
      "mean_token_accuracy": 0.8565763145685196,
      "num_tokens": 171634316.0,
      "step": 142660
    },
    {
      "entropy": 1.9388285502791405,
      "epoch": 0.4422645170840053,
      "grad_norm": 3.6656668186187744,
      "learning_rate": 3.8041107564894777e-06,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8403656214475632,
      "num_tokens": 171646140.0,
      "step": 142670
    },
    {
      "entropy": 1.9169174507260323,
      "epoch": 0.442295516209055,
      "grad_norm": 7.208846092224121,
      "learning_rate": 3.803977444039366e-06,
      "loss": 0.4243,
      "mean_token_accuracy": 0.8567602500319481,
      "num_tokens": 171657641.0,
      "step": 142680
    },
    {
      "entropy": 1.7894317671656608,
      "epoch": 0.4423265153341047,
      "grad_norm": 4.454036712646484,
      "learning_rate": 3.803844145603802e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8561735481023789,
      "num_tokens": 171671211.0,
      "step": 142690
    },
    {
      "entropy": 1.924549500644207,
      "epoch": 0.4423575144591544,
      "grad_norm": 4.878185272216797,
      "learning_rate": 3.803710861180331e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8529748469591141,
      "num_tokens": 171683021.0,
      "step": 142700
    },
    {
      "entropy": 1.926971609890461,
      "epoch": 0.4423885135842041,
      "grad_norm": 8.438478469848633,
      "learning_rate": 3.803577590766498e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8476522505283356,
      "num_tokens": 171694922.0,
      "step": 142710
    },
    {
      "entropy": 1.869713193178177,
      "epoch": 0.4424195127092538,
      "grad_norm": 8.490039825439453,
      "learning_rate": 3.8034443343598484e-06,
      "loss": 0.4056,
      "mean_token_accuracy": 0.8578268960118294,
      "num_tokens": 171707523.0,
      "step": 142720
    },
    {
      "entropy": 1.89549450725317,
      "epoch": 0.44245051183430345,
      "grad_norm": 3.9622802734375,
      "learning_rate": 3.8033110919579275e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.8647204831242561,
      "num_tokens": 171719274.0,
      "step": 142730
    },
    {
      "entropy": 1.9119060546159745,
      "epoch": 0.4424815109593532,
      "grad_norm": 10.171487808227539,
      "learning_rate": 3.8031778635582846e-06,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8421938866376877,
      "num_tokens": 171731484.0,
      "step": 142740
    },
    {
      "entropy": 1.9421584740281106,
      "epoch": 0.44251251008440284,
      "grad_norm": 7.809762001037598,
      "learning_rate": 3.803044649158467e-06,
      "loss": 0.4406,
      "mean_token_accuracy": 0.8524343490600585,
      "num_tokens": 171742555.0,
      "step": 142750
    },
    {
      "entropy": 1.8323628097772597,
      "epoch": 0.44254350920945257,
      "grad_norm": 3.0407345294952393,
      "learning_rate": 3.802911448756022e-06,
      "loss": 0.3788,
      "mean_token_accuracy": 0.8646909952163696,
      "num_tokens": 171756753.0,
      "step": 142760
    },
    {
      "entropy": 1.949809417128563,
      "epoch": 0.44257450833450224,
      "grad_norm": 8.330705642700195,
      "learning_rate": 3.8027782623484994e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8445712998509407,
      "num_tokens": 171768227.0,
      "step": 142770
    },
    {
      "entropy": 1.8988068416714667,
      "epoch": 0.44260550745955196,
      "grad_norm": 7.558119297027588,
      "learning_rate": 3.802645089933448e-06,
      "loss": 0.4338,
      "mean_token_accuracy": 0.8599656626582146,
      "num_tokens": 171779928.0,
      "step": 142780
    },
    {
      "entropy": 1.8191672816872597,
      "epoch": 0.44263650658460163,
      "grad_norm": 8.800172805786133,
      "learning_rate": 3.8025119315084186e-06,
      "loss": 0.408,
      "mean_token_accuracy": 0.8557739660143853,
      "num_tokens": 171793531.0,
      "step": 142790
    },
    {
      "entropy": 1.912230722606182,
      "epoch": 0.44266750570965135,
      "grad_norm": 8.029808044433594,
      "learning_rate": 3.802378787070961e-06,
      "loss": 0.4513,
      "mean_token_accuracy": 0.8631566151976585,
      "num_tokens": 171804841.0,
      "step": 142800
    },
    {
      "entropy": 1.9423863098025322,
      "epoch": 0.442698504834701,
      "grad_norm": 7.330270290374756,
      "learning_rate": 3.8022456566186274e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8568160951137542,
      "num_tokens": 171816045.0,
      "step": 142810
    },
    {
      "entropy": 1.909077961742878,
      "epoch": 0.44272950395975075,
      "grad_norm": 8.023503303527832,
      "learning_rate": 3.8021125401489695e-06,
      "loss": 0.4576,
      "mean_token_accuracy": 0.8499825984239578,
      "num_tokens": 171827873.0,
      "step": 142820
    },
    {
      "entropy": 1.914843738079071,
      "epoch": 0.4427605030848004,
      "grad_norm": 7.886246681213379,
      "learning_rate": 3.8019794376595393e-06,
      "loss": 0.4809,
      "mean_token_accuracy": 0.854096032679081,
      "num_tokens": 171839212.0,
      "step": 142830
    },
    {
      "entropy": 1.747120851278305,
      "epoch": 0.44279150220985014,
      "grad_norm": 4.407447338104248,
      "learning_rate": 3.8018463491478903e-06,
      "loss": 0.3271,
      "mean_token_accuracy": 0.8662238627672195,
      "num_tokens": 171853351.0,
      "step": 142840
    },
    {
      "entropy": 1.9104099452495575,
      "epoch": 0.4428225013348998,
      "grad_norm": 7.366025924682617,
      "learning_rate": 3.801713274611577e-06,
      "loss": 0.4701,
      "mean_token_accuracy": 0.8494726955890656,
      "num_tokens": 171865251.0,
      "step": 142850
    },
    {
      "entropy": 1.9125598162412643,
      "epoch": 0.44285350045994953,
      "grad_norm": 8.276859283447266,
      "learning_rate": 3.8015802140481523e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8597826391458512,
      "num_tokens": 171877204.0,
      "step": 142860
    },
    {
      "entropy": 1.9104377076029777,
      "epoch": 0.4428844995849992,
      "grad_norm": 3.7604010105133057,
      "learning_rate": 3.801447167455171e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8584300577640533,
      "num_tokens": 171889490.0,
      "step": 142870
    },
    {
      "entropy": 1.7467719167470932,
      "epoch": 0.4429154987100489,
      "grad_norm": 2.625401258468628,
      "learning_rate": 3.8013141348301903e-06,
      "loss": 0.361,
      "mean_token_accuracy": 0.8759482517838478,
      "num_tokens": 171904139.0,
      "step": 142880
    },
    {
      "entropy": 1.8149724557995797,
      "epoch": 0.4429464978350986,
      "grad_norm": 7.951505184173584,
      "learning_rate": 3.801181116170765e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8639876395463943,
      "num_tokens": 171918069.0,
      "step": 142890
    },
    {
      "entropy": 1.9812359169125557,
      "epoch": 0.4429774969601483,
      "grad_norm": 6.9604926109313965,
      "learning_rate": 3.801048111474452e-06,
      "loss": 0.4941,
      "mean_token_accuracy": 0.8483646467328072,
      "num_tokens": 171929606.0,
      "step": 142900
    },
    {
      "entropy": 1.9642615109682082,
      "epoch": 0.443008496085198,
      "grad_norm": 7.556371688842773,
      "learning_rate": 3.800915120738809e-06,
      "loss": 0.4788,
      "mean_token_accuracy": 0.8475595816969872,
      "num_tokens": 171941190.0,
      "step": 142910
    },
    {
      "entropy": 1.8470994770526885,
      "epoch": 0.4430394952102477,
      "grad_norm": 7.835377216339111,
      "learning_rate": 3.8007821439613928e-06,
      "loss": 0.3891,
      "mean_token_accuracy": 0.8662957057356835,
      "num_tokens": 171954065.0,
      "step": 142920
    },
    {
      "entropy": 1.8816244021058082,
      "epoch": 0.4430704943352974,
      "grad_norm": 3.198092222213745,
      "learning_rate": 3.8006491811397643e-06,
      "loss": 0.4295,
      "mean_token_accuracy": 0.8502931743860245,
      "num_tokens": 171966243.0,
      "step": 142930
    },
    {
      "entropy": 1.8951477900147438,
      "epoch": 0.4431014934603471,
      "grad_norm": 6.3186750411987305,
      "learning_rate": 3.80051623227148e-06,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8593979939818382,
      "num_tokens": 171979052.0,
      "step": 142940
    },
    {
      "entropy": 1.939696778357029,
      "epoch": 0.4431324925853968,
      "grad_norm": 9.299203872680664,
      "learning_rate": 3.8003832973541e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8494555294513703,
      "num_tokens": 171990440.0,
      "step": 142950
    },
    {
      "entropy": 1.8888702988624573,
      "epoch": 0.4431634917104465,
      "grad_norm": 3.8916590213775635,
      "learning_rate": 3.800250376385186e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8480198219418525,
      "num_tokens": 172002952.0,
      "step": 142960
    },
    {
      "entropy": 1.9692298248410225,
      "epoch": 0.44319449083549617,
      "grad_norm": 7.702281475067139,
      "learning_rate": 3.8001174693622976e-06,
      "loss": 0.4755,
      "mean_token_accuracy": 0.8490083158016205,
      "num_tokens": 172013634.0,
      "step": 142970
    },
    {
      "entropy": 1.796003994345665,
      "epoch": 0.44322548996054584,
      "grad_norm": 2.3298850059509277,
      "learning_rate": 3.7999845762829975e-06,
      "loss": 0.3878,
      "mean_token_accuracy": 0.8645777225494384,
      "num_tokens": 172027808.0,
      "step": 142980
    },
    {
      "entropy": 1.8192441150546075,
      "epoch": 0.44325648908559556,
      "grad_norm": 8.177521705627441,
      "learning_rate": 3.7998516971448463e-06,
      "loss": 0.4098,
      "mean_token_accuracy": 0.8649537086486816,
      "num_tokens": 172041098.0,
      "step": 142990
    },
    {
      "entropy": 1.7670658022165298,
      "epoch": 0.44328748821064523,
      "grad_norm": 3.8939199447631836,
      "learning_rate": 3.7997188319454082e-06,
      "loss": 0.333,
      "mean_token_accuracy": 0.8773150816559792,
      "num_tokens": 172054398.0,
      "step": 143000
    },
    {
      "entropy": 1.839756967127323,
      "epoch": 0.44331848733569496,
      "grad_norm": 8.327499389648438,
      "learning_rate": 3.7995859806822448e-06,
      "loss": 0.4289,
      "mean_token_accuracy": 0.8523253813385964,
      "num_tokens": 172067290.0,
      "step": 143010
    },
    {
      "entropy": 1.9034264922142028,
      "epoch": 0.4433494864607446,
      "grad_norm": 7.8952531814575195,
      "learning_rate": 3.7994531433529215e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.852616871893406,
      "num_tokens": 172079439.0,
      "step": 143020
    },
    {
      "entropy": 1.9509797424077988,
      "epoch": 0.44338048558579435,
      "grad_norm": 8.539716720581055,
      "learning_rate": 3.7993203199550016e-06,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8532881319522858,
      "num_tokens": 172090400.0,
      "step": 143030
    },
    {
      "entropy": 1.8199230879545212,
      "epoch": 0.443411484710844,
      "grad_norm": 7.303118705749512,
      "learning_rate": 3.7991875104860506e-06,
      "loss": 0.4115,
      "mean_token_accuracy": 0.8639900296926498,
      "num_tokens": 172103631.0,
      "step": 143040
    },
    {
      "entropy": 1.9047269806265832,
      "epoch": 0.44344248383589374,
      "grad_norm": 3.573137044906616,
      "learning_rate": 3.799054714943634e-06,
      "loss": 0.4069,
      "mean_token_accuracy": 0.8529732450842857,
      "num_tokens": 172115897.0,
      "step": 143050
    },
    {
      "entropy": 1.927889946103096,
      "epoch": 0.4434734829609434,
      "grad_norm": 6.2143635749816895,
      "learning_rate": 3.798921933325319e-06,
      "loss": 0.5202,
      "mean_token_accuracy": 0.845208078622818,
      "num_tokens": 172127650.0,
      "step": 143060
    },
    {
      "entropy": 1.8747559115290642,
      "epoch": 0.44350448208599313,
      "grad_norm": 8.309805870056152,
      "learning_rate": 3.798789165628671e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.849570831656456,
      "num_tokens": 172139526.0,
      "step": 143070
    },
    {
      "entropy": 1.8134203180670738,
      "epoch": 0.4435354812110428,
      "grad_norm": 10.744686126708984,
      "learning_rate": 3.7986564118512593e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8602747023105621,
      "num_tokens": 172152170.0,
      "step": 143080
    },
    {
      "entropy": 1.920543058216572,
      "epoch": 0.44356648033609253,
      "grad_norm": 8.824508666992188,
      "learning_rate": 3.79852367199065e-06,
      "loss": 0.4664,
      "mean_token_accuracy": 0.8488981783390045,
      "num_tokens": 172163758.0,
      "step": 143090
    },
    {
      "entropy": 1.95814551115036,
      "epoch": 0.4435974794611422,
      "grad_norm": 5.424408912658691,
      "learning_rate": 3.798390946044413e-06,
      "loss": 0.4833,
      "mean_token_accuracy": 0.8507681146264077,
      "num_tokens": 172175098.0,
      "step": 143100
    },
    {
      "entropy": 1.9256742283701898,
      "epoch": 0.4436284785861919,
      "grad_norm": 8.446990966796875,
      "learning_rate": 3.7982582340101166e-06,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8471813514828682,
      "num_tokens": 172186173.0,
      "step": 143110
    },
    {
      "entropy": 1.914665700495243,
      "epoch": 0.4436594777112416,
      "grad_norm": 4.053114414215088,
      "learning_rate": 3.7981255358853308e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8550711467862129,
      "num_tokens": 172198247.0,
      "step": 143120
    },
    {
      "entropy": 1.9240732803940772,
      "epoch": 0.4436904768362913,
      "grad_norm": 6.92840576171875,
      "learning_rate": 3.7979928516676266e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8500184774398803,
      "num_tokens": 172210111.0,
      "step": 143130
    },
    {
      "entropy": 1.9857992932200432,
      "epoch": 0.443721475961341,
      "grad_norm": 9.123396873474121,
      "learning_rate": 3.7978601813545756e-06,
      "loss": 0.5182,
      "mean_token_accuracy": 0.8403029024600983,
      "num_tokens": 172221785.0,
      "step": 143140
    },
    {
      "entropy": 1.88204335719347,
      "epoch": 0.4437524750863907,
      "grad_norm": 4.223636627197266,
      "learning_rate": 3.797727524943748e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8584297001361847,
      "num_tokens": 172234848.0,
      "step": 143150
    },
    {
      "entropy": 1.8985393255949021,
      "epoch": 0.4437834742114404,
      "grad_norm": 10.910396575927734,
      "learning_rate": 3.797594882432716e-06,
      "loss": 0.4021,
      "mean_token_accuracy": 0.856754156947136,
      "num_tokens": 172247385.0,
      "step": 143160
    },
    {
      "entropy": 1.9117623910307884,
      "epoch": 0.4438144733364901,
      "grad_norm": 7.727200508117676,
      "learning_rate": 3.7974622538190527e-06,
      "loss": 0.4536,
      "mean_token_accuracy": 0.845762537419796,
      "num_tokens": 172260013.0,
      "step": 143170
    },
    {
      "entropy": 1.9631491884589196,
      "epoch": 0.44384547246153977,
      "grad_norm": 8.514778137207031,
      "learning_rate": 3.7973296391003324e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8480650544166565,
      "num_tokens": 172271771.0,
      "step": 143180
    },
    {
      "entropy": 1.8932307675480842,
      "epoch": 0.4438764715865895,
      "grad_norm": 8.17857551574707,
      "learning_rate": 3.797197038274128e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8522116959095001,
      "num_tokens": 172284879.0,
      "step": 143190
    },
    {
      "entropy": 2.053502270579338,
      "epoch": 0.44390747071163916,
      "grad_norm": 8.686943054199219,
      "learning_rate": 3.7970644513380144e-06,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8443288967013359,
      "num_tokens": 172295803.0,
      "step": 143200
    },
    {
      "entropy": 2.0258347660303118,
      "epoch": 0.4439384698366889,
      "grad_norm": 8.006467819213867,
      "learning_rate": 3.7969318782895674e-06,
      "loss": 0.536,
      "mean_token_accuracy": 0.8328389957547188,
      "num_tokens": 172306930.0,
      "step": 143210
    },
    {
      "entropy": 1.9896105587482453,
      "epoch": 0.44396946896173856,
      "grad_norm": 3.218825340270996,
      "learning_rate": 3.7967993191263625e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8439623191952705,
      "num_tokens": 172318217.0,
      "step": 143220
    },
    {
      "entropy": 1.9521566152572631,
      "epoch": 0.4440004680867882,
      "grad_norm": 7.5543389320373535,
      "learning_rate": 3.7966667738459746e-06,
      "loss": 0.4407,
      "mean_token_accuracy": 0.8609140679240227,
      "num_tokens": 172330536.0,
      "step": 143230
    },
    {
      "entropy": 1.9294056311249732,
      "epoch": 0.44403146721183795,
      "grad_norm": 9.052746772766113,
      "learning_rate": 3.7965342424459822e-06,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8596820458769798,
      "num_tokens": 172341999.0,
      "step": 143240
    },
    {
      "entropy": 1.869072140753269,
      "epoch": 0.4440624663368876,
      "grad_norm": 7.041997909545898,
      "learning_rate": 3.7964017249239623e-06,
      "loss": 0.4067,
      "mean_token_accuracy": 0.8588750317692757,
      "num_tokens": 172355292.0,
      "step": 143250
    },
    {
      "entropy": 1.9508719503879548,
      "epoch": 0.44409346546193734,
      "grad_norm": 4.241085529327393,
      "learning_rate": 3.796269221277493e-06,
      "loss": 0.4702,
      "mean_token_accuracy": 0.8481484889984131,
      "num_tokens": 172367036.0,
      "step": 143260
    },
    {
      "entropy": 1.9314745679497718,
      "epoch": 0.444124464586987,
      "grad_norm": 4.061893939971924,
      "learning_rate": 3.7961367315041535e-06,
      "loss": 0.425,
      "mean_token_accuracy": 0.8549939930438996,
      "num_tokens": 172379192.0,
      "step": 143270
    },
    {
      "entropy": 1.9073988541960716,
      "epoch": 0.44415546371203674,
      "grad_norm": 7.23845100402832,
      "learning_rate": 3.7960042556015226e-06,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8475894257426262,
      "num_tokens": 172391173.0,
      "step": 143280
    },
    {
      "entropy": 1.9345348447561264,
      "epoch": 0.4441864628370864,
      "grad_norm": 8.21992015838623,
      "learning_rate": 3.79587179356718e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8488015040755272,
      "num_tokens": 172403416.0,
      "step": 143290
    },
    {
      "entropy": 1.9360866978764535,
      "epoch": 0.44421746196213613,
      "grad_norm": 6.462368011474609,
      "learning_rate": 3.7957393453987075e-06,
      "loss": 0.4593,
      "mean_token_accuracy": 0.8466382816433906,
      "num_tokens": 172415684.0,
      "step": 143300
    },
    {
      "entropy": 1.9160262137651443,
      "epoch": 0.4442484610871858,
      "grad_norm": 6.9671454429626465,
      "learning_rate": 3.795606911093684e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8536192387342453,
      "num_tokens": 172427475.0,
      "step": 143310
    },
    {
      "entropy": 1.9292744249105453,
      "epoch": 0.4442794602122355,
      "grad_norm": 3.4620540142059326,
      "learning_rate": 3.7954744906496932e-06,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8594009026885032,
      "num_tokens": 172439631.0,
      "step": 143320
    },
    {
      "entropy": 1.9033573985099792,
      "epoch": 0.4443104593372852,
      "grad_norm": 8.030845642089844,
      "learning_rate": 3.795342084064316e-06,
      "loss": 0.4125,
      "mean_token_accuracy": 0.86111601293087,
      "num_tokens": 172451878.0,
      "step": 143330
    },
    {
      "entropy": 1.851330216228962,
      "epoch": 0.4443414584623349,
      "grad_norm": 8.43782901763916,
      "learning_rate": 3.795209691335136e-06,
      "loss": 0.3983,
      "mean_token_accuracy": 0.8640111222863197,
      "num_tokens": 172465066.0,
      "step": 143340
    },
    {
      "entropy": 1.991729763150215,
      "epoch": 0.4443724575873846,
      "grad_norm": 7.826473236083984,
      "learning_rate": 3.7950773124597375e-06,
      "loss": 0.488,
      "mean_token_accuracy": 0.8467871636152268,
      "num_tokens": 172476252.0,
      "step": 143350
    },
    {
      "entropy": 1.9685329020023346,
      "epoch": 0.4444034567124343,
      "grad_norm": 8.118627548217773,
      "learning_rate": 3.794944947435702e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8508369252085686,
      "num_tokens": 172487279.0,
      "step": 143360
    },
    {
      "entropy": 2.045986759662628,
      "epoch": 0.444434455837484,
      "grad_norm": 8.419778823852539,
      "learning_rate": 3.794812596260616e-06,
      "loss": 0.4839,
      "mean_token_accuracy": 0.8484291791915893,
      "num_tokens": 172498181.0,
      "step": 143370
    },
    {
      "entropy": 1.9107683286070825,
      "epoch": 0.4444654549625337,
      "grad_norm": 8.121953964233398,
      "learning_rate": 3.794680258932064e-06,
      "loss": 0.4352,
      "mean_token_accuracy": 0.8511162519454956,
      "num_tokens": 172510612.0,
      "step": 143380
    },
    {
      "entropy": 1.9096698313951492,
      "epoch": 0.44449645408758337,
      "grad_norm": 9.873578071594238,
      "learning_rate": 3.7945479354476334e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.867861407995224,
      "num_tokens": 172522684.0,
      "step": 143390
    },
    {
      "entropy": 1.9415227890014648,
      "epoch": 0.4445274532126331,
      "grad_norm": 9.934027671813965,
      "learning_rate": 3.794415625804908e-06,
      "loss": 0.477,
      "mean_token_accuracy": 0.8525919511914253,
      "num_tokens": 172533993.0,
      "step": 143400
    },
    {
      "entropy": 1.9405535489320755,
      "epoch": 0.44455845233768276,
      "grad_norm": 7.328090667724609,
      "learning_rate": 3.794283330001477e-06,
      "loss": 0.4226,
      "mean_token_accuracy": 0.8587682485580445,
      "num_tokens": 172545914.0,
      "step": 143410
    },
    {
      "entropy": 1.984012272953987,
      "epoch": 0.4445894514627325,
      "grad_norm": 8.643956184387207,
      "learning_rate": 3.7941510480349265e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8469142064452171,
      "num_tokens": 172556997.0,
      "step": 143420
    },
    {
      "entropy": 1.9300741687417031,
      "epoch": 0.44462045058778216,
      "grad_norm": 4.657692909240723,
      "learning_rate": 3.794018779902845e-06,
      "loss": 0.4216,
      "mean_token_accuracy": 0.860441392660141,
      "num_tokens": 172568929.0,
      "step": 143430
    },
    {
      "entropy": 1.8880173966288567,
      "epoch": 0.4446514497128319,
      "grad_norm": 8.421919822692871,
      "learning_rate": 3.793886525602822e-06,
      "loss": 0.395,
      "mean_token_accuracy": 0.8604597702622414,
      "num_tokens": 172580791.0,
      "step": 143440
    },
    {
      "entropy": 1.933186987042427,
      "epoch": 0.44468244883788155,
      "grad_norm": 3.5404715538024902,
      "learning_rate": 3.793754285132446e-06,
      "loss": 0.4859,
      "mean_token_accuracy": 0.8485592529177666,
      "num_tokens": 172593075.0,
      "step": 143450
    },
    {
      "entropy": 1.950406338274479,
      "epoch": 0.4447134479629312,
      "grad_norm": 8.130172729492188,
      "learning_rate": 3.7936220584893074e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8546596810221672,
      "num_tokens": 172604452.0,
      "step": 143460
    },
    {
      "entropy": 1.9698896750807762,
      "epoch": 0.44474444708798094,
      "grad_norm": 8.126727104187012,
      "learning_rate": 3.7934898456709963e-06,
      "loss": 0.4745,
      "mean_token_accuracy": 0.8500417619943619,
      "num_tokens": 172616103.0,
      "step": 143470
    },
    {
      "entropy": 1.8846001833677293,
      "epoch": 0.4447754462130306,
      "grad_norm": 7.103678226470947,
      "learning_rate": 3.793357646675104e-06,
      "loss": 0.4099,
      "mean_token_accuracy": 0.8594605028629303,
      "num_tokens": 172627939.0,
      "step": 143480
    },
    {
      "entropy": 1.9624834671616553,
      "epoch": 0.44480644533808034,
      "grad_norm": 7.368971824645996,
      "learning_rate": 3.7932254614992225e-06,
      "loss": 0.4464,
      "mean_token_accuracy": 0.853089140355587,
      "num_tokens": 172639602.0,
      "step": 143490
    },
    {
      "entropy": 2.0040226548910143,
      "epoch": 0.44483744446313,
      "grad_norm": 8.010761260986328,
      "learning_rate": 3.7930932901409443e-06,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8468669816851616,
      "num_tokens": 172650585.0,
      "step": 143500
    },
    {
      "entropy": 1.9335386529564857,
      "epoch": 0.44486844358817973,
      "grad_norm": 8.675997734069824,
      "learning_rate": 3.7929611325978612e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8535260155797004,
      "num_tokens": 172662166.0,
      "step": 143510
    },
    {
      "entropy": 1.975813153386116,
      "epoch": 0.4448994427132294,
      "grad_norm": 9.182839393615723,
      "learning_rate": 3.7928289888675668e-06,
      "loss": 0.4709,
      "mean_token_accuracy": 0.855238126218319,
      "num_tokens": 172673577.0,
      "step": 143520
    },
    {
      "entropy": 1.982006560266018,
      "epoch": 0.4449304418382791,
      "grad_norm": 8.914770126342773,
      "learning_rate": 3.7926968589476558e-06,
      "loss": 0.4923,
      "mean_token_accuracy": 0.8514761239290237,
      "num_tokens": 172685618.0,
      "step": 143530
    },
    {
      "entropy": 2.004455064237118,
      "epoch": 0.4449614409633288,
      "grad_norm": 8.590579986572266,
      "learning_rate": 3.7925647428357226e-06,
      "loss": 0.4645,
      "mean_token_accuracy": 0.8540130868554116,
      "num_tokens": 172697021.0,
      "step": 143540
    },
    {
      "entropy": 2.02477196007967,
      "epoch": 0.4449924400883785,
      "grad_norm": 8.24437427520752,
      "learning_rate": 3.7924326405293627e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8468568831682205,
      "num_tokens": 172708199.0,
      "step": 143550
    },
    {
      "entropy": 1.9295995250344276,
      "epoch": 0.4450234392134282,
      "grad_norm": 4.574173927307129,
      "learning_rate": 3.7923005520261708e-06,
      "loss": 0.4307,
      "mean_token_accuracy": 0.8547417059540748,
      "num_tokens": 172720296.0,
      "step": 143560
    },
    {
      "entropy": 1.7906879603862762,
      "epoch": 0.4450544383384779,
      "grad_norm": 8.181215286254883,
      "learning_rate": 3.792168477323745e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.8623343601822853,
      "num_tokens": 172734098.0,
      "step": 143570
    },
    {
      "entropy": 1.7868395507335664,
      "epoch": 0.4450854374635276,
      "grad_norm": 4.160261631011963,
      "learning_rate": 3.7920364164196805e-06,
      "loss": 0.3373,
      "mean_token_accuracy": 0.8650955274701119,
      "num_tokens": 172748197.0,
      "step": 143580
    },
    {
      "entropy": 1.906904463469982,
      "epoch": 0.4451164365885773,
      "grad_norm": 7.085610389709473,
      "learning_rate": 3.7919043693115757e-06,
      "loss": 0.425,
      "mean_token_accuracy": 0.857354860007763,
      "num_tokens": 172760940.0,
      "step": 143590
    },
    {
      "entropy": 1.8628742694854736,
      "epoch": 0.44514743571362697,
      "grad_norm": 7.324219226837158,
      "learning_rate": 3.7917723359970293e-06,
      "loss": 0.3814,
      "mean_token_accuracy": 0.8632989749312401,
      "num_tokens": 172773595.0,
      "step": 143600
    },
    {
      "entropy": 1.9510155633091926,
      "epoch": 0.4451784348386767,
      "grad_norm": 7.612024784088135,
      "learning_rate": 3.791640316473639e-06,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8439678400754929,
      "num_tokens": 172784596.0,
      "step": 143610
    },
    {
      "entropy": 1.9409842744469643,
      "epoch": 0.44520943396372636,
      "grad_norm": 7.95268440246582,
      "learning_rate": 3.791508310739005e-06,
      "loss": 0.4449,
      "mean_token_accuracy": 0.8481158643960953,
      "num_tokens": 172796663.0,
      "step": 143620
    },
    {
      "entropy": 1.8901426151394844,
      "epoch": 0.4452404330887761,
      "grad_norm": 7.622109889984131,
      "learning_rate": 3.7913763187907265e-06,
      "loss": 0.4143,
      "mean_token_accuracy": 0.8628189340233803,
      "num_tokens": 172809208.0,
      "step": 143630
    },
    {
      "entropy": 1.9054123714566231,
      "epoch": 0.44527143221382576,
      "grad_norm": 8.614324569702148,
      "learning_rate": 3.791244340626404e-06,
      "loss": 0.4353,
      "mean_token_accuracy": 0.8581098973751068,
      "num_tokens": 172821386.0,
      "step": 143640
    },
    {
      "entropy": 1.8443352833390236,
      "epoch": 0.4453024313388755,
      "grad_norm": 4.017133712768555,
      "learning_rate": 3.7911123762436393e-06,
      "loss": 0.4023,
      "mean_token_accuracy": 0.8598392441868782,
      "num_tokens": 172834753.0,
      "step": 143650
    },
    {
      "entropy": 1.9353301167488097,
      "epoch": 0.44533343046392515,
      "grad_norm": 4.68195104598999,
      "learning_rate": 3.7909804256400327e-06,
      "loss": 0.444,
      "mean_token_accuracy": 0.8590727686882019,
      "num_tokens": 172846687.0,
      "step": 143660
    },
    {
      "entropy": 2.013177511096001,
      "epoch": 0.4453644295889749,
      "grad_norm": 8.711843490600586,
      "learning_rate": 3.7908484888131875e-06,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8534877151250839,
      "num_tokens": 172857233.0,
      "step": 143670
    },
    {
      "entropy": 1.9266994565725326,
      "epoch": 0.44539542871402454,
      "grad_norm": 8.333199501037598,
      "learning_rate": 3.7907165657607066e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8665392875671387,
      "num_tokens": 172869433.0,
      "step": 143680
    },
    {
      "entropy": 1.9714792117476463,
      "epoch": 0.44542642783907427,
      "grad_norm": 8.7294340133667,
      "learning_rate": 3.7905846564801923e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8482206985354424,
      "num_tokens": 172880658.0,
      "step": 143690
    },
    {
      "entropy": 1.8752370357513428,
      "epoch": 0.44545742696412394,
      "grad_norm": 8.233158111572266,
      "learning_rate": 3.7904527609692506e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.8617415338754654,
      "num_tokens": 172892973.0,
      "step": 143700
    },
    {
      "entropy": 2.0074197858572007,
      "epoch": 0.4454884260891736,
      "grad_norm": 7.99542760848999,
      "learning_rate": 3.7903208792254844e-06,
      "loss": 0.4904,
      "mean_token_accuracy": 0.8521378964185715,
      "num_tokens": 172904130.0,
      "step": 143710
    },
    {
      "entropy": 1.99031350761652,
      "epoch": 0.44551942521422333,
      "grad_norm": 8.73692798614502,
      "learning_rate": 3.7901890112464982e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8600781843066215,
      "num_tokens": 172915284.0,
      "step": 143720
    },
    {
      "entropy": 1.9789996802806855,
      "epoch": 0.445550424339273,
      "grad_norm": 9.458690643310547,
      "learning_rate": 3.7900571570299e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8530536726117134,
      "num_tokens": 172926970.0,
      "step": 143730
    },
    {
      "entropy": 1.834419848024845,
      "epoch": 0.4455814234643227,
      "grad_norm": 4.028777599334717,
      "learning_rate": 3.789925316573294e-06,
      "loss": 0.4093,
      "mean_token_accuracy": 0.8613479465246201,
      "num_tokens": 172939899.0,
      "step": 143740
    },
    {
      "entropy": 1.945051771402359,
      "epoch": 0.4456124225893724,
      "grad_norm": 4.0096306800842285,
      "learning_rate": 3.7897934898742885e-06,
      "loss": 0.4787,
      "mean_token_accuracy": 0.8497995778918266,
      "num_tokens": 172951805.0,
      "step": 143750
    },
    {
      "entropy": 1.8734006211161613,
      "epoch": 0.4456434217144221,
      "grad_norm": 3.48857045173645,
      "learning_rate": 3.7896616769304905e-06,
      "loss": 0.3735,
      "mean_token_accuracy": 0.8659715816378594,
      "num_tokens": 172964673.0,
      "step": 143760
    },
    {
      "entropy": 1.9327120184898376,
      "epoch": 0.4456744208394718,
      "grad_norm": 8.14692497253418,
      "learning_rate": 3.7895298777395078e-06,
      "loss": 0.46,
      "mean_token_accuracy": 0.8533953115344047,
      "num_tokens": 172975682.0,
      "step": 143770
    },
    {
      "entropy": 1.9580781385302544,
      "epoch": 0.4457054199645215,
      "grad_norm": 8.688072204589844,
      "learning_rate": 3.7893980922989496e-06,
      "loss": 0.4946,
      "mean_token_accuracy": 0.839683572947979,
      "num_tokens": 172987936.0,
      "step": 143780
    },
    {
      "entropy": 1.908944171667099,
      "epoch": 0.4457364190895712,
      "grad_norm": 3.971160888671875,
      "learning_rate": 3.7892663206064244e-06,
      "loss": 0.4127,
      "mean_token_accuracy": 0.8549481183290482,
      "num_tokens": 173000930.0,
      "step": 143790
    },
    {
      "entropy": 2.0028712913393973,
      "epoch": 0.4457674182146209,
      "grad_norm": 8.951005935668945,
      "learning_rate": 3.789134562659543e-06,
      "loss": 0.4679,
      "mean_token_accuracy": 0.845504654943943,
      "num_tokens": 173012916.0,
      "step": 143800
    },
    {
      "entropy": 1.9837856590747833,
      "epoch": 0.44579841733967057,
      "grad_norm": 7.293043613433838,
      "learning_rate": 3.7890028184559154e-06,
      "loss": 0.4261,
      "mean_token_accuracy": 0.8591191872954369,
      "num_tokens": 173024193.0,
      "step": 143810
    },
    {
      "entropy": 1.9127214059233666,
      "epoch": 0.4458294164647203,
      "grad_norm": 3.8465218544006348,
      "learning_rate": 3.7888710879931517e-06,
      "loss": 0.4417,
      "mean_token_accuracy": 0.8534463986754417,
      "num_tokens": 173035733.0,
      "step": 143820
    },
    {
      "entropy": 1.949730084836483,
      "epoch": 0.44586041558976997,
      "grad_norm": 10.93399429321289,
      "learning_rate": 3.788739371268865e-06,
      "loss": 0.4748,
      "mean_token_accuracy": 0.8493082284927368,
      "num_tokens": 173047215.0,
      "step": 143830
    },
    {
      "entropy": 2.0025800704956054,
      "epoch": 0.4458914147148197,
      "grad_norm": 8.645318984985352,
      "learning_rate": 3.788607668280666e-06,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8457403063774109,
      "num_tokens": 173058282.0,
      "step": 143840
    },
    {
      "entropy": 1.910311257839203,
      "epoch": 0.44592241383986936,
      "grad_norm": 3.202796220779419,
      "learning_rate": 3.7884759790261683e-06,
      "loss": 0.4434,
      "mean_token_accuracy": 0.8490109801292419,
      "num_tokens": 173070834.0,
      "step": 143850
    },
    {
      "entropy": 1.9238232627511025,
      "epoch": 0.4459534129649191,
      "grad_norm": 8.433602333068848,
      "learning_rate": 3.7883443035029844e-06,
      "loss": 0.4331,
      "mean_token_accuracy": 0.8548846036195755,
      "num_tokens": 173083400.0,
      "step": 143860
    },
    {
      "entropy": 1.961086443066597,
      "epoch": 0.44598441208996875,
      "grad_norm": 3.7970497608184814,
      "learning_rate": 3.7882126417087294e-06,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8421978339552879,
      "num_tokens": 173094716.0,
      "step": 143870
    },
    {
      "entropy": 1.9548719599843025,
      "epoch": 0.4460154112150185,
      "grad_norm": 8.201971054077148,
      "learning_rate": 3.788080993641017e-06,
      "loss": 0.5078,
      "mean_token_accuracy": 0.8327627673745155,
      "num_tokens": 173107004.0,
      "step": 143880
    },
    {
      "entropy": 1.8781091898679734,
      "epoch": 0.44604641034006814,
      "grad_norm": 3.5687875747680664,
      "learning_rate": 3.7879493592974612e-06,
      "loss": 0.4453,
      "mean_token_accuracy": 0.8458937138319016,
      "num_tokens": 173119649.0,
      "step": 143890
    },
    {
      "entropy": 1.9691094398498534,
      "epoch": 0.44607740946511787,
      "grad_norm": 10.030198097229004,
      "learning_rate": 3.7878177386756796e-06,
      "loss": 0.516,
      "mean_token_accuracy": 0.8466410353779793,
      "num_tokens": 173131103.0,
      "step": 143900
    },
    {
      "entropy": 2.0002141326665877,
      "epoch": 0.44610840859016754,
      "grad_norm": 8.030385971069336,
      "learning_rate": 3.7876861317732877e-06,
      "loss": 0.5004,
      "mean_token_accuracy": 0.8476643145084382,
      "num_tokens": 173142013.0,
      "step": 143910
    },
    {
      "entropy": 1.9814749881625175,
      "epoch": 0.44613940771521726,
      "grad_norm": 9.209390640258789,
      "learning_rate": 3.787554538587902e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8513798832893371,
      "num_tokens": 173153145.0,
      "step": 143920
    },
    {
      "entropy": 1.953411616384983,
      "epoch": 0.44617040684026693,
      "grad_norm": 10.11378288269043,
      "learning_rate": 3.7874229591171395e-06,
      "loss": 0.467,
      "mean_token_accuracy": 0.845984798669815,
      "num_tokens": 173165304.0,
      "step": 143930
    },
    {
      "entropy": 1.8993015423417092,
      "epoch": 0.44620140596531666,
      "grad_norm": 7.429069995880127,
      "learning_rate": 3.787291393358619e-06,
      "loss": 0.4043,
      "mean_token_accuracy": 0.8644404307007789,
      "num_tokens": 173177675.0,
      "step": 143940
    },
    {
      "entropy": 1.8486544996500016,
      "epoch": 0.4462324050903663,
      "grad_norm": 8.14201545715332,
      "learning_rate": 3.7871598413099593e-06,
      "loss": 0.379,
      "mean_token_accuracy": 0.8645230799913406,
      "num_tokens": 173191122.0,
      "step": 143950
    },
    {
      "entropy": 2.023305447399616,
      "epoch": 0.446263404215416,
      "grad_norm": 10.556729316711426,
      "learning_rate": 3.7870283029687777e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.847667895257473,
      "num_tokens": 173202142.0,
      "step": 143960
    },
    {
      "entropy": 1.910950092971325,
      "epoch": 0.4462944033404657,
      "grad_norm": 7.93691349029541,
      "learning_rate": 3.7868967783326956e-06,
      "loss": 0.4291,
      "mean_token_accuracy": 0.856875829398632,
      "num_tokens": 173214829.0,
      "step": 143970
    },
    {
      "entropy": 1.9166125014424324,
      "epoch": 0.4463254024655154,
      "grad_norm": 9.42168140411377,
      "learning_rate": 3.7867652673993318e-06,
      "loss": 0.4674,
      "mean_token_accuracy": 0.8535972073674202,
      "num_tokens": 173226844.0,
      "step": 143980
    },
    {
      "entropy": 1.8531441867351532,
      "epoch": 0.4463564015905651,
      "grad_norm": 4.800891399383545,
      "learning_rate": 3.7866337701663088e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8521735593676567,
      "num_tokens": 173239753.0,
      "step": 143990
    },
    {
      "entropy": 1.881640262901783,
      "epoch": 0.4463874007156148,
      "grad_norm": 9.612931251525879,
      "learning_rate": 3.7865022866312468e-06,
      "loss": 0.4001,
      "mean_token_accuracy": 0.8631109833717346,
      "num_tokens": 173252177.0,
      "step": 144000
    },
    {
      "entropy": 1.8542085379362105,
      "epoch": 0.4464183998406645,
      "grad_norm": 5.085047245025635,
      "learning_rate": 3.7863708167917686e-06,
      "loss": 0.3643,
      "mean_token_accuracy": 0.8695595130324364,
      "num_tokens": 173264694.0,
      "step": 144010
    },
    {
      "entropy": 2.0253135934472084,
      "epoch": 0.4464493989657142,
      "grad_norm": 9.341069221496582,
      "learning_rate": 3.7862393606454958e-06,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8394965454936028,
      "num_tokens": 173275727.0,
      "step": 144020
    },
    {
      "entropy": 1.8591209024190902,
      "epoch": 0.4464803980907639,
      "grad_norm": 8.249305725097656,
      "learning_rate": 3.786107918190052e-06,
      "loss": 0.4193,
      "mean_token_accuracy": 0.8542803794145584,
      "num_tokens": 173288795.0,
      "step": 144030
    },
    {
      "entropy": 1.9186673119664193,
      "epoch": 0.44651139721581357,
      "grad_norm": 9.003576278686523,
      "learning_rate": 3.785976489423061e-06,
      "loss": 0.4394,
      "mean_token_accuracy": 0.8566552370786666,
      "num_tokens": 173300153.0,
      "step": 144040
    },
    {
      "entropy": 1.9807184368371964,
      "epoch": 0.4465423963408633,
      "grad_norm": 8.3126859664917,
      "learning_rate": 3.785845074342148e-06,
      "loss": 0.4589,
      "mean_token_accuracy": 0.8607570424675941,
      "num_tokens": 173311456.0,
      "step": 144050
    },
    {
      "entropy": 1.9973660945892333,
      "epoch": 0.44657339546591296,
      "grad_norm": 8.058760643005371,
      "learning_rate": 3.7857136729449356e-06,
      "loss": 0.4827,
      "mean_token_accuracy": 0.8486983701586723,
      "num_tokens": 173322684.0,
      "step": 144060
    },
    {
      "entropy": 1.9315310716629028,
      "epoch": 0.4466043945909627,
      "grad_norm": 9.327740669250488,
      "learning_rate": 3.7855822852290523e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8483046740293503,
      "num_tokens": 173334375.0,
      "step": 144070
    },
    {
      "entropy": 1.854925161600113,
      "epoch": 0.44663539371601235,
      "grad_norm": 8.95000171661377,
      "learning_rate": 3.785450911192121e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8594622611999512,
      "num_tokens": 173347587.0,
      "step": 144080
    },
    {
      "entropy": 1.9255126759409904,
      "epoch": 0.4466663928410621,
      "grad_norm": 9.044776916503906,
      "learning_rate": 3.7853195508317707e-06,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8483678221702575,
      "num_tokens": 173359548.0,
      "step": 144090
    },
    {
      "entropy": 1.9388061597943307,
      "epoch": 0.44669739196611175,
      "grad_norm": 8.007100105285645,
      "learning_rate": 3.785188204145627e-06,
      "loss": 0.4604,
      "mean_token_accuracy": 0.8521351292729378,
      "num_tokens": 173371042.0,
      "step": 144100
    },
    {
      "entropy": 1.9809336930513382,
      "epoch": 0.44672839109116147,
      "grad_norm": 8.529678344726562,
      "learning_rate": 3.7850568711313192e-06,
      "loss": 0.4665,
      "mean_token_accuracy": 0.8555900052189827,
      "num_tokens": 173382290.0,
      "step": 144110
    },
    {
      "entropy": 1.9656935811042786,
      "epoch": 0.44675939021621114,
      "grad_norm": 4.3543219566345215,
      "learning_rate": 3.7849255517864743e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8531507521867752,
      "num_tokens": 173393550.0,
      "step": 144120
    },
    {
      "entropy": 1.8287263602018355,
      "epoch": 0.44679038934126086,
      "grad_norm": 3.763613700866699,
      "learning_rate": 3.784794246108721e-06,
      "loss": 0.3894,
      "mean_token_accuracy": 0.859270915389061,
      "num_tokens": 173406666.0,
      "step": 144130
    },
    {
      "entropy": 1.93349888920784,
      "epoch": 0.44682138846631053,
      "grad_norm": 8.177099227905273,
      "learning_rate": 3.784662954095691e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8471749663352967,
      "num_tokens": 173418993.0,
      "step": 144140
    },
    {
      "entropy": 1.9891784444451333,
      "epoch": 0.44685238759136026,
      "grad_norm": 9.054378509521484,
      "learning_rate": 3.7845316757450114e-06,
      "loss": 0.4624,
      "mean_token_accuracy": 0.8486263379454613,
      "num_tokens": 173430436.0,
      "step": 144150
    },
    {
      "entropy": 1.9303235545754434,
      "epoch": 0.4468833867164099,
      "grad_norm": 8.891922950744629,
      "learning_rate": 3.7844004110543148e-06,
      "loss": 0.4609,
      "mean_token_accuracy": 0.8510625973343849,
      "num_tokens": 173442415.0,
      "step": 144160
    },
    {
      "entropy": 1.94441519677639,
      "epoch": 0.44691438584145965,
      "grad_norm": 8.909342765808105,
      "learning_rate": 3.7842691600212316e-06,
      "loss": 0.4587,
      "mean_token_accuracy": 0.8438147768378258,
      "num_tokens": 173454807.0,
      "step": 144170
    },
    {
      "entropy": 2.0386078268289567,
      "epoch": 0.4469453849665093,
      "grad_norm": 10.605982780456543,
      "learning_rate": 3.784137922643394e-06,
      "loss": 0.5531,
      "mean_token_accuracy": 0.8398728102445603,
      "num_tokens": 173465671.0,
      "step": 144180
    },
    {
      "entropy": 1.9403854593634606,
      "epoch": 0.44697638409155904,
      "grad_norm": 7.102710723876953,
      "learning_rate": 3.7840066989184345e-06,
      "loss": 0.4316,
      "mean_token_accuracy": 0.8576330199837685,
      "num_tokens": 173478051.0,
      "step": 144190
    },
    {
      "entropy": 1.9223571211099624,
      "epoch": 0.4470073832166087,
      "grad_norm": 7.868150234222412,
      "learning_rate": 3.7838754888439857e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8626696810126304,
      "num_tokens": 173489382.0,
      "step": 144200
    },
    {
      "entropy": 1.9799730449914932,
      "epoch": 0.4470383823416584,
      "grad_norm": 6.588014602661133,
      "learning_rate": 3.7837442924176804e-06,
      "loss": 0.5044,
      "mean_token_accuracy": 0.8521138474345207,
      "num_tokens": 173501335.0,
      "step": 144210
    },
    {
      "entropy": 1.9104075238108635,
      "epoch": 0.4470693814667081,
      "grad_norm": 8.533629417419434,
      "learning_rate": 3.783613109637155e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.8599554657936096,
      "num_tokens": 173513423.0,
      "step": 144220
    },
    {
      "entropy": 1.884183020889759,
      "epoch": 0.4471003805917578,
      "grad_norm": 2.680689573287964,
      "learning_rate": 3.7834819405000404e-06,
      "loss": 0.4075,
      "mean_token_accuracy": 0.8564264923334122,
      "num_tokens": 173525703.0,
      "step": 144230
    },
    {
      "entropy": 1.998404061794281,
      "epoch": 0.4471313797168075,
      "grad_norm": 7.732211112976074,
      "learning_rate": 3.7833507850039757e-06,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8460897743701935,
      "num_tokens": 173536387.0,
      "step": 144240
    },
    {
      "entropy": 1.888758347928524,
      "epoch": 0.44716237884185717,
      "grad_norm": 9.650225639343262,
      "learning_rate": 3.783219643146595e-06,
      "loss": 0.4148,
      "mean_token_accuracy": 0.8614796027541161,
      "num_tokens": 173548843.0,
      "step": 144250
    },
    {
      "entropy": 1.9674888044595717,
      "epoch": 0.4471933779669069,
      "grad_norm": 7.974256992340088,
      "learning_rate": 3.7830885149255346e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8619727566838264,
      "num_tokens": 173560036.0,
      "step": 144260
    },
    {
      "entropy": 1.854830276221037,
      "epoch": 0.44722437709195656,
      "grad_norm": 4.305238246917725,
      "learning_rate": 3.7829574003384317e-06,
      "loss": 0.4224,
      "mean_token_accuracy": 0.8533999145030975,
      "num_tokens": 173572981.0,
      "step": 144270
    },
    {
      "entropy": 1.8603426963090897,
      "epoch": 0.4472553762170063,
      "grad_norm": 8.774975776672363,
      "learning_rate": 3.7828262993829235e-06,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8556037560105324,
      "num_tokens": 173586083.0,
      "step": 144280
    },
    {
      "entropy": 1.8292605131864548,
      "epoch": 0.44728637534205595,
      "grad_norm": 3.801150321960449,
      "learning_rate": 3.7826952120566485e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.8639505878090858,
      "num_tokens": 173599424.0,
      "step": 144290
    },
    {
      "entropy": 1.8792633160948753,
      "epoch": 0.4473173744671057,
      "grad_norm": 4.3725409507751465,
      "learning_rate": 3.7825641383572448e-06,
      "loss": 0.4181,
      "mean_token_accuracy": 0.8518676936626435,
      "num_tokens": 173611821.0,
      "step": 144300
    },
    {
      "entropy": 1.9342765644192697,
      "epoch": 0.44734837359215535,
      "grad_norm": 9.642680168151855,
      "learning_rate": 3.7824330782823524e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8533917725086212,
      "num_tokens": 173623485.0,
      "step": 144310
    },
    {
      "entropy": 2.012094184756279,
      "epoch": 0.44737937271720507,
      "grad_norm": 7.291985988616943,
      "learning_rate": 3.7823020318296104e-06,
      "loss": 0.5282,
      "mean_token_accuracy": 0.8319395646452904,
      "num_tokens": 173634217.0,
      "step": 144320
    },
    {
      "entropy": 1.9318621635437012,
      "epoch": 0.44741037184225474,
      "grad_norm": 6.990114212036133,
      "learning_rate": 3.7821709989966605e-06,
      "loss": 0.4082,
      "mean_token_accuracy": 0.8600582510232926,
      "num_tokens": 173646108.0,
      "step": 144330
    },
    {
      "entropy": 1.9527502954006195,
      "epoch": 0.44744137096730446,
      "grad_norm": 7.637629508972168,
      "learning_rate": 3.782039979781142e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8539629653096199,
      "num_tokens": 173657100.0,
      "step": 144340
    },
    {
      "entropy": 1.969818153977394,
      "epoch": 0.44747237009235413,
      "grad_norm": 8.846341133117676,
      "learning_rate": 3.7819089741806974e-06,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8485133573412895,
      "num_tokens": 173668376.0,
      "step": 144350
    },
    {
      "entropy": 1.903004801273346,
      "epoch": 0.44750336921740386,
      "grad_norm": 6.75057315826416,
      "learning_rate": 3.781777982192969e-06,
      "loss": 0.5438,
      "mean_token_accuracy": 0.8346272155642509,
      "num_tokens": 173681042.0,
      "step": 144360
    },
    {
      "entropy": 1.8782558515667915,
      "epoch": 0.4475343683424535,
      "grad_norm": 6.844717025756836,
      "learning_rate": 3.781647003815598e-06,
      "loss": 0.4359,
      "mean_token_accuracy": 0.8552762180566787,
      "num_tokens": 173693000.0,
      "step": 144370
    },
    {
      "entropy": 1.9393670424818992,
      "epoch": 0.44756536746750325,
      "grad_norm": 9.035807609558105,
      "learning_rate": 3.7815160390462298e-06,
      "loss": 0.4454,
      "mean_token_accuracy": 0.8559998840093612,
      "num_tokens": 173704983.0,
      "step": 144380
    },
    {
      "entropy": 1.9297925919294356,
      "epoch": 0.4475963665925529,
      "grad_norm": 6.4936137199401855,
      "learning_rate": 3.7813850878825064e-06,
      "loss": 0.4642,
      "mean_token_accuracy": 0.8614618703722954,
      "num_tokens": 173716355.0,
      "step": 144390
    },
    {
      "entropy": 1.8133862063288688,
      "epoch": 0.44762736571760264,
      "grad_norm": 4.11335563659668,
      "learning_rate": 3.7812541503220733e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8672507256269455,
      "num_tokens": 173730318.0,
      "step": 144400
    },
    {
      "entropy": 1.956637793779373,
      "epoch": 0.4476583648426523,
      "grad_norm": 6.994054794311523,
      "learning_rate": 3.7811232263625753e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8508543521165848,
      "num_tokens": 173741321.0,
      "step": 144410
    },
    {
      "entropy": 1.92156320810318,
      "epoch": 0.44768936396770204,
      "grad_norm": 7.512271404266357,
      "learning_rate": 3.780992316001657e-06,
      "loss": 0.5093,
      "mean_token_accuracy": 0.845100137591362,
      "num_tokens": 173753512.0,
      "step": 144420
    },
    {
      "entropy": 1.8369450643658638,
      "epoch": 0.4477203630927517,
      "grad_norm": 8.064924240112305,
      "learning_rate": 3.7808614192369664e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8641338348388672,
      "num_tokens": 173767543.0,
      "step": 144430
    },
    {
      "entropy": 1.8546362400054932,
      "epoch": 0.44775136221780143,
      "grad_norm": 9.19071102142334,
      "learning_rate": 3.7807305360661485e-06,
      "loss": 0.3772,
      "mean_token_accuracy": 0.8610058188438415,
      "num_tokens": 173780631.0,
      "step": 144440
    },
    {
      "entropy": 1.9356670066714288,
      "epoch": 0.4477823613428511,
      "grad_norm": 11.3148832321167,
      "learning_rate": 3.780599666486851e-06,
      "loss": 0.4627,
      "mean_token_accuracy": 0.852401140332222,
      "num_tokens": 173792247.0,
      "step": 144450
    },
    {
      "entropy": 1.8100459277629852,
      "epoch": 0.44781336046790077,
      "grad_norm": 3.687837839126587,
      "learning_rate": 3.780468810496722e-06,
      "loss": 0.3704,
      "mean_token_accuracy": 0.8491495341062546,
      "num_tokens": 173806105.0,
      "step": 144460
    },
    {
      "entropy": 1.9653054475784302,
      "epoch": 0.4478443595929505,
      "grad_norm": 8.540299415588379,
      "learning_rate": 3.7803379680934092e-06,
      "loss": 0.4749,
      "mean_token_accuracy": 0.8488686487078667,
      "num_tokens": 173818024.0,
      "step": 144470
    },
    {
      "entropy": 1.9154264122247695,
      "epoch": 0.44787535871800016,
      "grad_norm": 8.398823738098145,
      "learning_rate": 3.7802071392745628e-06,
      "loss": 0.4439,
      "mean_token_accuracy": 0.8501715928316116,
      "num_tokens": 173830403.0,
      "step": 144480
    },
    {
      "entropy": 1.9817367240786552,
      "epoch": 0.4479063578430499,
      "grad_norm": 8.067160606384277,
      "learning_rate": 3.7800763240378307e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8624659746885299,
      "num_tokens": 173841447.0,
      "step": 144490
    },
    {
      "entropy": 1.826142853498459,
      "epoch": 0.44793735696809955,
      "grad_norm": 8.450629234313965,
      "learning_rate": 3.7799455223808647e-06,
      "loss": 0.3764,
      "mean_token_accuracy": 0.8624396830797195,
      "num_tokens": 173854816.0,
      "step": 144500
    },
    {
      "entropy": 1.9071953654289246,
      "epoch": 0.4479683560931493,
      "grad_norm": 8.542773246765137,
      "learning_rate": 3.7798147343013134e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8587081953883171,
      "num_tokens": 173866813.0,
      "step": 144510
    },
    {
      "entropy": 1.9302286952733994,
      "epoch": 0.44799935521819895,
      "grad_norm": 8.922464370727539,
      "learning_rate": 3.7796839597968305e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8490835353732109,
      "num_tokens": 173878864.0,
      "step": 144520
    },
    {
      "entropy": 1.841838812828064,
      "epoch": 0.44803035434324867,
      "grad_norm": 10.363226890563965,
      "learning_rate": 3.7795531988650663e-06,
      "loss": 0.4245,
      "mean_token_accuracy": 0.8650688961148262,
      "num_tokens": 173892098.0,
      "step": 144530
    },
    {
      "entropy": 1.9519219264388084,
      "epoch": 0.44806135346829834,
      "grad_norm": 8.940476417541504,
      "learning_rate": 3.7794224515036733e-06,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8444177433848381,
      "num_tokens": 173904195.0,
      "step": 144540
    },
    {
      "entropy": 1.998878961801529,
      "epoch": 0.44809235259334806,
      "grad_norm": 8.424077033996582,
      "learning_rate": 3.7792917177103043e-06,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8384242698550224,
      "num_tokens": 173915400.0,
      "step": 144550
    },
    {
      "entropy": 1.9215024933218956,
      "epoch": 0.44812335171839773,
      "grad_norm": 8.416211128234863,
      "learning_rate": 3.7791609974826136e-06,
      "loss": 0.4661,
      "mean_token_accuracy": 0.8600714936852455,
      "num_tokens": 173927090.0,
      "step": 144560
    },
    {
      "entropy": 1.891715730726719,
      "epoch": 0.44815435084344746,
      "grad_norm": 9.589639663696289,
      "learning_rate": 3.7790302908182543e-06,
      "loss": 0.4641,
      "mean_token_accuracy": 0.8505370214581489,
      "num_tokens": 173938920.0,
      "step": 144570
    },
    {
      "entropy": 1.8214015498757363,
      "epoch": 0.4481853499684971,
      "grad_norm": 8.072266578674316,
      "learning_rate": 3.7788995977148823e-06,
      "loss": 0.348,
      "mean_token_accuracy": 0.8764893263578415,
      "num_tokens": 173951952.0,
      "step": 144580
    },
    {
      "entropy": 1.9523023292422295,
      "epoch": 0.44821634909354685,
      "grad_norm": 8.023730278015137,
      "learning_rate": 3.7787689181701514e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8551726579666138,
      "num_tokens": 173963990.0,
      "step": 144590
    },
    {
      "entropy": 1.975418707728386,
      "epoch": 0.4482473482185965,
      "grad_norm": 8.128300666809082,
      "learning_rate": 3.7786382521817178e-06,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8471471086144448,
      "num_tokens": 173974720.0,
      "step": 144600
    },
    {
      "entropy": 1.935040408372879,
      "epoch": 0.44827834734364624,
      "grad_norm": 8.031977653503418,
      "learning_rate": 3.7785075997472385e-06,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8642361879348754,
      "num_tokens": 173986483.0,
      "step": 144610
    },
    {
      "entropy": 1.9660716861486436,
      "epoch": 0.4483093464686959,
      "grad_norm": 7.869020938873291,
      "learning_rate": 3.7783769608643696e-06,
      "loss": 0.4754,
      "mean_token_accuracy": 0.8499971255660057,
      "num_tokens": 173998365.0,
      "step": 144620
    },
    {
      "entropy": 1.998292076587677,
      "epoch": 0.44834034559374564,
      "grad_norm": 8.130206108093262,
      "learning_rate": 3.778246335530769e-06,
      "loss": 0.5028,
      "mean_token_accuracy": 0.8486690282821655,
      "num_tokens": 174009884.0,
      "step": 144630
    },
    {
      "entropy": 1.890485832095146,
      "epoch": 0.4483713447187953,
      "grad_norm": 7.7400946617126465,
      "learning_rate": 3.778115723744095e-06,
      "loss": 0.4081,
      "mean_token_accuracy": 0.8541572690010071,
      "num_tokens": 174022402.0,
      "step": 144640
    },
    {
      "entropy": 1.897654327750206,
      "epoch": 0.44840234384384503,
      "grad_norm": 4.250067710876465,
      "learning_rate": 3.7779851255020057e-06,
      "loss": 0.4291,
      "mean_token_accuracy": 0.8561486795544624,
      "num_tokens": 174034391.0,
      "step": 144650
    },
    {
      "entropy": 1.917440117895603,
      "epoch": 0.4484333429688947,
      "grad_norm": 4.559934139251709,
      "learning_rate": 3.7778545408021607e-06,
      "loss": 0.4682,
      "mean_token_accuracy": 0.8511192619800567,
      "num_tokens": 174046427.0,
      "step": 144660
    },
    {
      "entropy": 1.9716055259108543,
      "epoch": 0.4484643420939444,
      "grad_norm": 7.858667850494385,
      "learning_rate": 3.77772396964222e-06,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8432918503880501,
      "num_tokens": 174057999.0,
      "step": 144670
    },
    {
      "entropy": 1.928077156841755,
      "epoch": 0.4484953412189941,
      "grad_norm": 9.245266914367676,
      "learning_rate": 3.777593412019842e-06,
      "loss": 0.4646,
      "mean_token_accuracy": 0.8530062898993492,
      "num_tokens": 174069219.0,
      "step": 144680
    },
    {
      "entropy": 2.0191250085830688,
      "epoch": 0.4485263403440438,
      "grad_norm": 8.808843612670898,
      "learning_rate": 3.7774628679326895e-06,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8358566120266915,
      "num_tokens": 174079632.0,
      "step": 144690
    },
    {
      "entropy": 1.9445554435253143,
      "epoch": 0.4485573394690935,
      "grad_norm": 10.58156681060791,
      "learning_rate": 3.7773323373784244e-06,
      "loss": 0.4865,
      "mean_token_accuracy": 0.8430225223302841,
      "num_tokens": 174091815.0,
      "step": 144700
    },
    {
      "entropy": 1.9787197232246398,
      "epoch": 0.44858833859414315,
      "grad_norm": 8.029949188232422,
      "learning_rate": 3.777201820354707e-06,
      "loss": 0.4623,
      "mean_token_accuracy": 0.8574790298938751,
      "num_tokens": 174102183.0,
      "step": 144710
    },
    {
      "entropy": 1.8836344301700592,
      "epoch": 0.4486193377191929,
      "grad_norm": 4.910955429077148,
      "learning_rate": 3.777071316859201e-06,
      "loss": 0.4363,
      "mean_token_accuracy": 0.8546113297343254,
      "num_tokens": 174114589.0,
      "step": 144720
    },
    {
      "entropy": 1.8803539738059043,
      "epoch": 0.44865033684424255,
      "grad_norm": 7.87885046005249,
      "learning_rate": 3.776940826889569e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8489265456795693,
      "num_tokens": 174126765.0,
      "step": 144730
    },
    {
      "entropy": 1.9403909876942635,
      "epoch": 0.44868133596929227,
      "grad_norm": 9.444135665893555,
      "learning_rate": 3.776810350443475e-06,
      "loss": 0.489,
      "mean_token_accuracy": 0.8454124689102173,
      "num_tokens": 174138010.0,
      "step": 144740
    },
    {
      "entropy": 1.869876691699028,
      "epoch": 0.44871233509434194,
      "grad_norm": 10.500418663024902,
      "learning_rate": 3.776679887518583e-06,
      "loss": 0.4856,
      "mean_token_accuracy": 0.8416117459535599,
      "num_tokens": 174151284.0,
      "step": 144750
    },
    {
      "entropy": 1.8063198134303093,
      "epoch": 0.44874333421939167,
      "grad_norm": 7.03587532043457,
      "learning_rate": 3.776549438112559e-06,
      "loss": 0.3784,
      "mean_token_accuracy": 0.8664273172616959,
      "num_tokens": 174164715.0,
      "step": 144760
    },
    {
      "entropy": 1.9691897720098495,
      "epoch": 0.44877433334444133,
      "grad_norm": 8.313720703125,
      "learning_rate": 3.7764190022230658e-06,
      "loss": 0.4791,
      "mean_token_accuracy": 0.8534282594919205,
      "num_tokens": 174175915.0,
      "step": 144770
    },
    {
      "entropy": 1.8201162710785865,
      "epoch": 0.44880533246949106,
      "grad_norm": 3.464089870452881,
      "learning_rate": 3.7762885798477715e-06,
      "loss": 0.3616,
      "mean_token_accuracy": 0.8693333268165588,
      "num_tokens": 174188460.0,
      "step": 144780
    },
    {
      "entropy": 1.968066155910492,
      "epoch": 0.4488363315945407,
      "grad_norm": 9.63811206817627,
      "learning_rate": 3.776158170984343e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8475179970264435,
      "num_tokens": 174199357.0,
      "step": 144790
    },
    {
      "entropy": 1.8865119606256484,
      "epoch": 0.44886733071959045,
      "grad_norm": 8.431649208068848,
      "learning_rate": 3.7760277756304458e-06,
      "loss": 0.4675,
      "mean_token_accuracy": 0.8538785234093667,
      "num_tokens": 174211184.0,
      "step": 144800
    },
    {
      "entropy": 1.9904786467552185,
      "epoch": 0.4488983298446401,
      "grad_norm": 8.829815864562988,
      "learning_rate": 3.775897393783749e-06,
      "loss": 0.4995,
      "mean_token_accuracy": 0.849235288798809,
      "num_tokens": 174222072.0,
      "step": 144810
    },
    {
      "entropy": 1.8386689230799675,
      "epoch": 0.44892932896968984,
      "grad_norm": 8.532641410827637,
      "learning_rate": 3.775767025441919e-06,
      "loss": 0.412,
      "mean_token_accuracy": 0.8608880117535591,
      "num_tokens": 174235310.0,
      "step": 144820
    },
    {
      "entropy": 1.9428915694355964,
      "epoch": 0.4489603280947395,
      "grad_norm": 8.705090522766113,
      "learning_rate": 3.7756366706026264e-06,
      "loss": 0.5264,
      "mean_token_accuracy": 0.8452722027897834,
      "num_tokens": 174247069.0,
      "step": 144830
    },
    {
      "entropy": 1.9664770871400834,
      "epoch": 0.44899132721978924,
      "grad_norm": 9.778718948364258,
      "learning_rate": 3.7755063292635396e-06,
      "loss": 0.5638,
      "mean_token_accuracy": 0.8389669686555863,
      "num_tokens": 174257829.0,
      "step": 144840
    },
    {
      "entropy": 1.9240783050656318,
      "epoch": 0.4490223263448389,
      "grad_norm": 7.943138122558594,
      "learning_rate": 3.775376001422329e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8448940262198448,
      "num_tokens": 174269741.0,
      "step": 144850
    },
    {
      "entropy": 1.9432480692863465,
      "epoch": 0.44905332546988863,
      "grad_norm": 3.412449598312378,
      "learning_rate": 3.7752456870766646e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8651240512728691,
      "num_tokens": 174280859.0,
      "step": 144860
    },
    {
      "entropy": 1.8958119705319405,
      "epoch": 0.4490843245949383,
      "grad_norm": 4.008243083953857,
      "learning_rate": 3.775115386224218e-06,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8487641841173172,
      "num_tokens": 174293162.0,
      "step": 144870
    },
    {
      "entropy": 1.9293140321969986,
      "epoch": 0.449115323719988,
      "grad_norm": 8.646997451782227,
      "learning_rate": 3.7749850988626597e-06,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8494552314281464,
      "num_tokens": 174304631.0,
      "step": 144880
    },
    {
      "entropy": 1.8826806560158729,
      "epoch": 0.4491463228450377,
      "grad_norm": 7.910927772521973,
      "learning_rate": 3.7748548249896632e-06,
      "loss": 0.3843,
      "mean_token_accuracy": 0.8651323825120926,
      "num_tokens": 174316973.0,
      "step": 144890
    },
    {
      "entropy": 1.9384344890713692,
      "epoch": 0.4491773219700874,
      "grad_norm": 8.303495407104492,
      "learning_rate": 3.7747245646029e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8473970159888268,
      "num_tokens": 174328666.0,
      "step": 144900
    },
    {
      "entropy": 1.9702891185879707,
      "epoch": 0.4492083210951371,
      "grad_norm": 7.635629177093506,
      "learning_rate": 3.7745943177000442e-06,
      "loss": 0.4746,
      "mean_token_accuracy": 0.84871247112751,
      "num_tokens": 174340268.0,
      "step": 144910
    },
    {
      "entropy": 1.9801568925380706,
      "epoch": 0.4492393202201868,
      "grad_norm": 7.393730640411377,
      "learning_rate": 3.7744640842787706e-06,
      "loss": 0.4899,
      "mean_token_accuracy": 0.8527787283062935,
      "num_tokens": 174350919.0,
      "step": 144920
    },
    {
      "entropy": 1.8521282449364662,
      "epoch": 0.4492703193452365,
      "grad_norm": 7.78226900100708,
      "learning_rate": 3.774333864336751e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.8685169115662574,
      "num_tokens": 174363562.0,
      "step": 144930
    },
    {
      "entropy": 1.8877981454133987,
      "epoch": 0.44930131847028615,
      "grad_norm": 7.9879302978515625,
      "learning_rate": 3.774203657871663e-06,
      "loss": 0.4636,
      "mean_token_accuracy": 0.8551206976175308,
      "num_tokens": 174375369.0,
      "step": 144940
    },
    {
      "entropy": 1.8575106739997864,
      "epoch": 0.4493323175953359,
      "grad_norm": 7.526092529296875,
      "learning_rate": 3.7740734648811805e-06,
      "loss": 0.42,
      "mean_token_accuracy": 0.8627775356173515,
      "num_tokens": 174387464.0,
      "step": 144950
    },
    {
      "entropy": 1.873704120516777,
      "epoch": 0.44936331672038554,
      "grad_norm": 8.647795677185059,
      "learning_rate": 3.7739432853629796e-06,
      "loss": 0.4119,
      "mean_token_accuracy": 0.8583318084478379,
      "num_tokens": 174399542.0,
      "step": 144960
    },
    {
      "entropy": 1.9029436275362968,
      "epoch": 0.44939431584543527,
      "grad_norm": 7.637712001800537,
      "learning_rate": 3.773813119314738e-06,
      "loss": 0.454,
      "mean_token_accuracy": 0.8516765266656876,
      "num_tokens": 174410772.0,
      "step": 144970
    },
    {
      "entropy": 1.8933783307671548,
      "epoch": 0.44942531497048493,
      "grad_norm": 4.113735675811768,
      "learning_rate": 3.7736829667341326e-06,
      "loss": 0.5039,
      "mean_token_accuracy": 0.843044999241829,
      "num_tokens": 174423995.0,
      "step": 144980
    },
    {
      "entropy": 1.9160526722669602,
      "epoch": 0.44945631409553466,
      "grad_norm": 9.520600318908691,
      "learning_rate": 3.7735528276188404e-06,
      "loss": 0.4766,
      "mean_token_accuracy": 0.8556877493858337,
      "num_tokens": 174435604.0,
      "step": 144990
    },
    {
      "entropy": 1.9604547709226607,
      "epoch": 0.44948731322058433,
      "grad_norm": 8.132418632507324,
      "learning_rate": 3.773422701966541e-06,
      "loss": 0.4974,
      "mean_token_accuracy": 0.8487904667854309,
      "num_tokens": 174446584.0,
      "step": 145000
    },
    {
      "entropy": 1.8689698219299316,
      "epoch": 0.44951831234563405,
      "grad_norm": 8.19633960723877,
      "learning_rate": 3.773292589774912e-06,
      "loss": 0.4134,
      "mean_token_accuracy": 0.8521596968173981,
      "num_tokens": 174458869.0,
      "step": 145010
    },
    {
      "entropy": 1.9306217849254608,
      "epoch": 0.4495493114706837,
      "grad_norm": 6.501941204071045,
      "learning_rate": 3.773162491041633e-06,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8544813707470894,
      "num_tokens": 174470488.0,
      "step": 145020
    },
    {
      "entropy": 1.9626944810152054,
      "epoch": 0.44958031059573345,
      "grad_norm": 6.954222202301025,
      "learning_rate": 3.7730324057643857e-06,
      "loss": 0.4843,
      "mean_token_accuracy": 0.8428614303469658,
      "num_tokens": 174480946.0,
      "step": 145030
    },
    {
      "entropy": 1.9757573395967483,
      "epoch": 0.4496113097207831,
      "grad_norm": 7.623712539672852,
      "learning_rate": 3.7729023339408476e-06,
      "loss": 0.5138,
      "mean_token_accuracy": 0.8365162685513496,
      "num_tokens": 174492039.0,
      "step": 145040
    },
    {
      "entropy": 1.8950822830200196,
      "epoch": 0.44964230884583284,
      "grad_norm": 4.679478645324707,
      "learning_rate": 3.7727722755687034e-06,
      "loss": 0.4306,
      "mean_token_accuracy": 0.84790218770504,
      "num_tokens": 174504318.0,
      "step": 145050
    },
    {
      "entropy": 1.8425507709383964,
      "epoch": 0.4496733079708825,
      "grad_norm": 3.7281038761138916,
      "learning_rate": 3.7726422306456324e-06,
      "loss": 0.421,
      "mean_token_accuracy": 0.8493483513593674,
      "num_tokens": 174516826.0,
      "step": 145060
    },
    {
      "entropy": 1.8533750414848327,
      "epoch": 0.44970430709593223,
      "grad_norm": 3.924395799636841,
      "learning_rate": 3.7725121991693183e-06,
      "loss": 0.4451,
      "mean_token_accuracy": 0.8450315102934838,
      "num_tokens": 174530110.0,
      "step": 145070
    },
    {
      "entropy": 1.9302792519330978,
      "epoch": 0.4497353062209819,
      "grad_norm": 12.66701889038086,
      "learning_rate": 3.772382181137442e-06,
      "loss": 0.4751,
      "mean_token_accuracy": 0.846200980246067,
      "num_tokens": 174542159.0,
      "step": 145080
    },
    {
      "entropy": 1.9027525156736373,
      "epoch": 0.4497663053460316,
      "grad_norm": 7.88311243057251,
      "learning_rate": 3.7722521765476877e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8578101888298988,
      "num_tokens": 174552779.0,
      "step": 145090
    },
    {
      "entropy": 1.8144456431269647,
      "epoch": 0.4497973044710813,
      "grad_norm": 7.007596015930176,
      "learning_rate": 3.77212218539774e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8665218591690064,
      "num_tokens": 174566254.0,
      "step": 145100
    },
    {
      "entropy": 1.9618958830833435,
      "epoch": 0.449828303596131,
      "grad_norm": 7.736656665802002,
      "learning_rate": 3.771992207685284e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.836806908249855,
      "num_tokens": 174578147.0,
      "step": 145110
    },
    {
      "entropy": 1.8933872073888778,
      "epoch": 0.4498593027211807,
      "grad_norm": 7.00977087020874,
      "learning_rate": 3.771862243408003e-06,
      "loss": 0.3912,
      "mean_token_accuracy": 0.8770283341407776,
      "num_tokens": 174589833.0,
      "step": 145120
    },
    {
      "entropy": 1.9726043611764907,
      "epoch": 0.4498903018462304,
      "grad_norm": 8.317329406738281,
      "learning_rate": 3.7717322925635836e-06,
      "loss": 0.4768,
      "mean_token_accuracy": 0.8538321673870086,
      "num_tokens": 174600627.0,
      "step": 145130
    },
    {
      "entropy": 1.8826549246907234,
      "epoch": 0.4499213009712801,
      "grad_norm": 4.149418354034424,
      "learning_rate": 3.7716023551497116e-06,
      "loss": 0.3934,
      "mean_token_accuracy": 0.8519825637340546,
      "num_tokens": 174613125.0,
      "step": 145140
    },
    {
      "entropy": 2.011317655444145,
      "epoch": 0.4499523000963298,
      "grad_norm": 7.87576961517334,
      "learning_rate": 3.7714724311640744e-06,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8533690482378006,
      "num_tokens": 174623674.0,
      "step": 145150
    },
    {
      "entropy": 1.9144355922937393,
      "epoch": 0.4499832992213795,
      "grad_norm": 7.214294910430908,
      "learning_rate": 3.771342520604358e-06,
      "loss": 0.4861,
      "mean_token_accuracy": 0.851697339117527,
      "num_tokens": 174635535.0,
      "step": 145160
    },
    {
      "entropy": 1.9184280708432198,
      "epoch": 0.4500142983464292,
      "grad_norm": 4.311148166656494,
      "learning_rate": 3.771212623468252e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8484656348824501,
      "num_tokens": 174647262.0,
      "step": 145170
    },
    {
      "entropy": 1.9903872221708299,
      "epoch": 0.45004529747147887,
      "grad_norm": 8.605646133422852,
      "learning_rate": 3.771082739753443e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8503182783722878,
      "num_tokens": 174658244.0,
      "step": 145180
    },
    {
      "entropy": 1.9099340721964837,
      "epoch": 0.45007629659652854,
      "grad_norm": 9.464656829833984,
      "learning_rate": 3.7709528694576214e-06,
      "loss": 0.4633,
      "mean_token_accuracy": 0.8493123814463616,
      "num_tokens": 174670230.0,
      "step": 145190
    },
    {
      "entropy": 1.962883660197258,
      "epoch": 0.45010729572157826,
      "grad_norm": 8.372581481933594,
      "learning_rate": 3.7708230125784757e-06,
      "loss": 0.4834,
      "mean_token_accuracy": 0.8490772858262062,
      "num_tokens": 174681923.0,
      "step": 145200
    },
    {
      "entropy": 1.8974821627140046,
      "epoch": 0.45013829484662793,
      "grad_norm": 3.493914842605591,
      "learning_rate": 3.7706931691136962e-06,
      "loss": 0.4276,
      "mean_token_accuracy": 0.861706106364727,
      "num_tokens": 174693772.0,
      "step": 145210
    },
    {
      "entropy": 1.9498993948101997,
      "epoch": 0.45016929397167765,
      "grad_norm": 7.030703067779541,
      "learning_rate": 3.7705633390609737e-06,
      "loss": 0.4377,
      "mean_token_accuracy": 0.8613954395055771,
      "num_tokens": 174705327.0,
      "step": 145220
    },
    {
      "entropy": 1.92320823520422,
      "epoch": 0.4502002930967273,
      "grad_norm": 11.262289047241211,
      "learning_rate": 3.770433522418e-06,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8487863168120384,
      "num_tokens": 174717121.0,
      "step": 145230
    },
    {
      "entropy": 1.8713699102401733,
      "epoch": 0.45023129222177705,
      "grad_norm": 9.125288963317871,
      "learning_rate": 3.770303719182465e-06,
      "loss": 0.4408,
      "mean_token_accuracy": 0.8564379021525383,
      "num_tokens": 174729447.0,
      "step": 145240
    },
    {
      "entropy": 1.916731895506382,
      "epoch": 0.4502622913468267,
      "grad_norm": 8.092860221862793,
      "learning_rate": 3.7701739293520634e-06,
      "loss": 0.446,
      "mean_token_accuracy": 0.8585136890411377,
      "num_tokens": 174741652.0,
      "step": 145250
    },
    {
      "entropy": 1.8323576033115387,
      "epoch": 0.45029329047187644,
      "grad_norm": 5.573619365692139,
      "learning_rate": 3.7700441529244865e-06,
      "loss": 0.4288,
      "mean_token_accuracy": 0.858129957318306,
      "num_tokens": 174754464.0,
      "step": 145260
    },
    {
      "entropy": 1.8868038043379785,
      "epoch": 0.4503242895969261,
      "grad_norm": 7.117466926574707,
      "learning_rate": 3.7699143898974273e-06,
      "loss": 0.4213,
      "mean_token_accuracy": 0.8598243370652199,
      "num_tokens": 174767280.0,
      "step": 145270
    },
    {
      "entropy": 1.9327538147568704,
      "epoch": 0.45035528872197583,
      "grad_norm": 7.857034206390381,
      "learning_rate": 3.769784640268581e-06,
      "loss": 0.5033,
      "mean_token_accuracy": 0.8374062940478325,
      "num_tokens": 174779680.0,
      "step": 145280
    },
    {
      "entropy": 1.8759731560945512,
      "epoch": 0.4503862878470255,
      "grad_norm": 9.070082664489746,
      "learning_rate": 3.769654904035642e-06,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8461429521441459,
      "num_tokens": 174792617.0,
      "step": 145290
    },
    {
      "entropy": 1.9318383768200875,
      "epoch": 0.4504172869720752,
      "grad_norm": 8.199836730957031,
      "learning_rate": 3.7695251811963052e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8561635687947273,
      "num_tokens": 174804624.0,
      "step": 145300
    },
    {
      "entropy": 1.9782269150018692,
      "epoch": 0.4504482860971249,
      "grad_norm": 3.9492764472961426,
      "learning_rate": 3.7693954717482656e-06,
      "loss": 0.4845,
      "mean_token_accuracy": 0.847953063249588,
      "num_tokens": 174816707.0,
      "step": 145310
    },
    {
      "entropy": 1.8973633468151092,
      "epoch": 0.4504792852221746,
      "grad_norm": 8.541032791137695,
      "learning_rate": 3.76926577568922e-06,
      "loss": 0.407,
      "mean_token_accuracy": 0.8600034147500992,
      "num_tokens": 174829220.0,
      "step": 145320
    },
    {
      "entropy": 1.868040455877781,
      "epoch": 0.4505102843472243,
      "grad_norm": 8.232982635498047,
      "learning_rate": 3.769136093016865e-06,
      "loss": 0.4017,
      "mean_token_accuracy": 0.860238878428936,
      "num_tokens": 174841040.0,
      "step": 145330
    },
    {
      "entropy": 1.8973990380764008,
      "epoch": 0.450541283472274,
      "grad_norm": 7.8517165184021,
      "learning_rate": 3.769006423728897e-06,
      "loss": 0.4159,
      "mean_token_accuracy": 0.8573246464133263,
      "num_tokens": 174852510.0,
      "step": 145340
    },
    {
      "entropy": 1.88846056163311,
      "epoch": 0.4505722825973237,
      "grad_norm": 7.867887496948242,
      "learning_rate": 3.7688767678230155e-06,
      "loss": 0.4817,
      "mean_token_accuracy": 0.8506530284881592,
      "num_tokens": 174864689.0,
      "step": 145350
    },
    {
      "entropy": 1.9104204386472703,
      "epoch": 0.4506032817223734,
      "grad_norm": 9.259328842163086,
      "learning_rate": 3.768747125296918e-06,
      "loss": 0.463,
      "mean_token_accuracy": 0.8554364711046218,
      "num_tokens": 174876233.0,
      "step": 145360
    },
    {
      "entropy": 1.9462157368659974,
      "epoch": 0.4506342808474231,
      "grad_norm": 8.897783279418945,
      "learning_rate": 3.7686174961483033e-06,
      "loss": 0.4928,
      "mean_token_accuracy": 0.8470346301794052,
      "num_tokens": 174886851.0,
      "step": 145370
    },
    {
      "entropy": 1.9113350063562393,
      "epoch": 0.4506652799724728,
      "grad_norm": 9.343399047851562,
      "learning_rate": 3.768487880374872e-06,
      "loss": 0.4542,
      "mean_token_accuracy": 0.8516996219754219,
      "num_tokens": 174898934.0,
      "step": 145380
    },
    {
      "entropy": 1.9105153515934945,
      "epoch": 0.45069627909752247,
      "grad_norm": 8.577310562133789,
      "learning_rate": 3.768358277974323e-06,
      "loss": 0.4668,
      "mean_token_accuracy": 0.8571393460035324,
      "num_tokens": 174910924.0,
      "step": 145390
    },
    {
      "entropy": 1.7790193900465965,
      "epoch": 0.4507272782225722,
      "grad_norm": 3.691340446472168,
      "learning_rate": 3.7682286889443563e-06,
      "loss": 0.3487,
      "mean_token_accuracy": 0.8683654963970184,
      "num_tokens": 174924748.0,
      "step": 145400
    },
    {
      "entropy": 1.853318177163601,
      "epoch": 0.45075827734762186,
      "grad_norm": 3.717491388320923,
      "learning_rate": 3.768099113282675e-06,
      "loss": 0.419,
      "mean_token_accuracy": 0.8503245025873184,
      "num_tokens": 174937623.0,
      "step": 145410
    },
    {
      "entropy": 1.8263938404619693,
      "epoch": 0.4507892764726716,
      "grad_norm": 3.7551968097686768,
      "learning_rate": 3.7679695509869798e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8548484072089195,
      "num_tokens": 174951248.0,
      "step": 145420
    },
    {
      "entropy": 1.9366182684898376,
      "epoch": 0.45082027559772125,
      "grad_norm": 7.332760810852051,
      "learning_rate": 3.7678400020549727e-06,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8432886257767678,
      "num_tokens": 174962653.0,
      "step": 145430
    },
    {
      "entropy": 1.9753572344779968,
      "epoch": 0.4508512747227709,
      "grad_norm": 8.431512832641602,
      "learning_rate": 3.767710466484357e-06,
      "loss": 0.5259,
      "mean_token_accuracy": 0.839150819182396,
      "num_tokens": 174973717.0,
      "step": 145440
    },
    {
      "entropy": 1.9610534459352493,
      "epoch": 0.45088227384782065,
      "grad_norm": 6.784049034118652,
      "learning_rate": 3.767580944272836e-06,
      "loss": 0.5492,
      "mean_token_accuracy": 0.845959635078907,
      "num_tokens": 174985280.0,
      "step": 145450
    },
    {
      "entropy": 1.8077823013067245,
      "epoch": 0.4509132729728703,
      "grad_norm": 9.173667907714844,
      "learning_rate": 3.767451435418114e-06,
      "loss": 0.3778,
      "mean_token_accuracy": 0.8598848000168801,
      "num_tokens": 174999259.0,
      "step": 145460
    },
    {
      "entropy": 1.8944556072354317,
      "epoch": 0.45094427209792004,
      "grad_norm": 8.268648147583008,
      "learning_rate": 3.767321939917894e-06,
      "loss": 0.4506,
      "mean_token_accuracy": 0.861379300057888,
      "num_tokens": 175010366.0,
      "step": 145470
    },
    {
      "entropy": 1.8945633813738822,
      "epoch": 0.4509752712229697,
      "grad_norm": 3.7836310863494873,
      "learning_rate": 3.7671924577698832e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8532457053661346,
      "num_tokens": 175022036.0,
      "step": 145480
    },
    {
      "entropy": 1.9272257044911385,
      "epoch": 0.45100627034801943,
      "grad_norm": 9.02260971069336,
      "learning_rate": 3.767062988971786e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8532265886664391,
      "num_tokens": 175034153.0,
      "step": 145490
    },
    {
      "entropy": 1.9854293823242188,
      "epoch": 0.4510372694730691,
      "grad_norm": 7.62827730178833,
      "learning_rate": 3.766933533521308e-06,
      "loss": 0.4918,
      "mean_token_accuracy": 0.8465104550123215,
      "num_tokens": 175045152.0,
      "step": 145500
    },
    {
      "entropy": 1.888518001139164,
      "epoch": 0.4510682685981188,
      "grad_norm": 7.6335883140563965,
      "learning_rate": 3.766804091416157e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8494478777050972,
      "num_tokens": 175057796.0,
      "step": 145510
    },
    {
      "entropy": 1.916823922097683,
      "epoch": 0.4510992677231685,
      "grad_norm": 9.0162935256958,
      "learning_rate": 3.76667466265404e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8522469475865364,
      "num_tokens": 175069430.0,
      "step": 145520
    },
    {
      "entropy": 1.8178383886814118,
      "epoch": 0.4511302668482182,
      "grad_norm": 8.224920272827148,
      "learning_rate": 3.766545247232664e-06,
      "loss": 0.4528,
      "mean_token_accuracy": 0.8601442322134971,
      "num_tokens": 175082985.0,
      "step": 145530
    },
    {
      "entropy": 1.898731729388237,
      "epoch": 0.4511612659732679,
      "grad_norm": 3.8143157958984375,
      "learning_rate": 3.7664158451497383e-06,
      "loss": 0.4486,
      "mean_token_accuracy": 0.8475686386227608,
      "num_tokens": 175095045.0,
      "step": 145540
    },
    {
      "entropy": 1.9468153685331344,
      "epoch": 0.4511922650983176,
      "grad_norm": 6.286230564117432,
      "learning_rate": 3.766286456402971e-06,
      "loss": 0.4355,
      "mean_token_accuracy": 0.8552075073122978,
      "num_tokens": 175106619.0,
      "step": 145550
    },
    {
      "entropy": 1.8661633163690567,
      "epoch": 0.4512232642233673,
      "grad_norm": 3.8550450801849365,
      "learning_rate": 3.766157080990073e-06,
      "loss": 0.4514,
      "mean_token_accuracy": 0.8487115263938904,
      "num_tokens": 175119296.0,
      "step": 145560
    },
    {
      "entropy": 1.928366206586361,
      "epoch": 0.451254263348417,
      "grad_norm": 7.825737953186035,
      "learning_rate": 3.7660277189087524e-06,
      "loss": 0.4441,
      "mean_token_accuracy": 0.8565060004591942,
      "num_tokens": 175130424.0,
      "step": 145570
    },
    {
      "entropy": 1.9019853085279466,
      "epoch": 0.4512852624734667,
      "grad_norm": 4.290144920349121,
      "learning_rate": 3.7658983701567215e-06,
      "loss": 0.4669,
      "mean_token_accuracy": 0.8472353518009186,
      "num_tokens": 175143118.0,
      "step": 145580
    },
    {
      "entropy": 1.811234064400196,
      "epoch": 0.4513162615985164,
      "grad_norm": 8.767404556274414,
      "learning_rate": 3.7657690347316896e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.859492601454258,
      "num_tokens": 175155763.0,
      "step": 145590
    },
    {
      "entropy": 1.9169629096984864,
      "epoch": 0.45134726072356607,
      "grad_norm": 8.782833099365234,
      "learning_rate": 3.7656397126313704e-06,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8517961174249649,
      "num_tokens": 175167606.0,
      "step": 145600
    },
    {
      "entropy": 1.9656015813350678,
      "epoch": 0.4513782598486158,
      "grad_norm": 7.532365322113037,
      "learning_rate": 3.765510403853474e-06,
      "loss": 0.5174,
      "mean_token_accuracy": 0.8420636489987373,
      "num_tokens": 175178779.0,
      "step": 145610
    },
    {
      "entropy": 1.89768455773592,
      "epoch": 0.45140925897366546,
      "grad_norm": 7.2777419090271,
      "learning_rate": 3.765381108395715e-06,
      "loss": 0.3982,
      "mean_token_accuracy": 0.8631177753210068,
      "num_tokens": 175190689.0,
      "step": 145620
    },
    {
      "entropy": 1.8941746070981025,
      "epoch": 0.4514402580987152,
      "grad_norm": 7.515166282653809,
      "learning_rate": 3.7652518262558054e-06,
      "loss": 0.4465,
      "mean_token_accuracy": 0.8515808820724488,
      "num_tokens": 175202846.0,
      "step": 145630
    },
    {
      "entropy": 1.8148242741823197,
      "epoch": 0.45147125722376485,
      "grad_norm": 3.4633376598358154,
      "learning_rate": 3.7651225574314597e-06,
      "loss": 0.3591,
      "mean_token_accuracy": 0.8688716858625412,
      "num_tokens": 175215871.0,
      "step": 145640
    },
    {
      "entropy": 1.897299911081791,
      "epoch": 0.4515022563488146,
      "grad_norm": 8.746367454528809,
      "learning_rate": 3.764993301920392e-06,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8508375376462937,
      "num_tokens": 175227683.0,
      "step": 145650
    },
    {
      "entropy": 1.9884761601686478,
      "epoch": 0.45153325547386425,
      "grad_norm": 7.514569282531738,
      "learning_rate": 3.764864059720317e-06,
      "loss": 0.4895,
      "mean_token_accuracy": 0.8484263613820076,
      "num_tokens": 175238305.0,
      "step": 145660
    },
    {
      "entropy": 1.8774521455168725,
      "epoch": 0.45156425459891397,
      "grad_norm": 4.357245922088623,
      "learning_rate": 3.7647348308289522e-06,
      "loss": 0.4094,
      "mean_token_accuracy": 0.8594071462750434,
      "num_tokens": 175250175.0,
      "step": 145670
    },
    {
      "entropy": 1.9650772213935852,
      "epoch": 0.45159525372396364,
      "grad_norm": 9.48290729522705,
      "learning_rate": 3.7646056152440104e-06,
      "loss": 0.5147,
      "mean_token_accuracy": 0.8337369963526726,
      "num_tokens": 175261255.0,
      "step": 145680
    },
    {
      "entropy": 1.8208548158407212,
      "epoch": 0.4516262528490133,
      "grad_norm": 9.698999404907227,
      "learning_rate": 3.7644764129632104e-06,
      "loss": 0.404,
      "mean_token_accuracy": 0.8611343666911125,
      "num_tokens": 175274442.0,
      "step": 145690
    },
    {
      "entropy": 1.8542212471365929,
      "epoch": 0.45165725197406303,
      "grad_norm": 8.707727432250977,
      "learning_rate": 3.7643472239842692e-06,
      "loss": 0.4699,
      "mean_token_accuracy": 0.8457312777638435,
      "num_tokens": 175287615.0,
      "step": 145700
    },
    {
      "entropy": 1.947345346212387,
      "epoch": 0.4516882510991127,
      "grad_norm": 9.62697696685791,
      "learning_rate": 3.7642180483049036e-06,
      "loss": 0.5165,
      "mean_token_accuracy": 0.8280424475669861,
      "num_tokens": 175299235.0,
      "step": 145710
    },
    {
      "entropy": 1.7479452803730964,
      "epoch": 0.4517192502241624,
      "grad_norm": 7.1266350746154785,
      "learning_rate": 3.7640888859228326e-06,
      "loss": 0.3448,
      "mean_token_accuracy": 0.8756140992045403,
      "num_tokens": 175312946.0,
      "step": 145720
    },
    {
      "entropy": 1.9307225465774536,
      "epoch": 0.4517502493492121,
      "grad_norm": 10.900723457336426,
      "learning_rate": 3.7639597368357745e-06,
      "loss": 0.4533,
      "mean_token_accuracy": 0.8503560811281204,
      "num_tokens": 175324628.0,
      "step": 145730
    },
    {
      "entropy": 1.9207808002829552,
      "epoch": 0.4517812484742618,
      "grad_norm": 8.007662773132324,
      "learning_rate": 3.763830601041448e-06,
      "loss": 0.4391,
      "mean_token_accuracy": 0.8547277525067329,
      "num_tokens": 175336268.0,
      "step": 145740
    },
    {
      "entropy": 1.9525502398610115,
      "epoch": 0.4518122475993115,
      "grad_norm": 7.941483974456787,
      "learning_rate": 3.763701478537576e-06,
      "loss": 0.4872,
      "mean_token_accuracy": 0.8508265346288681,
      "num_tokens": 175347435.0,
      "step": 145750
    },
    {
      "entropy": 1.8660466879606248,
      "epoch": 0.4518432467243612,
      "grad_norm": 4.241132736206055,
      "learning_rate": 3.763572369321876e-06,
      "loss": 0.4325,
      "mean_token_accuracy": 0.8603238686919212,
      "num_tokens": 175359501.0,
      "step": 145760
    },
    {
      "entropy": 1.9864515900611877,
      "epoch": 0.4518742458494109,
      "grad_norm": 8.52209758758545,
      "learning_rate": 3.763443273392069e-06,
      "loss": 0.539,
      "mean_token_accuracy": 0.8401314124464989,
      "num_tokens": 175370454.0,
      "step": 145770
    },
    {
      "entropy": 1.917511025071144,
      "epoch": 0.4519052449744606,
      "grad_norm": 4.007366180419922,
      "learning_rate": 3.7633141907458774e-06,
      "loss": 0.4836,
      "mean_token_accuracy": 0.8430694580078125,
      "num_tokens": 175382097.0,
      "step": 145780
    },
    {
      "entropy": 1.8464987218379973,
      "epoch": 0.4519362440995103,
      "grad_norm": 4.244717121124268,
      "learning_rate": 3.7631851213810237e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8494131848216057,
      "num_tokens": 175396376.0,
      "step": 145790
    },
    {
      "entropy": 1.8551280453801156,
      "epoch": 0.45196724322456,
      "grad_norm": 8.366869926452637,
      "learning_rate": 3.76305606529523e-06,
      "loss": 0.4078,
      "mean_token_accuracy": 0.8622759610414505,
      "num_tokens": 175408942.0,
      "step": 145800
    },
    {
      "entropy": 1.8460657626390458,
      "epoch": 0.45199824234960967,
      "grad_norm": 7.980034351348877,
      "learning_rate": 3.7629270224862198e-06,
      "loss": 0.3803,
      "mean_token_accuracy": 0.8638658255338669,
      "num_tokens": 175421526.0,
      "step": 145810
    },
    {
      "entropy": 1.8811449334025383,
      "epoch": 0.4520292414746594,
      "grad_norm": 8.846792221069336,
      "learning_rate": 3.762797992951716e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8668080389499664,
      "num_tokens": 175434163.0,
      "step": 145820
    },
    {
      "entropy": 1.8398293122649192,
      "epoch": 0.45206024059970906,
      "grad_norm": 8.336902618408203,
      "learning_rate": 3.762668976689443e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8575309097766877,
      "num_tokens": 175447165.0,
      "step": 145830
    },
    {
      "entropy": 1.8911280773580075,
      "epoch": 0.4520912397247588,
      "grad_norm": 3.6037964820861816,
      "learning_rate": 3.7625399736971264e-06,
      "loss": 0.4532,
      "mean_token_accuracy": 0.8530662402510643,
      "num_tokens": 175458732.0,
      "step": 145840
    },
    {
      "entropy": 1.937391071021557,
      "epoch": 0.45212223884980846,
      "grad_norm": 7.425451755523682,
      "learning_rate": 3.7624109839724915e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8544071182608605,
      "num_tokens": 175470385.0,
      "step": 145850
    },
    {
      "entropy": 1.9437742054462432,
      "epoch": 0.4521532379748582,
      "grad_norm": 7.55548095703125,
      "learning_rate": 3.762282007513263e-06,
      "loss": 0.459,
      "mean_token_accuracy": 0.8576256543397903,
      "num_tokens": 175482257.0,
      "step": 145860
    },
    {
      "entropy": 1.879883836209774,
      "epoch": 0.45218423709990785,
      "grad_norm": 9.295843124389648,
      "learning_rate": 3.7621530443171695e-06,
      "loss": 0.4445,
      "mean_token_accuracy": 0.8495224505662918,
      "num_tokens": 175494991.0,
      "step": 145870
    },
    {
      "entropy": 1.8533268101513385,
      "epoch": 0.4522152362249576,
      "grad_norm": 2.5859084129333496,
      "learning_rate": 3.7620240943819353e-06,
      "loss": 0.4178,
      "mean_token_accuracy": 0.8556357622146606,
      "num_tokens": 175507885.0,
      "step": 145880
    },
    {
      "entropy": 1.8991135403513908,
      "epoch": 0.45224623535000724,
      "grad_norm": 8.07573127746582,
      "learning_rate": 3.7618951577052897e-06,
      "loss": 0.4416,
      "mean_token_accuracy": 0.8474540561437607,
      "num_tokens": 175519719.0,
      "step": 145890
    },
    {
      "entropy": 1.9330607324838638,
      "epoch": 0.45227723447505697,
      "grad_norm": 7.152818202972412,
      "learning_rate": 3.7617662342849608e-06,
      "loss": 0.4621,
      "mean_token_accuracy": 0.8559790030121803,
      "num_tokens": 175531157.0,
      "step": 145900
    },
    {
      "entropy": 1.875030305981636,
      "epoch": 0.45230823360010664,
      "grad_norm": 10.458932876586914,
      "learning_rate": 3.7616373241186765e-06,
      "loss": 0.4402,
      "mean_token_accuracy": 0.85587347894907,
      "num_tokens": 175542756.0,
      "step": 145910
    },
    {
      "entropy": 1.9337833374738693,
      "epoch": 0.45233923272515636,
      "grad_norm": 9.064762115478516,
      "learning_rate": 3.7615084272041664e-06,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8471284449100495,
      "num_tokens": 175555072.0,
      "step": 145920
    },
    {
      "entropy": 1.9829118058085442,
      "epoch": 0.45237023185020603,
      "grad_norm": 7.892855644226074,
      "learning_rate": 3.7613795435391603e-06,
      "loss": 0.4718,
      "mean_token_accuracy": 0.8524794027209281,
      "num_tokens": 175566512.0,
      "step": 145930
    },
    {
      "entropy": 1.8689216911792754,
      "epoch": 0.4524012309752557,
      "grad_norm": 7.025180816650391,
      "learning_rate": 3.761250673121388e-06,
      "loss": 0.4345,
      "mean_token_accuracy": 0.8556947037577629,
      "num_tokens": 175577952.0,
      "step": 145940
    },
    {
      "entropy": 1.8746563777327538,
      "epoch": 0.4524322301003054,
      "grad_norm": 8.948837280273438,
      "learning_rate": 3.7611218159485807e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.8540759429335594,
      "num_tokens": 175589918.0,
      "step": 145950
    },
    {
      "entropy": 1.906599646806717,
      "epoch": 0.4524632292253551,
      "grad_norm": 8.048293113708496,
      "learning_rate": 3.7609929720184695e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8491839274764061,
      "num_tokens": 175602264.0,
      "step": 145960
    },
    {
      "entropy": 1.8783759981393815,
      "epoch": 0.4524942283504048,
      "grad_norm": 8.030713081359863,
      "learning_rate": 3.7608641413287865e-06,
      "loss": 0.3877,
      "mean_token_accuracy": 0.8676090762019157,
      "num_tokens": 175613895.0,
      "step": 145970
    },
    {
      "entropy": 1.8621134147047997,
      "epoch": 0.4525252274754545,
      "grad_norm": 9.925078392028809,
      "learning_rate": 3.760735323877264e-06,
      "loss": 0.4474,
      "mean_token_accuracy": 0.8541931763291359,
      "num_tokens": 175625978.0,
      "step": 145980
    },
    {
      "entropy": 1.9505731910467148,
      "epoch": 0.4525562266005042,
      "grad_norm": 8.451035499572754,
      "learning_rate": 3.7606065196616353e-06,
      "loss": 0.4809,
      "mean_token_accuracy": 0.8430397853255271,
      "num_tokens": 175637809.0,
      "step": 145990
    },
    {
      "entropy": 1.9288512140512466,
      "epoch": 0.4525872257255539,
      "grad_norm": 9.2060546875,
      "learning_rate": 3.7604777286796333e-06,
      "loss": 0.4286,
      "mean_token_accuracy": 0.8581225946545601,
      "num_tokens": 175649771.0,
      "step": 146000
    },
    {
      "entropy": 1.9088809505105018,
      "epoch": 0.4526182248506036,
      "grad_norm": 9.723130226135254,
      "learning_rate": 3.7603489509289924e-06,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8498413473367691,
      "num_tokens": 175661361.0,
      "step": 146010
    },
    {
      "entropy": 1.8807017832994462,
      "epoch": 0.45264922397565327,
      "grad_norm": 7.284520626068115,
      "learning_rate": 3.7602201864074476e-06,
      "loss": 0.4296,
      "mean_token_accuracy": 0.8585941612720489,
      "num_tokens": 175674036.0,
      "step": 146020
    },
    {
      "entropy": 1.9656174287199975,
      "epoch": 0.452680223100703,
      "grad_norm": 9.173661231994629,
      "learning_rate": 3.7600914351127343e-06,
      "loss": 0.4841,
      "mean_token_accuracy": 0.8418927356600762,
      "num_tokens": 175685277.0,
      "step": 146030
    },
    {
      "entropy": 1.9602088913321496,
      "epoch": 0.45271122222575266,
      "grad_norm": 8.204484939575195,
      "learning_rate": 3.759962697042587e-06,
      "loss": 0.4398,
      "mean_token_accuracy": 0.8636955425143242,
      "num_tokens": 175696654.0,
      "step": 146040
    },
    {
      "entropy": 1.886138205230236,
      "epoch": 0.4527422213508024,
      "grad_norm": 8.34501838684082,
      "learning_rate": 3.7598339721947426e-06,
      "loss": 0.4302,
      "mean_token_accuracy": 0.85477264970541,
      "num_tokens": 175708775.0,
      "step": 146050
    },
    {
      "entropy": 1.8968783617019653,
      "epoch": 0.45277322047585206,
      "grad_norm": 8.124290466308594,
      "learning_rate": 3.7597052605669376e-06,
      "loss": 0.4263,
      "mean_token_accuracy": 0.8634179666638374,
      "num_tokens": 175720239.0,
      "step": 146060
    },
    {
      "entropy": 1.9461852222681046,
      "epoch": 0.4528042196009018,
      "grad_norm": 10.085100173950195,
      "learning_rate": 3.7595765621569098e-06,
      "loss": 0.4547,
      "mean_token_accuracy": 0.8562173038721085,
      "num_tokens": 175731786.0,
      "step": 146070
    },
    {
      "entropy": 1.8739292353391648,
      "epoch": 0.45283521872595145,
      "grad_norm": 8.647810935974121,
      "learning_rate": 3.7594478769623967e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.8552744179964066,
      "num_tokens": 175744743.0,
      "step": 146080
    },
    {
      "entropy": 1.972387745976448,
      "epoch": 0.4528662178510012,
      "grad_norm": 8.841172218322754,
      "learning_rate": 3.759319204981136e-06,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8444529309868812,
      "num_tokens": 175756143.0,
      "step": 146090
    },
    {
      "entropy": 1.804822953045368,
      "epoch": 0.45289721697605084,
      "grad_norm": 7.164949893951416,
      "learning_rate": 3.7591905462108686e-06,
      "loss": 0.374,
      "mean_token_accuracy": 0.8675896748900414,
      "num_tokens": 175768824.0,
      "step": 146100
    },
    {
      "entropy": 1.858502623438835,
      "epoch": 0.45292821610110057,
      "grad_norm": 7.452975749969482,
      "learning_rate": 3.7590619006493333e-06,
      "loss": 0.4396,
      "mean_token_accuracy": 0.8505475029349328,
      "num_tokens": 175781202.0,
      "step": 146110
    },
    {
      "entropy": 1.9403677567839623,
      "epoch": 0.45295921522615024,
      "grad_norm": 9.553077697753906,
      "learning_rate": 3.7589332682942687e-06,
      "loss": 0.4697,
      "mean_token_accuracy": 0.8545985773205758,
      "num_tokens": 175792433.0,
      "step": 146120
    },
    {
      "entropy": 1.929755797982216,
      "epoch": 0.45299021435119996,
      "grad_norm": 7.850072860717773,
      "learning_rate": 3.7588046491434164e-06,
      "loss": 0.4831,
      "mean_token_accuracy": 0.8516537502408028,
      "num_tokens": 175803014.0,
      "step": 146130
    },
    {
      "entropy": 1.8702018111944199,
      "epoch": 0.45302121347624963,
      "grad_norm": 4.102813243865967,
      "learning_rate": 3.758676043194518e-06,
      "loss": 0.3885,
      "mean_token_accuracy": 0.8598198860883712,
      "num_tokens": 175814852.0,
      "step": 146140
    },
    {
      "entropy": 1.875620885193348,
      "epoch": 0.45305221260129935,
      "grad_norm": 7.794433116912842,
      "learning_rate": 3.7585474504453145e-06,
      "loss": 0.4321,
      "mean_token_accuracy": 0.8569299295544625,
      "num_tokens": 175826845.0,
      "step": 146150
    },
    {
      "entropy": 1.9491507172584535,
      "epoch": 0.453083211726349,
      "grad_norm": 8.696412086486816,
      "learning_rate": 3.758418870893548e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8512806192040443,
      "num_tokens": 175838457.0,
      "step": 146160
    },
    {
      "entropy": 1.8456061393022538,
      "epoch": 0.45311421085139875,
      "grad_norm": 3.9336066246032715,
      "learning_rate": 3.758290304536962e-06,
      "loss": 0.3994,
      "mean_token_accuracy": 0.861776913702488,
      "num_tokens": 175850893.0,
      "step": 146170
    },
    {
      "entropy": 1.8336857661604882,
      "epoch": 0.4531452099764484,
      "grad_norm": 4.638808727264404,
      "learning_rate": 3.7581617513732994e-06,
      "loss": 0.3938,
      "mean_token_accuracy": 0.8632370933890343,
      "num_tokens": 175863816.0,
      "step": 146180
    },
    {
      "entropy": 1.8227697402238845,
      "epoch": 0.4531762091014981,
      "grad_norm": 8.638792991638184,
      "learning_rate": 3.7580332114003033e-06,
      "loss": 0.475,
      "mean_token_accuracy": 0.8531124874949455,
      "num_tokens": 175877462.0,
      "step": 146190
    },
    {
      "entropy": 1.938033263385296,
      "epoch": 0.4532072082265478,
      "grad_norm": 9.167952537536621,
      "learning_rate": 3.757904684615719e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8607913628220558,
      "num_tokens": 175889063.0,
      "step": 146200
    },
    {
      "entropy": 1.8250967353582381,
      "epoch": 0.4532382073515975,
      "grad_norm": 7.031805515289307,
      "learning_rate": 3.7577761710172912e-06,
      "loss": 0.3626,
      "mean_token_accuracy": 0.871262151002884,
      "num_tokens": 175902376.0,
      "step": 146210
    },
    {
      "entropy": 1.872345322370529,
      "epoch": 0.4532692064766472,
      "grad_norm": 7.401670932769775,
      "learning_rate": 3.757647670602764e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8529538735747337,
      "num_tokens": 175915132.0,
      "step": 146220
    },
    {
      "entropy": 1.8786262601613999,
      "epoch": 0.45330020560169687,
      "grad_norm": 4.830646514892578,
      "learning_rate": 3.757519183369886e-06,
      "loss": 0.4165,
      "mean_token_accuracy": 0.85924152135849,
      "num_tokens": 175927680.0,
      "step": 146230
    },
    {
      "entropy": 1.9301627531647683,
      "epoch": 0.4533312047267466,
      "grad_norm": 4.050126552581787,
      "learning_rate": 3.757390709316402e-06,
      "loss": 0.457,
      "mean_token_accuracy": 0.8585702002048492,
      "num_tokens": 175938798.0,
      "step": 146240
    },
    {
      "entropy": 1.909451249241829,
      "epoch": 0.45336220385179626,
      "grad_norm": 7.195918560028076,
      "learning_rate": 3.757262248440059e-06,
      "loss": 0.4654,
      "mean_token_accuracy": 0.8493908762931823,
      "num_tokens": 175949970.0,
      "step": 146250
    },
    {
      "entropy": 1.9241393342614175,
      "epoch": 0.453393202976846,
      "grad_norm": 3.823732614517212,
      "learning_rate": 3.7571338007386053e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.830666047334671,
      "num_tokens": 175961893.0,
      "step": 146260
    },
    {
      "entropy": 1.9584316313266754,
      "epoch": 0.45342420210189566,
      "grad_norm": 7.582314968109131,
      "learning_rate": 3.7570053662097884e-06,
      "loss": 0.4784,
      "mean_token_accuracy": 0.8390909880399704,
      "num_tokens": 175973692.0,
      "step": 146270
    },
    {
      "entropy": 1.8017342045903206,
      "epoch": 0.4534552012269454,
      "grad_norm": 4.599339008331299,
      "learning_rate": 3.7568769448513577e-06,
      "loss": 0.336,
      "mean_token_accuracy": 0.87629035115242,
      "num_tokens": 175987073.0,
      "step": 146280
    },
    {
      "entropy": 1.9104434236884118,
      "epoch": 0.45348620035199505,
      "grad_norm": 5.8684210777282715,
      "learning_rate": 3.756748536661061e-06,
      "loss": 0.4021,
      "mean_token_accuracy": 0.8615088537335396,
      "num_tokens": 175999179.0,
      "step": 146290
    },
    {
      "entropy": 1.9020302399992943,
      "epoch": 0.4535171994770448,
      "grad_norm": 8.238286972045898,
      "learning_rate": 3.7566201416366497e-06,
      "loss": 0.4678,
      "mean_token_accuracy": 0.850467374920845,
      "num_tokens": 176010704.0,
      "step": 146300
    },
    {
      "entropy": 1.9250214383006097,
      "epoch": 0.45354819860209444,
      "grad_norm": 7.398449897766113,
      "learning_rate": 3.756491759775874e-06,
      "loss": 0.4741,
      "mean_token_accuracy": 0.8439469993114471,
      "num_tokens": 176023049.0,
      "step": 146310
    },
    {
      "entropy": 1.9645772278308868,
      "epoch": 0.45357919772714417,
      "grad_norm": 7.6212334632873535,
      "learning_rate": 3.7563633910764837e-06,
      "loss": 0.5461,
      "mean_token_accuracy": 0.8408754646778107,
      "num_tokens": 176034032.0,
      "step": 146320
    },
    {
      "entropy": 1.9224450066685677,
      "epoch": 0.45361019685219384,
      "grad_norm": 5.688684463500977,
      "learning_rate": 3.7562350355362297e-06,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8347508668899536,
      "num_tokens": 176046264.0,
      "step": 146330
    },
    {
      "entropy": 1.9145216554403306,
      "epoch": 0.45364119597724356,
      "grad_norm": 8.109264373779297,
      "learning_rate": 3.7561066931528657e-06,
      "loss": 0.4583,
      "mean_token_accuracy": 0.8559110507369041,
      "num_tokens": 176057654.0,
      "step": 146340
    },
    {
      "entropy": 1.903997114300728,
      "epoch": 0.45367219510229323,
      "grad_norm": 7.708643436431885,
      "learning_rate": 3.755978363924143e-06,
      "loss": 0.4314,
      "mean_token_accuracy": 0.8482723668217659,
      "num_tokens": 176070619.0,
      "step": 146350
    },
    {
      "entropy": 1.8557177141308785,
      "epoch": 0.45370319422734295,
      "grad_norm": 4.477802753448486,
      "learning_rate": 3.755850047847815e-06,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8525144010782242,
      "num_tokens": 176083576.0,
      "step": 146360
    },
    {
      "entropy": 1.9063315868377686,
      "epoch": 0.4537341933523926,
      "grad_norm": 8.484750747680664,
      "learning_rate": 3.7557217449216354e-06,
      "loss": 0.4557,
      "mean_token_accuracy": 0.8580427676439285,
      "num_tokens": 176095193.0,
      "step": 146370
    },
    {
      "entropy": 1.8667533531785012,
      "epoch": 0.45376519247744235,
      "grad_norm": 9.92103099822998,
      "learning_rate": 3.755593455143357e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8585796818137169,
      "num_tokens": 176107763.0,
      "step": 146380
    },
    {
      "entropy": 1.9320159569382667,
      "epoch": 0.453796191602492,
      "grad_norm": 9.859814643859863,
      "learning_rate": 3.7554651785107367e-06,
      "loss": 0.4778,
      "mean_token_accuracy": 0.8518664732575416,
      "num_tokens": 176119971.0,
      "step": 146390
    },
    {
      "entropy": 1.8948560282588005,
      "epoch": 0.45382719072754174,
      "grad_norm": 3.578590154647827,
      "learning_rate": 3.755336915021527e-06,
      "loss": 0.4498,
      "mean_token_accuracy": 0.8504968300461769,
      "num_tokens": 176132352.0,
      "step": 146400
    },
    {
      "entropy": 1.9224641382694245,
      "epoch": 0.4538581898525914,
      "grad_norm": 9.709887504577637,
      "learning_rate": 3.755208664673485e-06,
      "loss": 0.452,
      "mean_token_accuracy": 0.849474447965622,
      "num_tokens": 176145215.0,
      "step": 146410
    },
    {
      "entropy": 1.9500353157520294,
      "epoch": 0.45388918897764113,
      "grad_norm": 7.521528720855713,
      "learning_rate": 3.7550804274643673e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8548279941082001,
      "num_tokens": 176156902.0,
      "step": 146420
    },
    {
      "entropy": 1.9636742144823074,
      "epoch": 0.4539201881026908,
      "grad_norm": 9.059736251831055,
      "learning_rate": 3.7549522033919293e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8536644339561462,
      "num_tokens": 176167981.0,
      "step": 146430
    },
    {
      "entropy": 1.9579711258411407,
      "epoch": 0.45395118722774047,
      "grad_norm": 7.5242390632629395,
      "learning_rate": 3.7548239924539294e-06,
      "loss": 0.4569,
      "mean_token_accuracy": 0.855312067270279,
      "num_tokens": 176179862.0,
      "step": 146440
    },
    {
      "entropy": 1.9306029558181763,
      "epoch": 0.4539821863527902,
      "grad_norm": 8.735157012939453,
      "learning_rate": 3.754695794648125e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8461642101407051,
      "num_tokens": 176191558.0,
      "step": 146450
    },
    {
      "entropy": 1.8903416648507119,
      "epoch": 0.45401318547783986,
      "grad_norm": 8.93207836151123,
      "learning_rate": 3.7545676099722737e-06,
      "loss": 0.4224,
      "mean_token_accuracy": 0.8568781584501266,
      "num_tokens": 176203878.0,
      "step": 146460
    },
    {
      "entropy": 1.918420398235321,
      "epoch": 0.4540441846028896,
      "grad_norm": 8.027023315429688,
      "learning_rate": 3.7544394384241366e-06,
      "loss": 0.4635,
      "mean_token_accuracy": 0.8519258230924607,
      "num_tokens": 176215578.0,
      "step": 146470
    },
    {
      "entropy": 1.9900656551122666,
      "epoch": 0.45407518372793926,
      "grad_norm": 6.475451469421387,
      "learning_rate": 3.754311280001471e-06,
      "loss": 0.4894,
      "mean_token_accuracy": 0.8443711012601852,
      "num_tokens": 176226440.0,
      "step": 146480
    },
    {
      "entropy": 1.9735152557492257,
      "epoch": 0.454106182852989,
      "grad_norm": 8.966446876525879,
      "learning_rate": 3.7541831347020374e-06,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8361330017447471,
      "num_tokens": 176237579.0,
      "step": 146490
    },
    {
      "entropy": 1.8347552955150603,
      "epoch": 0.45413718197803865,
      "grad_norm": 4.222280502319336,
      "learning_rate": 3.754055002523596e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8571354284882545,
      "num_tokens": 176251272.0,
      "step": 146500
    },
    {
      "entropy": 1.8998313397169113,
      "epoch": 0.4541681811030884,
      "grad_norm": 8.564806938171387,
      "learning_rate": 3.7539268834639085e-06,
      "loss": 0.4753,
      "mean_token_accuracy": 0.8468169465661048,
      "num_tokens": 176263481.0,
      "step": 146510
    },
    {
      "entropy": 1.912764126062393,
      "epoch": 0.45419918022813804,
      "grad_norm": 8.024970054626465,
      "learning_rate": 3.7537987775207373e-06,
      "loss": 0.3862,
      "mean_token_accuracy": 0.8708274886012077,
      "num_tokens": 176275846.0,
      "step": 146520
    },
    {
      "entropy": 1.9775932729244232,
      "epoch": 0.45423017935318777,
      "grad_norm": 10.816838264465332,
      "learning_rate": 3.753670684691842e-06,
      "loss": 0.5207,
      "mean_token_accuracy": 0.842138460278511,
      "num_tokens": 176286757.0,
      "step": 146530
    },
    {
      "entropy": 1.9837248116731643,
      "epoch": 0.45426117847823744,
      "grad_norm": 8.28238296508789,
      "learning_rate": 3.7535426049749867e-06,
      "loss": 0.4965,
      "mean_token_accuracy": 0.8460395529866218,
      "num_tokens": 176297210.0,
      "step": 146540
    },
    {
      "entropy": 1.9528759866952896,
      "epoch": 0.45429217760328716,
      "grad_norm": 8.31442642211914,
      "learning_rate": 3.7534145383679354e-06,
      "loss": 0.4845,
      "mean_token_accuracy": 0.8538484081625939,
      "num_tokens": 176309108.0,
      "step": 146550
    },
    {
      "entropy": 1.8652475848793983,
      "epoch": 0.45432317672833683,
      "grad_norm": 8.39297866821289,
      "learning_rate": 3.7532864848684496e-06,
      "loss": 0.4237,
      "mean_token_accuracy": 0.8631422132253647,
      "num_tokens": 176322014.0,
      "step": 146560
    },
    {
      "entropy": 1.9841240167617797,
      "epoch": 0.45435417585338655,
      "grad_norm": 9.03348159790039,
      "learning_rate": 3.7531584444742956e-06,
      "loss": 0.4887,
      "mean_token_accuracy": 0.8504109963774681,
      "num_tokens": 176332381.0,
      "step": 146570
    },
    {
      "entropy": 1.8902300730347634,
      "epoch": 0.4543851749784362,
      "grad_norm": 3.9884934425354004,
      "learning_rate": 3.753030417183237e-06,
      "loss": 0.4038,
      "mean_token_accuracy": 0.8569736734032631,
      "num_tokens": 176345267.0,
      "step": 146580
    },
    {
      "entropy": 1.8768309980630875,
      "epoch": 0.45441617410348595,
      "grad_norm": 8.625255584716797,
      "learning_rate": 3.75290240299304e-06,
      "loss": 0.4368,
      "mean_token_accuracy": 0.8578463882207871,
      "num_tokens": 176357705.0,
      "step": 146590
    },
    {
      "entropy": 1.9548759251832961,
      "epoch": 0.4544471732285356,
      "grad_norm": 4.450130462646484,
      "learning_rate": 3.7527744019014693e-06,
      "loss": 0.4713,
      "mean_token_accuracy": 0.855126628279686,
      "num_tokens": 176369275.0,
      "step": 146600
    },
    {
      "entropy": 1.8565620198845862,
      "epoch": 0.45447817235358534,
      "grad_norm": 8.501686096191406,
      "learning_rate": 3.7526464139062918e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.863280688226223,
      "num_tokens": 176382234.0,
      "step": 146610
    },
    {
      "entropy": 1.8947626411914826,
      "epoch": 0.454509171478635,
      "grad_norm": 9.0901460647583,
      "learning_rate": 3.752518439005274e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8616373479366303,
      "num_tokens": 176394636.0,
      "step": 146620
    },
    {
      "entropy": 1.9476829752326013,
      "epoch": 0.45454017060368473,
      "grad_norm": 8.009326934814453,
      "learning_rate": 3.752390477196185e-06,
      "loss": 0.4846,
      "mean_token_accuracy": 0.8480390936136246,
      "num_tokens": 176406091.0,
      "step": 146630
    },
    {
      "entropy": 1.9551826044917107,
      "epoch": 0.4545711697287344,
      "grad_norm": 3.430755853652954,
      "learning_rate": 3.752262528476791e-06,
      "loss": 0.5021,
      "mean_token_accuracy": 0.8439038008451462,
      "num_tokens": 176417335.0,
      "step": 146640
    },
    {
      "entropy": 1.9948366075754165,
      "epoch": 0.4546021688537841,
      "grad_norm": 6.840229034423828,
      "learning_rate": 3.752134592844861e-06,
      "loss": 0.5027,
      "mean_token_accuracy": 0.8471700206398964,
      "num_tokens": 176428379.0,
      "step": 146650
    },
    {
      "entropy": 1.964817936718464,
      "epoch": 0.4546331679788338,
      "grad_norm": 4.002964496612549,
      "learning_rate": 3.7520066702981637e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8483250498771667,
      "num_tokens": 176439775.0,
      "step": 146660
    },
    {
      "entropy": 1.9319669753313065,
      "epoch": 0.45466416710388347,
      "grad_norm": 4.155783653259277,
      "learning_rate": 3.7518787608344694e-06,
      "loss": 0.4414,
      "mean_token_accuracy": 0.8619429409503937,
      "num_tokens": 176451549.0,
      "step": 146670
    },
    {
      "entropy": 1.9431686520576477,
      "epoch": 0.4546951662289332,
      "grad_norm": 8.80432415008545,
      "learning_rate": 3.7517508644515476e-06,
      "loss": 0.4715,
      "mean_token_accuracy": 0.8515708222985268,
      "num_tokens": 176463221.0,
      "step": 146680
    },
    {
      "entropy": 1.912970869243145,
      "epoch": 0.45472616535398286,
      "grad_norm": 9.149775505065918,
      "learning_rate": 3.7516229811471686e-06,
      "loss": 0.4347,
      "mean_token_accuracy": 0.8584818720817566,
      "num_tokens": 176475849.0,
      "step": 146690
    },
    {
      "entropy": 1.948892466723919,
      "epoch": 0.4547571644790326,
      "grad_norm": 8.775398254394531,
      "learning_rate": 3.751495110919105e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8604862332344055,
      "num_tokens": 176487394.0,
      "step": 146700
    },
    {
      "entropy": 1.879041202366352,
      "epoch": 0.45478816360408225,
      "grad_norm": 8.988895416259766,
      "learning_rate": 3.7513672537651273e-06,
      "loss": 0.4596,
      "mean_token_accuracy": 0.8561202257871627,
      "num_tokens": 176499781.0,
      "step": 146710
    },
    {
      "entropy": 1.9582669615745545,
      "epoch": 0.454819162729132,
      "grad_norm": 10.033254623413086,
      "learning_rate": 3.751239409683008e-06,
      "loss": 0.5125,
      "mean_token_accuracy": 0.8357150718569756,
      "num_tokens": 176512409.0,
      "step": 146720
    },
    {
      "entropy": 1.9103184998035432,
      "epoch": 0.45485016185418164,
      "grad_norm": 8.347360610961914,
      "learning_rate": 3.75111157867052e-06,
      "loss": 0.4322,
      "mean_token_accuracy": 0.8483709633350373,
      "num_tokens": 176524784.0,
      "step": 146730
    },
    {
      "entropy": 1.9160448759794235,
      "epoch": 0.45488116097923137,
      "grad_norm": 4.41226863861084,
      "learning_rate": 3.7509837607254356e-06,
      "loss": 0.4349,
      "mean_token_accuracy": 0.8565223976969719,
      "num_tokens": 176536691.0,
      "step": 146740
    },
    {
      "entropy": 1.8885368049144744,
      "epoch": 0.45491216010428104,
      "grad_norm": 7.640499591827393,
      "learning_rate": 3.75085595584553e-06,
      "loss": 0.4101,
      "mean_token_accuracy": 0.8599068447947502,
      "num_tokens": 176549077.0,
      "step": 146750
    },
    {
      "entropy": 1.9121224999427795,
      "epoch": 0.45494315922933076,
      "grad_norm": 8.662951469421387,
      "learning_rate": 3.750728164028577e-06,
      "loss": 0.4483,
      "mean_token_accuracy": 0.8458693385124206,
      "num_tokens": 176561257.0,
      "step": 146760
    },
    {
      "entropy": 1.8658764064311981,
      "epoch": 0.45497415835438043,
      "grad_norm": 8.34899616241455,
      "learning_rate": 3.7506003852723517e-06,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8492805927991867,
      "num_tokens": 176574547.0,
      "step": 146770
    },
    {
      "entropy": 2.010634405910969,
      "epoch": 0.45500515747943016,
      "grad_norm": 4.921254634857178,
      "learning_rate": 3.7504726195746287e-06,
      "loss": 0.4775,
      "mean_token_accuracy": 0.8483109161257744,
      "num_tokens": 176585789.0,
      "step": 146780
    },
    {
      "entropy": 2.0231185287237166,
      "epoch": 0.4550361566044798,
      "grad_norm": 7.859384536743164,
      "learning_rate": 3.750344866933185e-06,
      "loss": 0.482,
      "mean_token_accuracy": 0.854818707704544,
      "num_tokens": 176596403.0,
      "step": 146790
    },
    {
      "entropy": 1.809226544201374,
      "epoch": 0.45506715572952955,
      "grad_norm": 7.986112594604492,
      "learning_rate": 3.750217127345796e-06,
      "loss": 0.3894,
      "mean_token_accuracy": 0.8633719727396965,
      "num_tokens": 176610424.0,
      "step": 146800
    },
    {
      "entropy": 1.9953494429588319,
      "epoch": 0.4550981548545792,
      "grad_norm": 8.195384979248047,
      "learning_rate": 3.7500894008102395e-06,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8488559857010841,
      "num_tokens": 176621493.0,
      "step": 146810
    },
    {
      "entropy": 1.7815378695726394,
      "epoch": 0.45512915397962894,
      "grad_norm": 6.289517402648926,
      "learning_rate": 3.7499616873242926e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.8637578442692757,
      "num_tokens": 176635418.0,
      "step": 146820
    },
    {
      "entropy": 1.9626459717750548,
      "epoch": 0.4551601531046786,
      "grad_norm": 8.439818382263184,
      "learning_rate": 3.7498339868857342e-06,
      "loss": 0.508,
      "mean_token_accuracy": 0.852840892970562,
      "num_tokens": 176646632.0,
      "step": 146830
    },
    {
      "entropy": 1.906586892902851,
      "epoch": 0.45519115222972834,
      "grad_norm": 8.007729530334473,
      "learning_rate": 3.749706299492342e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8653773680329323,
      "num_tokens": 176657728.0,
      "step": 146840
    },
    {
      "entropy": 1.9015120595693589,
      "epoch": 0.455222151354778,
      "grad_norm": 7.211781978607178,
      "learning_rate": 3.749578625141895e-06,
      "loss": 0.4305,
      "mean_token_accuracy": 0.8567000299692153,
      "num_tokens": 176670194.0,
      "step": 146850
    },
    {
      "entropy": 1.9847611114382744,
      "epoch": 0.45525315047982773,
      "grad_norm": 6.767532825469971,
      "learning_rate": 3.7494509638321734e-06,
      "loss": 0.4531,
      "mean_token_accuracy": 0.8509671688079834,
      "num_tokens": 176681480.0,
      "step": 146860
    },
    {
      "entropy": 1.8333649218082428,
      "epoch": 0.4552841496048774,
      "grad_norm": 7.665855407714844,
      "learning_rate": 3.749323315560957e-06,
      "loss": 0.4189,
      "mean_token_accuracy": 0.856324379146099,
      "num_tokens": 176694603.0,
      "step": 146870
    },
    {
      "entropy": 1.9811863422393798,
      "epoch": 0.4553151487299271,
      "grad_norm": 8.544697761535645,
      "learning_rate": 3.7491956803260273e-06,
      "loss": 0.4924,
      "mean_token_accuracy": 0.8474650055170059,
      "num_tokens": 176705780.0,
      "step": 146880
    },
    {
      "entropy": 1.8731940254569053,
      "epoch": 0.4553461478549768,
      "grad_norm": 8.085098266601562,
      "learning_rate": 3.7490680581251637e-06,
      "loss": 0.4266,
      "mean_token_accuracy": 0.8562427669763565,
      "num_tokens": 176717194.0,
      "step": 146890
    },
    {
      "entropy": 1.862533800303936,
      "epoch": 0.4553771469800265,
      "grad_norm": 4.403972625732422,
      "learning_rate": 3.74894044895615e-06,
      "loss": 0.4545,
      "mean_token_accuracy": 0.8544600263237954,
      "num_tokens": 176730041.0,
      "step": 146900
    },
    {
      "entropy": 1.8781135827302933,
      "epoch": 0.4554081461050762,
      "grad_norm": 3.9804108142852783,
      "learning_rate": 3.7488128528167672e-06,
      "loss": 0.428,
      "mean_token_accuracy": 0.8630628094077111,
      "num_tokens": 176742512.0,
      "step": 146910
    },
    {
      "entropy": 1.9072934925556182,
      "epoch": 0.45543914523012585,
      "grad_norm": 7.206721305847168,
      "learning_rate": 3.7486852697047988e-06,
      "loss": 0.4205,
      "mean_token_accuracy": 0.8565845042467117,
      "num_tokens": 176754331.0,
      "step": 146920
    },
    {
      "entropy": 1.9471127331256866,
      "epoch": 0.4554701443551756,
      "grad_norm": 6.824718952178955,
      "learning_rate": 3.748557699618028e-06,
      "loss": 0.4853,
      "mean_token_accuracy": 0.8565745607018471,
      "num_tokens": 176765360.0,
      "step": 146930
    },
    {
      "entropy": 1.888777793943882,
      "epoch": 0.45550114348022525,
      "grad_norm": 8.23228645324707,
      "learning_rate": 3.748430142554238e-06,
      "loss": 0.4229,
      "mean_token_accuracy": 0.865308640897274,
      "num_tokens": 176777052.0,
      "step": 146940
    },
    {
      "entropy": 1.881396123766899,
      "epoch": 0.45553214260527497,
      "grad_norm": 7.634214878082275,
      "learning_rate": 3.748302598511214e-06,
      "loss": 0.4877,
      "mean_token_accuracy": 0.8448997780680656,
      "num_tokens": 176789075.0,
      "step": 146950
    },
    {
      "entropy": 1.888343983888626,
      "epoch": 0.45556314173032464,
      "grad_norm": 8.560070991516113,
      "learning_rate": 3.748175067486742e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8476789712905883,
      "num_tokens": 176801688.0,
      "step": 146960
    },
    {
      "entropy": 1.8695935264229775,
      "epoch": 0.45559414085537436,
      "grad_norm": 4.059643268585205,
      "learning_rate": 3.7480475494786045e-06,
      "loss": 0.3986,
      "mean_token_accuracy": 0.8586703911423683,
      "num_tokens": 176814689.0,
      "step": 146970
    },
    {
      "entropy": 1.8886354252696038,
      "epoch": 0.45562513998042403,
      "grad_norm": 3.8732738494873047,
      "learning_rate": 3.7479200444845893e-06,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8417670831084252,
      "num_tokens": 176827493.0,
      "step": 146980
    },
    {
      "entropy": 1.8563087373971938,
      "epoch": 0.45565613910547376,
      "grad_norm": 8.856595993041992,
      "learning_rate": 3.7477925525024837e-06,
      "loss": 0.4056,
      "mean_token_accuracy": 0.8550028428435326,
      "num_tokens": 176840753.0,
      "step": 146990
    },
    {
      "entropy": 1.941939078271389,
      "epoch": 0.4556871382305234,
      "grad_norm": 4.509472846984863,
      "learning_rate": 3.7476650735300728e-06,
      "loss": 0.4409,
      "mean_token_accuracy": 0.8547850877046586,
      "num_tokens": 176852241.0,
      "step": 147000
    },
    {
      "entropy": 1.9420098468661309,
      "epoch": 0.45571813735557315,
      "grad_norm": 4.272128582000732,
      "learning_rate": 3.747537607565146e-06,
      "loss": 0.4371,
      "mean_token_accuracy": 0.8548914834856987,
      "num_tokens": 176863804.0,
      "step": 147010
    },
    {
      "entropy": 1.897740714251995,
      "epoch": 0.4557491364806228,
      "grad_norm": 9.78155517578125,
      "learning_rate": 3.7474101546054897e-06,
      "loss": 0.4496,
      "mean_token_accuracy": 0.8575510829687119,
      "num_tokens": 176876007.0,
      "step": 147020
    },
    {
      "entropy": 1.901134905219078,
      "epoch": 0.45578013560567254,
      "grad_norm": 8.032532691955566,
      "learning_rate": 3.747282714648894e-06,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8620973736047745,
      "num_tokens": 176887951.0,
      "step": 147030
    },
    {
      "entropy": 1.9228925719857215,
      "epoch": 0.4558111347307222,
      "grad_norm": 3.396730422973633,
      "learning_rate": 3.747155287693148e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8453203931450843,
      "num_tokens": 176899964.0,
      "step": 147040
    },
    {
      "entropy": 1.9790107786655426,
      "epoch": 0.45584213385577194,
      "grad_norm": 6.975090026855469,
      "learning_rate": 3.7470278737360395e-06,
      "loss": 0.4874,
      "mean_token_accuracy": 0.8480637580156326,
      "num_tokens": 176911678.0,
      "step": 147050
    },
    {
      "entropy": 1.8822043985128403,
      "epoch": 0.4558731329808216,
      "grad_norm": 9.182622909545898,
      "learning_rate": 3.7469004727753605e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8651730179786682,
      "num_tokens": 176923300.0,
      "step": 147060
    },
    {
      "entropy": 1.9158610820770263,
      "epoch": 0.45590413210587133,
      "grad_norm": 3.5178349018096924,
      "learning_rate": 3.746773084808901e-06,
      "loss": 0.4353,
      "mean_token_accuracy": 0.8595368161797523,
      "num_tokens": 176935060.0,
      "step": 147070
    },
    {
      "entropy": 1.968504549562931,
      "epoch": 0.455935131230921,
      "grad_norm": 7.308728218078613,
      "learning_rate": 3.7466457098344528e-06,
      "loss": 0.4946,
      "mean_token_accuracy": 0.8455461367964745,
      "num_tokens": 176946490.0,
      "step": 147080
    },
    {
      "entropy": 1.8836671754717826,
      "epoch": 0.4559661303559707,
      "grad_norm": 8.155224800109863,
      "learning_rate": 3.7465183478498068e-06,
      "loss": 0.4312,
      "mean_token_accuracy": 0.8532806396484375,
      "num_tokens": 176959124.0,
      "step": 147090
    },
    {
      "entropy": 1.9047341987490654,
      "epoch": 0.4559971294810204,
      "grad_norm": 7.964391708374023,
      "learning_rate": 3.7463909988527563e-06,
      "loss": 0.4336,
      "mean_token_accuracy": 0.8509239420294762,
      "num_tokens": 176971598.0,
      "step": 147100
    },
    {
      "entropy": 1.9551147490739822,
      "epoch": 0.4560281286060701,
      "grad_norm": 7.997689723968506,
      "learning_rate": 3.7462636628410933e-06,
      "loss": 0.5301,
      "mean_token_accuracy": 0.8321620270609855,
      "num_tokens": 176982580.0,
      "step": 147110
    },
    {
      "entropy": 1.898515647649765,
      "epoch": 0.4560591277311198,
      "grad_norm": 2.4085474014282227,
      "learning_rate": 3.7461363398126123e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8540950760245323,
      "num_tokens": 176995367.0,
      "step": 147120
    },
    {
      "entropy": 1.841650950908661,
      "epoch": 0.4560901268561695,
      "grad_norm": 7.601720333099365,
      "learning_rate": 3.746009029765105e-06,
      "loss": 0.3595,
      "mean_token_accuracy": 0.8667580097913742,
      "num_tokens": 177008617.0,
      "step": 147130
    },
    {
      "entropy": 1.8938813239336014,
      "epoch": 0.4561211259812192,
      "grad_norm": 8.31351375579834,
      "learning_rate": 3.745881732696369e-06,
      "loss": 0.4297,
      "mean_token_accuracy": 0.8541634380817413,
      "num_tokens": 177021433.0,
      "step": 147140
    },
    {
      "entropy": 1.9346230700612068,
      "epoch": 0.4561521251062689,
      "grad_norm": 8.50265884399414,
      "learning_rate": 3.745754448604195e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8484914928674698,
      "num_tokens": 177033364.0,
      "step": 147150
    },
    {
      "entropy": 1.8723743125796317,
      "epoch": 0.45618312423131857,
      "grad_norm": 3.476085901260376,
      "learning_rate": 3.745627177486383e-06,
      "loss": 0.4031,
      "mean_token_accuracy": 0.8580270186066628,
      "num_tokens": 177046006.0,
      "step": 147160
    },
    {
      "entropy": 1.8366812959313392,
      "epoch": 0.45621412335636824,
      "grad_norm": 3.9904587268829346,
      "learning_rate": 3.745499919340726e-06,
      "loss": 0.3737,
      "mean_token_accuracy": 0.8696741983294487,
      "num_tokens": 177059038.0,
      "step": 147170
    },
    {
      "entropy": 1.8440161734819411,
      "epoch": 0.45624512248141796,
      "grad_norm": 9.400425910949707,
      "learning_rate": 3.745372674165021e-06,
      "loss": 0.3684,
      "mean_token_accuracy": 0.8613582566380501,
      "num_tokens": 177072384.0,
      "step": 147180
    },
    {
      "entropy": 1.9192558169364928,
      "epoch": 0.45627612160646763,
      "grad_norm": 10.3735933303833,
      "learning_rate": 3.7452454419570656e-06,
      "loss": 0.4259,
      "mean_token_accuracy": 0.8616964593529701,
      "num_tokens": 177084461.0,
      "step": 147190
    },
    {
      "entropy": 1.8800915464758874,
      "epoch": 0.45630712073151736,
      "grad_norm": 3.897484302520752,
      "learning_rate": 3.745118222714657e-06,
      "loss": 0.4315,
      "mean_token_accuracy": 0.8533887133002281,
      "num_tokens": 177096842.0,
      "step": 147200
    },
    {
      "entropy": 1.945913690328598,
      "epoch": 0.456338119856567,
      "grad_norm": 2.707284688949585,
      "learning_rate": 3.7449910164355936e-06,
      "loss": 0.4759,
      "mean_token_accuracy": 0.847554586827755,
      "num_tokens": 177109235.0,
      "step": 147210
    },
    {
      "entropy": 1.9585506185889243,
      "epoch": 0.45636911898161675,
      "grad_norm": 7.910290718078613,
      "learning_rate": 3.7448638231176737e-06,
      "loss": 0.4517,
      "mean_token_accuracy": 0.8551441594958306,
      "num_tokens": 177120779.0,
      "step": 147220
    },
    {
      "entropy": 1.8726617515087127,
      "epoch": 0.4564001181066664,
      "grad_norm": 3.750462770462036,
      "learning_rate": 3.7447366427586964e-06,
      "loss": 0.4599,
      "mean_token_accuracy": 0.8569039270281792,
      "num_tokens": 177133126.0,
      "step": 147230
    },
    {
      "entropy": 1.91754612326622,
      "epoch": 0.45643111723171614,
      "grad_norm": 7.520792007446289,
      "learning_rate": 3.7446094753564614e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8505095258355141,
      "num_tokens": 177145098.0,
      "step": 147240
    },
    {
      "entropy": 1.9441439002752303,
      "epoch": 0.4564621163567658,
      "grad_norm": 10.833855628967285,
      "learning_rate": 3.7444823209087682e-06,
      "loss": 0.4681,
      "mean_token_accuracy": 0.8531385481357574,
      "num_tokens": 177156290.0,
      "step": 147250
    },
    {
      "entropy": 1.912196746468544,
      "epoch": 0.45649311548181554,
      "grad_norm": 9.136136054992676,
      "learning_rate": 3.744355179413419e-06,
      "loss": 0.4201,
      "mean_token_accuracy": 0.8659820929169655,
      "num_tokens": 177167879.0,
      "step": 147260
    },
    {
      "entropy": 1.7876418299973011,
      "epoch": 0.4565241146068652,
      "grad_norm": 7.6198039054870605,
      "learning_rate": 3.7442280508682134e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8576739236712456,
      "num_tokens": 177181505.0,
      "step": 147270
    },
    {
      "entropy": 1.925425273180008,
      "epoch": 0.45655511373191493,
      "grad_norm": 7.062708377838135,
      "learning_rate": 3.7441009352709544e-06,
      "loss": 0.4873,
      "mean_token_accuracy": 0.8492991134524346,
      "num_tokens": 177193315.0,
      "step": 147280
    },
    {
      "entropy": 1.847182884812355,
      "epoch": 0.4565861128569646,
      "grad_norm": 7.065945625305176,
      "learning_rate": 3.7439738326194437e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8619561657309532,
      "num_tokens": 177205609.0,
      "step": 147290
    },
    {
      "entropy": 1.8271872013807298,
      "epoch": 0.4566171119820143,
      "grad_norm": 4.04852294921875,
      "learning_rate": 3.7438467429114837e-06,
      "loss": 0.3711,
      "mean_token_accuracy": 0.8713533550500869,
      "num_tokens": 177219174.0,
      "step": 147300
    },
    {
      "entropy": 1.9408521130681038,
      "epoch": 0.456648111107064,
      "grad_norm": 8.535037994384766,
      "learning_rate": 3.743719666144879e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.843614149093628,
      "num_tokens": 177230752.0,
      "step": 147310
    },
    {
      "entropy": 1.8946807518601418,
      "epoch": 0.4566791102321137,
      "grad_norm": 3.0620830059051514,
      "learning_rate": 3.743592602317431e-06,
      "loss": 0.4103,
      "mean_token_accuracy": 0.8616632342338562,
      "num_tokens": 177243742.0,
      "step": 147320
    },
    {
      "entropy": 1.8989785492420197,
      "epoch": 0.4567101093571634,
      "grad_norm": 7.388391017913818,
      "learning_rate": 3.743465551426947e-06,
      "loss": 0.4324,
      "mean_token_accuracy": 0.8464514210820198,
      "num_tokens": 177256265.0,
      "step": 147330
    },
    {
      "entropy": 1.7777711182832718,
      "epoch": 0.4567411084822131,
      "grad_norm": 8.797972679138184,
      "learning_rate": 3.7433385134712295e-06,
      "loss": 0.4015,
      "mean_token_accuracy": 0.8580346912145614,
      "num_tokens": 177270945.0,
      "step": 147340
    },
    {
      "entropy": 2.003285530209541,
      "epoch": 0.4567721076072628,
      "grad_norm": 8.642237663269043,
      "learning_rate": 3.7432114884480853e-06,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8421431943774224,
      "num_tokens": 177281633.0,
      "step": 147350
    },
    {
      "entropy": 1.9756213307380677,
      "epoch": 0.4568031067323125,
      "grad_norm": 9.405720710754395,
      "learning_rate": 3.743084476355319e-06,
      "loss": 0.4912,
      "mean_token_accuracy": 0.8499099537730217,
      "num_tokens": 177292337.0,
      "step": 147360
    },
    {
      "entropy": 1.8737393349409104,
      "epoch": 0.45683410585736217,
      "grad_norm": 6.993408203125,
      "learning_rate": 3.742957477190739e-06,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8587880119681358,
      "num_tokens": 177304763.0,
      "step": 147370
    },
    {
      "entropy": 1.934411568939686,
      "epoch": 0.4568651049824119,
      "grad_norm": 8.104722023010254,
      "learning_rate": 3.74283049095215e-06,
      "loss": 0.4842,
      "mean_token_accuracy": 0.8433166891336441,
      "num_tokens": 177316276.0,
      "step": 147380
    },
    {
      "entropy": 1.9655498325824738,
      "epoch": 0.45689610410746156,
      "grad_norm": 8.411802291870117,
      "learning_rate": 3.742703517637361e-06,
      "loss": 0.4673,
      "mean_token_accuracy": 0.8529037058353424,
      "num_tokens": 177327855.0,
      "step": 147390
    },
    {
      "entropy": 1.848972088098526,
      "epoch": 0.4569271032325113,
      "grad_norm": 8.530176162719727,
      "learning_rate": 3.74257655724418e-06,
      "loss": 0.4828,
      "mean_token_accuracy": 0.8518965750932693,
      "num_tokens": 177340947.0,
      "step": 147400
    },
    {
      "entropy": 1.8785428568720817,
      "epoch": 0.45695810235756096,
      "grad_norm": 8.014692306518555,
      "learning_rate": 3.742449609770415e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8454515695571899,
      "num_tokens": 177352583.0,
      "step": 147410
    },
    {
      "entropy": 1.9009640589356422,
      "epoch": 0.4569891014826106,
      "grad_norm": 8.411507606506348,
      "learning_rate": 3.7423226752138736e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8564955353736877,
      "num_tokens": 177364796.0,
      "step": 147420
    },
    {
      "entropy": 1.8695711359381675,
      "epoch": 0.45702010060766035,
      "grad_norm": 8.759578704833984,
      "learning_rate": 3.7421957535723686e-06,
      "loss": 0.4138,
      "mean_token_accuracy": 0.8556355476379395,
      "num_tokens": 177377130.0,
      "step": 147430
    },
    {
      "entropy": 1.882589338719845,
      "epoch": 0.45705109973271,
      "grad_norm": 5.764090538024902,
      "learning_rate": 3.7420688448437077e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8518973112106323,
      "num_tokens": 177389852.0,
      "step": 147440
    },
    {
      "entropy": 1.9209093794226646,
      "epoch": 0.45708209885775974,
      "grad_norm": 9.359416007995605,
      "learning_rate": 3.7419419490257012e-06,
      "loss": 0.475,
      "mean_token_accuracy": 0.8516506165266037,
      "num_tokens": 177401433.0,
      "step": 147450
    },
    {
      "entropy": 1.86415656208992,
      "epoch": 0.4571130979828094,
      "grad_norm": 8.562514305114746,
      "learning_rate": 3.741815066116162e-06,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8509325221180916,
      "num_tokens": 177414478.0,
      "step": 147460
    },
    {
      "entropy": 1.745816995203495,
      "epoch": 0.45714409710785914,
      "grad_norm": 6.630068302154541,
      "learning_rate": 3.7416881961129002e-06,
      "loss": 0.381,
      "mean_token_accuracy": 0.8737103626132011,
      "num_tokens": 177428315.0,
      "step": 147470
    },
    {
      "entropy": 1.8828457549214364,
      "epoch": 0.4571750962329088,
      "grad_norm": 4.396876335144043,
      "learning_rate": 3.7415613390137295e-06,
      "loss": 0.442,
      "mean_token_accuracy": 0.8575210615992546,
      "num_tokens": 177440392.0,
      "step": 147480
    },
    {
      "entropy": 1.9914973288774491,
      "epoch": 0.45720609535795853,
      "grad_norm": 6.813479423522949,
      "learning_rate": 3.7414344948164604e-06,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8524721592664719,
      "num_tokens": 177451268.0,
      "step": 147490
    },
    {
      "entropy": 1.8895151317119598,
      "epoch": 0.4572370944830082,
      "grad_norm": 3.594588279724121,
      "learning_rate": 3.741307663518908e-06,
      "loss": 0.4189,
      "mean_token_accuracy": 0.8583090662956238,
      "num_tokens": 177463489.0,
      "step": 147500
    },
    {
      "entropy": 1.9084604054689407,
      "epoch": 0.4572680936080579,
      "grad_norm": 4.002728462219238,
      "learning_rate": 3.741180845118885e-06,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8463861241936683,
      "num_tokens": 177475847.0,
      "step": 147510
    },
    {
      "entropy": 1.9138479217886926,
      "epoch": 0.4572990927331076,
      "grad_norm": 9.18737506866455,
      "learning_rate": 3.7410540396142063e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8641066178679466,
      "num_tokens": 177487582.0,
      "step": 147520
    },
    {
      "entropy": 1.87220705896616,
      "epoch": 0.4573300918581573,
      "grad_norm": 7.512433052062988,
      "learning_rate": 3.740927247002686e-06,
      "loss": 0.4415,
      "mean_token_accuracy": 0.858481977880001,
      "num_tokens": 177499543.0,
      "step": 147530
    },
    {
      "entropy": 1.9340161770582198,
      "epoch": 0.457361090983207,
      "grad_norm": 7.679035186767578,
      "learning_rate": 3.74080046728214e-06,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8467412024736405,
      "num_tokens": 177510615.0,
      "step": 147540
    },
    {
      "entropy": 1.8590017393231393,
      "epoch": 0.4573920901082567,
      "grad_norm": 9.347676277160645,
      "learning_rate": 3.7406737004503834e-06,
      "loss": 0.429,
      "mean_token_accuracy": 0.8628332495689393,
      "num_tokens": 177522638.0,
      "step": 147550
    },
    {
      "entropy": 1.911301201581955,
      "epoch": 0.4574230892333064,
      "grad_norm": 8.779614448547363,
      "learning_rate": 3.740546946505233e-06,
      "loss": 0.4369,
      "mean_token_accuracy": 0.8550707563757897,
      "num_tokens": 177534749.0,
      "step": 147560
    },
    {
      "entropy": 1.9754537165164947,
      "epoch": 0.4574540883583561,
      "grad_norm": 8.753040313720703,
      "learning_rate": 3.740420205444505e-06,
      "loss": 0.4771,
      "mean_token_accuracy": 0.853889499604702,
      "num_tokens": 177546067.0,
      "step": 147570
    },
    {
      "entropy": 1.9112480387091637,
      "epoch": 0.4574850874834058,
      "grad_norm": 4.434845447540283,
      "learning_rate": 3.740293477266017e-06,
      "loss": 0.4801,
      "mean_token_accuracy": 0.8533158525824547,
      "num_tokens": 177557969.0,
      "step": 147580
    },
    {
      "entropy": 1.877299964427948,
      "epoch": 0.4575160866084555,
      "grad_norm": 7.680180072784424,
      "learning_rate": 3.7401667619675876e-06,
      "loss": 0.3953,
      "mean_token_accuracy": 0.8607863545417785,
      "num_tokens": 177570545.0,
      "step": 147590
    },
    {
      "entropy": 1.9265247106552124,
      "epoch": 0.45754708573350517,
      "grad_norm": 8.468583106994629,
      "learning_rate": 3.7400400595470337e-06,
      "loss": 0.4765,
      "mean_token_accuracy": 0.8503778129816055,
      "num_tokens": 177582395.0,
      "step": 147600
    },
    {
      "entropy": 1.8167890459299088,
      "epoch": 0.4575780848585549,
      "grad_norm": 8.333720207214355,
      "learning_rate": 3.7399133700021756e-06,
      "loss": 0.4504,
      "mean_token_accuracy": 0.8408757716417312,
      "num_tokens": 177596261.0,
      "step": 147610
    },
    {
      "entropy": 1.8485788330435753,
      "epoch": 0.45760908398360456,
      "grad_norm": 7.62937593460083,
      "learning_rate": 3.7397866933308325e-06,
      "loss": 0.4469,
      "mean_token_accuracy": 0.8608444139361382,
      "num_tokens": 177608763.0,
      "step": 147620
    },
    {
      "entropy": 1.9072691813111304,
      "epoch": 0.4576400831086543,
      "grad_norm": 9.699867248535156,
      "learning_rate": 3.7396600295308234e-06,
      "loss": 0.447,
      "mean_token_accuracy": 0.8526267111301422,
      "num_tokens": 177621101.0,
      "step": 147630
    },
    {
      "entropy": 1.9000159114599229,
      "epoch": 0.45767108223370395,
      "grad_norm": 7.382110118865967,
      "learning_rate": 3.7395333785999692e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8545347273349762,
      "num_tokens": 177633245.0,
      "step": 147640
    },
    {
      "entropy": 1.9029620870947839,
      "epoch": 0.4577020813587537,
      "grad_norm": 9.210521697998047,
      "learning_rate": 3.739406740536092e-06,
      "loss": 0.4738,
      "mean_token_accuracy": 0.8504366457462311,
      "num_tokens": 177643987.0,
      "step": 147650
    },
    {
      "entropy": 1.9046365901827813,
      "epoch": 0.45773308048380335,
      "grad_norm": 4.034806251525879,
      "learning_rate": 3.739280115337011e-06,
      "loss": 0.4549,
      "mean_token_accuracy": 0.8487413614988327,
      "num_tokens": 177655795.0,
      "step": 147660
    },
    {
      "entropy": 1.8236022911965848,
      "epoch": 0.457764079608853,
      "grad_norm": 4.0919623374938965,
      "learning_rate": 3.73915350300055e-06,
      "loss": 0.3661,
      "mean_token_accuracy": 0.8643067702651024,
      "num_tokens": 177668862.0,
      "step": 147670
    },
    {
      "entropy": 1.9250768944621086,
      "epoch": 0.45779507873390274,
      "grad_norm": 8.812362670898438,
      "learning_rate": 3.7390269035245302e-06,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8513668492436409,
      "num_tokens": 177681125.0,
      "step": 147680
    },
    {
      "entropy": 1.8930511608719827,
      "epoch": 0.4578260778589524,
      "grad_norm": 7.733847141265869,
      "learning_rate": 3.738900316906776e-06,
      "loss": 0.4254,
      "mean_token_accuracy": 0.8619498953223228,
      "num_tokens": 177693301.0,
      "step": 147690
    },
    {
      "entropy": 1.9794332563877106,
      "epoch": 0.45785707698400213,
      "grad_norm": 7.670042514801025,
      "learning_rate": 3.7387737431451097e-06,
      "loss": 0.504,
      "mean_token_accuracy": 0.8443345949053764,
      "num_tokens": 177704344.0,
      "step": 147700
    },
    {
      "entropy": 1.9156940042972566,
      "epoch": 0.4578880761090518,
      "grad_norm": 7.951246738433838,
      "learning_rate": 3.738647182237357e-06,
      "loss": 0.4712,
      "mean_token_accuracy": 0.8424468949437142,
      "num_tokens": 177716411.0,
      "step": 147710
    },
    {
      "entropy": 1.8918796986341477,
      "epoch": 0.4579190752341015,
      "grad_norm": 6.88615083694458,
      "learning_rate": 3.738520634181341e-06,
      "loss": 0.414,
      "mean_token_accuracy": 0.861275652050972,
      "num_tokens": 177728341.0,
      "step": 147720
    },
    {
      "entropy": 1.849881935119629,
      "epoch": 0.4579500743591512,
      "grad_norm": 3.4234378337860107,
      "learning_rate": 3.738394098974886e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8636976391077041,
      "num_tokens": 177740803.0,
      "step": 147730
    },
    {
      "entropy": 1.8762715697288512,
      "epoch": 0.4579810734842009,
      "grad_norm": 8.60593318939209,
      "learning_rate": 3.7382675766158196e-06,
      "loss": 0.4829,
      "mean_token_accuracy": 0.8438662976026535,
      "num_tokens": 177752757.0,
      "step": 147740
    },
    {
      "entropy": 1.9411585584282876,
      "epoch": 0.4580120726092506,
      "grad_norm": 9.615153312683105,
      "learning_rate": 3.738141067101967e-06,
      "loss": 0.4989,
      "mean_token_accuracy": 0.8427041709423065,
      "num_tokens": 177764050.0,
      "step": 147750
    },
    {
      "entropy": 1.8888093829154968,
      "epoch": 0.4580430717343003,
      "grad_norm": 8.258987426757812,
      "learning_rate": 3.7380145704311548e-06,
      "loss": 0.4782,
      "mean_token_accuracy": 0.85269236266613,
      "num_tokens": 177775788.0,
      "step": 147760
    },
    {
      "entropy": 1.857784403860569,
      "epoch": 0.45807407085935,
      "grad_norm": 3.8808376789093018,
      "learning_rate": 3.73788808660121e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8614834606647491,
      "num_tokens": 177787727.0,
      "step": 147770
    },
    {
      "entropy": 1.9264632612466812,
      "epoch": 0.4581050699843997,
      "grad_norm": 8.179401397705078,
      "learning_rate": 3.7377616156099605e-06,
      "loss": 0.4947,
      "mean_token_accuracy": 0.8487723514437675,
      "num_tokens": 177799380.0,
      "step": 147780
    },
    {
      "entropy": 1.9536435410380364,
      "epoch": 0.4581360691094494,
      "grad_norm": 8.652231216430664,
      "learning_rate": 3.737635157455235e-06,
      "loss": 0.5055,
      "mean_token_accuracy": 0.8477381393313408,
      "num_tokens": 177810671.0,
      "step": 147790
    },
    {
      "entropy": 1.8677313759922982,
      "epoch": 0.4581670682344991,
      "grad_norm": 8.161964416503906,
      "learning_rate": 3.7375087121348613e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8476336553692818,
      "num_tokens": 177823218.0,
      "step": 147800
    },
    {
      "entropy": 1.8325920164585114,
      "epoch": 0.45819806735954877,
      "grad_norm": 7.421534538269043,
      "learning_rate": 3.737382279646669e-06,
      "loss": 0.367,
      "mean_token_accuracy": 0.8645919308066368,
      "num_tokens": 177836151.0,
      "step": 147810
    },
    {
      "entropy": 1.9766230046749116,
      "epoch": 0.4582290664845985,
      "grad_norm": 8.750480651855469,
      "learning_rate": 3.7372558599884873e-06,
      "loss": 0.5172,
      "mean_token_accuracy": 0.8399770334362984,
      "num_tokens": 177847098.0,
      "step": 147820
    },
    {
      "entropy": 1.9309228926897049,
      "epoch": 0.45826006560964816,
      "grad_norm": 7.476661205291748,
      "learning_rate": 3.737129453158147e-06,
      "loss": 0.4841,
      "mean_token_accuracy": 0.842421543598175,
      "num_tokens": 177859514.0,
      "step": 147830
    },
    {
      "entropy": 1.847913098335266,
      "epoch": 0.4582910647346979,
      "grad_norm": 6.960815906524658,
      "learning_rate": 3.737003059153479e-06,
      "loss": 0.5784,
      "mean_token_accuracy": 0.8363130848854781,
      "num_tokens": 177873072.0,
      "step": 147840
    },
    {
      "entropy": 1.8798335790634155,
      "epoch": 0.45832206385974755,
      "grad_norm": 7.6284894943237305,
      "learning_rate": 3.7368766779723135e-06,
      "loss": 0.4554,
      "mean_token_accuracy": 0.8519891336560249,
      "num_tokens": 177884925.0,
      "step": 147850
    },
    {
      "entropy": 1.9396116152405738,
      "epoch": 0.4583530629847973,
      "grad_norm": 7.631931304931641,
      "learning_rate": 3.7367503096124836e-06,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8487934455275535,
      "num_tokens": 177896276.0,
      "step": 147860
    },
    {
      "entropy": 1.9332348197698592,
      "epoch": 0.45838406210984695,
      "grad_norm": 9.10966682434082,
      "learning_rate": 3.7366239540718206e-06,
      "loss": 0.4686,
      "mean_token_accuracy": 0.8453480824828148,
      "num_tokens": 177907658.0,
      "step": 147870
    },
    {
      "entropy": 1.847151155769825,
      "epoch": 0.45841506123489667,
      "grad_norm": 7.519127368927002,
      "learning_rate": 3.736497611348158e-06,
      "loss": 0.4284,
      "mean_token_accuracy": 0.8468550607562065,
      "num_tokens": 177919761.0,
      "step": 147880
    },
    {
      "entropy": 1.854358959197998,
      "epoch": 0.45844606035994634,
      "grad_norm": 7.146642684936523,
      "learning_rate": 3.7363712814393296e-06,
      "loss": 0.4346,
      "mean_token_accuracy": 0.8587031260132789,
      "num_tokens": 177931858.0,
      "step": 147890
    },
    {
      "entropy": 1.8741928294301033,
      "epoch": 0.45847705948499606,
      "grad_norm": 8.542024612426758,
      "learning_rate": 3.7362449643431677e-06,
      "loss": 0.449,
      "mean_token_accuracy": 0.8568659141659737,
      "num_tokens": 177943416.0,
      "step": 147900
    },
    {
      "entropy": 1.9269966036081314,
      "epoch": 0.45850805861004573,
      "grad_norm": 5.178137302398682,
      "learning_rate": 3.736118660057506e-06,
      "loss": 0.5051,
      "mean_token_accuracy": 0.8503173589706421,
      "num_tokens": 177954691.0,
      "step": 147910
    },
    {
      "entropy": 1.856522725522518,
      "epoch": 0.4585390577350954,
      "grad_norm": 7.2768330574035645,
      "learning_rate": 3.735992368580183e-06,
      "loss": 0.4206,
      "mean_token_accuracy": 0.8675465360283852,
      "num_tokens": 177966749.0,
      "step": 147920
    },
    {
      "entropy": 1.8835629358887673,
      "epoch": 0.4585700568601451,
      "grad_norm": 3.9212841987609863,
      "learning_rate": 3.73586608990903e-06,
      "loss": 0.4512,
      "mean_token_accuracy": 0.853912553191185,
      "num_tokens": 177978879.0,
      "step": 147930
    },
    {
      "entropy": 1.8709630161523818,
      "epoch": 0.4586010559851948,
      "grad_norm": 9.633736610412598,
      "learning_rate": 3.735739824041885e-06,
      "loss": 0.4579,
      "mean_token_accuracy": 0.8498531699180603,
      "num_tokens": 177990951.0,
      "step": 147940
    },
    {
      "entropy": 1.9850156009197235,
      "epoch": 0.4586320551102445,
      "grad_norm": 8.021490097045898,
      "learning_rate": 3.735613570976584e-06,
      "loss": 0.5309,
      "mean_token_accuracy": 0.8361830487847328,
      "num_tokens": 178001723.0,
      "step": 147950
    },
    {
      "entropy": 1.9277898535132407,
      "epoch": 0.4586630542352942,
      "grad_norm": 8.133795738220215,
      "learning_rate": 3.735487330710964e-06,
      "loss": 0.5754,
      "mean_token_accuracy": 0.8342059582471848,
      "num_tokens": 178013683.0,
      "step": 147960
    },
    {
      "entropy": 1.9039981633424758,
      "epoch": 0.4586940533603439,
      "grad_norm": 9.529333114624023,
      "learning_rate": 3.735361103242862e-06,
      "loss": 0.4942,
      "mean_token_accuracy": 0.8599925115704536,
      "num_tokens": 178025035.0,
      "step": 147970
    },
    {
      "entropy": 1.8207675963640213,
      "epoch": 0.4587250524853936,
      "grad_norm": 4.995032787322998,
      "learning_rate": 3.735234888570117e-06,
      "loss": 0.3791,
      "mean_token_accuracy": 0.8648058965802192,
      "num_tokens": 178038165.0,
      "step": 147980
    },
    {
      "entropy": 1.7604272417724132,
      "epoch": 0.4587560516104433,
      "grad_norm": 7.621956825256348,
      "learning_rate": 3.735108686690566e-06,
      "loss": 0.3362,
      "mean_token_accuracy": 0.8773904070258141,
      "num_tokens": 178051753.0,
      "step": 147990
    },
    {
      "entropy": 1.8666868284344673,
      "epoch": 0.458787050735493,
      "grad_norm": 8.41880989074707,
      "learning_rate": 3.7349824976020483e-06,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8527174681425095,
      "num_tokens": 178063598.0,
      "step": 148000
    },
    {
      "entropy": 1.8387210443615913,
      "epoch": 0.4588180498605427,
      "grad_norm": 3.875763416290283,
      "learning_rate": 3.7348563213024038e-06,
      "loss": 0.4166,
      "mean_token_accuracy": 0.8554558053612709,
      "num_tokens": 178075606.0,
      "step": 148010
    },
    {
      "entropy": 1.8273346543312072,
      "epoch": 0.45884904898559237,
      "grad_norm": 9.78944206237793,
      "learning_rate": 3.7347301577894716e-06,
      "loss": 0.4098,
      "mean_token_accuracy": 0.8571007132530213,
      "num_tokens": 178088272.0,
      "step": 148020
    },
    {
      "entropy": 1.8279028117656708,
      "epoch": 0.4588800481106421,
      "grad_norm": 7.993279933929443,
      "learning_rate": 3.7346040070610935e-06,
      "loss": 0.4273,
      "mean_token_accuracy": 0.8609201908111572,
      "num_tokens": 178100592.0,
      "step": 148030
    },
    {
      "entropy": 1.9010518863797188,
      "epoch": 0.45891104723569176,
      "grad_norm": 8.148711204528809,
      "learning_rate": 3.73447786911511e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8482613220810891,
      "num_tokens": 178112256.0,
      "step": 148040
    },
    {
      "entropy": 1.84859948605299,
      "epoch": 0.4589420463607415,
      "grad_norm": 6.481073379516602,
      "learning_rate": 3.734351743949362e-06,
      "loss": 0.4582,
      "mean_token_accuracy": 0.8522376015782356,
      "num_tokens": 178124130.0,
      "step": 148050
    },
    {
      "entropy": 1.9222456485033035,
      "epoch": 0.45897304548579115,
      "grad_norm": 9.769478797912598,
      "learning_rate": 3.734225631561692e-06,
      "loss": 0.4303,
      "mean_token_accuracy": 0.858902907371521,
      "num_tokens": 178135001.0,
      "step": 148060
    },
    {
      "entropy": 1.9226790130138398,
      "epoch": 0.4590040446108409,
      "grad_norm": 7.749213218688965,
      "learning_rate": 3.7340995319499424e-06,
      "loss": 0.4852,
      "mean_token_accuracy": 0.8433895200490952,
      "num_tokens": 178146194.0,
      "step": 148070
    },
    {
      "entropy": 1.7752541199326515,
      "epoch": 0.45903504373589055,
      "grad_norm": 9.610624313354492,
      "learning_rate": 3.7339734451119556e-06,
      "loss": 0.423,
      "mean_token_accuracy": 0.8526219561696052,
      "num_tokens": 178159988.0,
      "step": 148080
    },
    {
      "entropy": 1.7483320951461792,
      "epoch": 0.45906604286094027,
      "grad_norm": 8.351910591125488,
      "learning_rate": 3.733847371045577e-06,
      "loss": 0.3534,
      "mean_token_accuracy": 0.8714915245771409,
      "num_tokens": 178173336.0,
      "step": 148090
    },
    {
      "entropy": 1.8959960505366324,
      "epoch": 0.45909704198598994,
      "grad_norm": 4.344943523406982,
      "learning_rate": 3.7337213097486484e-06,
      "loss": 0.4555,
      "mean_token_accuracy": 0.852771207690239,
      "num_tokens": 178184748.0,
      "step": 148100
    },
    {
      "entropy": 1.8576713547110557,
      "epoch": 0.45912804111103966,
      "grad_norm": 8.485532760620117,
      "learning_rate": 3.733595261219016e-06,
      "loss": 0.4161,
      "mean_token_accuracy": 0.8570611268281937,
      "num_tokens": 178196913.0,
      "step": 148110
    },
    {
      "entropy": 1.8324482500553132,
      "epoch": 0.45915904023608933,
      "grad_norm": 3.7649433612823486,
      "learning_rate": 3.7334692254545234e-06,
      "loss": 0.4328,
      "mean_token_accuracy": 0.8562913209199905,
      "num_tokens": 178209039.0,
      "step": 148120
    },
    {
      "entropy": 1.8635966286063195,
      "epoch": 0.45919003936113906,
      "grad_norm": 3.6343321800231934,
      "learning_rate": 3.733343202453017e-06,
      "loss": 0.4415,
      "mean_token_accuracy": 0.8553838461637497,
      "num_tokens": 178220749.0,
      "step": 148130
    },
    {
      "entropy": 1.8851610526442528,
      "epoch": 0.4592210384861887,
      "grad_norm": 6.960264682769775,
      "learning_rate": 3.7332171922123433e-06,
      "loss": 0.458,
      "mean_token_accuracy": 0.8590750455856323,
      "num_tokens": 178232494.0,
      "step": 148140
    },
    {
      "entropy": 1.8172919273376464,
      "epoch": 0.4592520376112384,
      "grad_norm": 8.392343521118164,
      "learning_rate": 3.733091194730348e-06,
      "loss": 0.4938,
      "mean_token_accuracy": 0.844761498272419,
      "num_tokens": 178245770.0,
      "step": 148150
    },
    {
      "entropy": 1.8854141429066658,
      "epoch": 0.4592830367362881,
      "grad_norm": 8.518744468688965,
      "learning_rate": 3.7329652100048796e-06,
      "loss": 0.4558,
      "mean_token_accuracy": 0.8478623673319816,
      "num_tokens": 178257483.0,
      "step": 148160
    },
    {
      "entropy": 1.9081405073404312,
      "epoch": 0.4593140358613378,
      "grad_norm": 8.384743690490723,
      "learning_rate": 3.7328392380337842e-06,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8525558218359948,
      "num_tokens": 178268727.0,
      "step": 148170
    },
    {
      "entropy": 1.8962554574012755,
      "epoch": 0.4593450349863875,
      "grad_norm": 9.84249496459961,
      "learning_rate": 3.73271327881491e-06,
      "loss": 0.4361,
      "mean_token_accuracy": 0.8616734966635704,
      "num_tokens": 178280654.0,
      "step": 148180
    },
    {
      "entropy": 1.9595209866762162,
      "epoch": 0.4593760341114372,
      "grad_norm": 7.133771896362305,
      "learning_rate": 3.732587332346106e-06,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8539301365613937,
      "num_tokens": 178292341.0,
      "step": 148190
    },
    {
      "entropy": 1.9443857610225677,
      "epoch": 0.4594070332364869,
      "grad_norm": 7.489109516143799,
      "learning_rate": 3.732461398625222e-06,
      "loss": 0.4481,
      "mean_token_accuracy": 0.8486480563879013,
      "num_tokens": 178303553.0,
      "step": 148200
    },
    {
      "entropy": 1.8499363631010055,
      "epoch": 0.4594380323615366,
      "grad_norm": 9.401103019714355,
      "learning_rate": 3.7323354776501063e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8625474691390991,
      "num_tokens": 178315044.0,
      "step": 148210
    },
    {
      "entropy": 1.9178618222475052,
      "epoch": 0.4594690314865863,
      "grad_norm": 7.870799541473389,
      "learning_rate": 3.73220956941861e-06,
      "loss": 0.4987,
      "mean_token_accuracy": 0.8432634264230728,
      "num_tokens": 178326928.0,
      "step": 148220
    },
    {
      "entropy": 1.948700188100338,
      "epoch": 0.45950003061163597,
      "grad_norm": 9.164449691772461,
      "learning_rate": 3.7320836739285838e-06,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8515329852700233,
      "num_tokens": 178338112.0,
      "step": 148230
    },
    {
      "entropy": 1.9085903450846673,
      "epoch": 0.4595310297366857,
      "grad_norm": 8.133758544921875,
      "learning_rate": 3.731957791177878e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8663553714752197,
      "num_tokens": 178349859.0,
      "step": 148240
    },
    {
      "entropy": 1.8798146709799766,
      "epoch": 0.45956202886173536,
      "grad_norm": 3.9677722454071045,
      "learning_rate": 3.7318319211643456e-06,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8445356607437133,
      "num_tokens": 178362252.0,
      "step": 148250
    },
    {
      "entropy": 1.843547348678112,
      "epoch": 0.4595930279867851,
      "grad_norm": 9.012165069580078,
      "learning_rate": 3.731706063885837e-06,
      "loss": 0.4151,
      "mean_token_accuracy": 0.8570879876613617,
      "num_tokens": 178374338.0,
      "step": 148260
    },
    {
      "entropy": 1.897803196310997,
      "epoch": 0.45962402711183475,
      "grad_norm": 2.348179817199707,
      "learning_rate": 3.731580219340206e-06,
      "loss": 0.4171,
      "mean_token_accuracy": 0.8629708766937256,
      "num_tokens": 178385867.0,
      "step": 148270
    },
    {
      "entropy": 1.755355989933014,
      "epoch": 0.4596550262368845,
      "grad_norm": 5.742542743682861,
      "learning_rate": 3.7314543875253065e-06,
      "loss": 0.359,
      "mean_token_accuracy": 0.8633789956569672,
      "num_tokens": 178400091.0,
      "step": 148280
    },
    {
      "entropy": 1.8990287438035012,
      "epoch": 0.45968602536193415,
      "grad_norm": 8.405163764953613,
      "learning_rate": 3.731328568438991e-06,
      "loss": 0.48,
      "mean_token_accuracy": 0.8494486227631569,
      "num_tokens": 178412761.0,
      "step": 148290
    },
    {
      "entropy": 1.8492143407464028,
      "epoch": 0.45971702448698387,
      "grad_norm": 8.039115905761719,
      "learning_rate": 3.731202762079114e-06,
      "loss": 0.3856,
      "mean_token_accuracy": 0.8641363605856895,
      "num_tokens": 178425735.0,
      "step": 148300
    },
    {
      "entropy": 1.9216509327292441,
      "epoch": 0.45974802361203354,
      "grad_norm": 8.291874885559082,
      "learning_rate": 3.7310769684435306e-06,
      "loss": 0.4854,
      "mean_token_accuracy": 0.8502373903989792,
      "num_tokens": 178437058.0,
      "step": 148310
    },
    {
      "entropy": 1.925723034143448,
      "epoch": 0.45977902273708326,
      "grad_norm": 9.9407320022583,
      "learning_rate": 3.730951187530095e-06,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8500989750027657,
      "num_tokens": 178448170.0,
      "step": 148320
    },
    {
      "entropy": 1.8789965465664864,
      "epoch": 0.45981002186213293,
      "grad_norm": 8.22014045715332,
      "learning_rate": 3.7308254193366646e-06,
      "loss": 0.4538,
      "mean_token_accuracy": 0.8639925360679627,
      "num_tokens": 178461198.0,
      "step": 148330
    },
    {
      "entropy": 1.9338722795248031,
      "epoch": 0.45984102098718266,
      "grad_norm": 7.892792224884033,
      "learning_rate": 3.7306996638610936e-06,
      "loss": 0.4897,
      "mean_token_accuracy": 0.8488512441515923,
      "num_tokens": 178471923.0,
      "step": 148340
    },
    {
      "entropy": 1.9255221277475356,
      "epoch": 0.4598720201122323,
      "grad_norm": 4.1723198890686035,
      "learning_rate": 3.7305739211012404e-06,
      "loss": 0.483,
      "mean_token_accuracy": 0.8413741737604141,
      "num_tokens": 178483299.0,
      "step": 148350
    },
    {
      "entropy": 1.9141563907265664,
      "epoch": 0.45990301923728205,
      "grad_norm": 7.0736260414123535,
      "learning_rate": 3.7304481910549613e-06,
      "loss": 0.4735,
      "mean_token_accuracy": 0.8529133036732673,
      "num_tokens": 178494625.0,
      "step": 148360
    },
    {
      "entropy": 1.8494025871157647,
      "epoch": 0.4599340183623317,
      "grad_norm": 7.612475872039795,
      "learning_rate": 3.7303224737201137e-06,
      "loss": 0.4706,
      "mean_token_accuracy": 0.8510186776518822,
      "num_tokens": 178507700.0,
      "step": 148370
    },
    {
      "entropy": 1.8061942994594573,
      "epoch": 0.45996501748738144,
      "grad_norm": 4.09743595123291,
      "learning_rate": 3.730196769094557e-06,
      "loss": 0.4241,
      "mean_token_accuracy": 0.8509301677346229,
      "num_tokens": 178521259.0,
      "step": 148380
    },
    {
      "entropy": 1.9035452425479888,
      "epoch": 0.4599960166124311,
      "grad_norm": 3.9658563137054443,
      "learning_rate": 3.73007107717615e-06,
      "loss": 0.4427,
      "mean_token_accuracy": 0.8570003524422646,
      "num_tokens": 178532802.0,
      "step": 148390
    },
    {
      "entropy": 1.821213935315609,
      "epoch": 0.4600270157374808,
      "grad_norm": 3.9787087440490723,
      "learning_rate": 3.729945397962751e-06,
      "loss": 0.3726,
      "mean_token_accuracy": 0.869298306107521,
      "num_tokens": 178545572.0,
      "step": 148400
    },
    {
      "entropy": 1.796069860458374,
      "epoch": 0.4600580148625305,
      "grad_norm": 4.056951999664307,
      "learning_rate": 3.72981973145222e-06,
      "loss": 0.4046,
      "mean_token_accuracy": 0.8616714045405388,
      "num_tokens": 178558915.0,
      "step": 148410
    },
    {
      "entropy": 1.9034935608506203,
      "epoch": 0.4600890139875802,
      "grad_norm": 8.9136381149292,
      "learning_rate": 3.7296940776424174e-06,
      "loss": 0.4599,
      "mean_token_accuracy": 0.851823453605175,
      "num_tokens": 178570522.0,
      "step": 148420
    },
    {
      "entropy": 1.921525527536869,
      "epoch": 0.4601200131126299,
      "grad_norm": 8.047652244567871,
      "learning_rate": 3.7295684365312045e-06,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8542478293180465,
      "num_tokens": 178581965.0,
      "step": 148430
    },
    {
      "entropy": 1.8355329155921936,
      "epoch": 0.46015101223767957,
      "grad_norm": 4.0909223556518555,
      "learning_rate": 3.7294428081164413e-06,
      "loss": 0.3998,
      "mean_token_accuracy": 0.8662302419543266,
      "num_tokens": 178594538.0,
      "step": 148440
    },
    {
      "entropy": 1.859887145459652,
      "epoch": 0.4601820113627293,
      "grad_norm": 9.011685371398926,
      "learning_rate": 3.729317192395991e-06,
      "loss": 0.453,
      "mean_token_accuracy": 0.8597676187753678,
      "num_tokens": 178606405.0,
      "step": 148450
    },
    {
      "entropy": 1.8634124889969825,
      "epoch": 0.46021301048777896,
      "grad_norm": 7.429311752319336,
      "learning_rate": 3.729191589367715e-06,
      "loss": 0.4054,
      "mean_token_accuracy": 0.8525898218154907,
      "num_tokens": 178619148.0,
      "step": 148460
    },
    {
      "entropy": 1.8275597229599954,
      "epoch": 0.4602440096128287,
      "grad_norm": 3.9588801860809326,
      "learning_rate": 3.729065999029476e-06,
      "loss": 0.3897,
      "mean_token_accuracy": 0.8704280957579613,
      "num_tokens": 178630852.0,
      "step": 148470
    },
    {
      "entropy": 1.8700731366872787,
      "epoch": 0.46027500873787836,
      "grad_norm": 9.530686378479004,
      "learning_rate": 3.728940421379138e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8597807481884956,
      "num_tokens": 178643067.0,
      "step": 148480
    },
    {
      "entropy": 1.8858630776405334,
      "epoch": 0.4603060078629281,
      "grad_norm": 8.52836799621582,
      "learning_rate": 3.7288148564145645e-06,
      "loss": 0.4496,
      "mean_token_accuracy": 0.84766735881567,
      "num_tokens": 178654791.0,
      "step": 148490
    },
    {
      "entropy": 1.9085488021373749,
      "epoch": 0.46033700698797775,
      "grad_norm": 7.8866095542907715,
      "learning_rate": 3.7286893041336187e-06,
      "loss": 0.512,
      "mean_token_accuracy": 0.8404016211628914,
      "num_tokens": 178666290.0,
      "step": 148500
    },
    {
      "entropy": 1.9258504971861838,
      "epoch": 0.4603680061130275,
      "grad_norm": 8.244267463684082,
      "learning_rate": 3.728563764534167e-06,
      "loss": 0.4795,
      "mean_token_accuracy": 0.8527882546186447,
      "num_tokens": 178677958.0,
      "step": 148510
    },
    {
      "entropy": 1.93827313631773,
      "epoch": 0.46039900523807714,
      "grad_norm": 7.484049320220947,
      "learning_rate": 3.728438237614075e-06,
      "loss": 0.4855,
      "mean_token_accuracy": 0.8486429572105407,
      "num_tokens": 178689438.0,
      "step": 148520
    },
    {
      "entropy": 1.8482612371444702,
      "epoch": 0.46043000436312687,
      "grad_norm": 6.9296650886535645,
      "learning_rate": 3.7283127233712067e-06,
      "loss": 0.405,
      "mean_token_accuracy": 0.8683997794985772,
      "num_tokens": 178702457.0,
      "step": 148530
    },
    {
      "entropy": 1.9950364649295806,
      "epoch": 0.46046100348817653,
      "grad_norm": 7.755855560302734,
      "learning_rate": 3.7281872218034292e-06,
      "loss": 0.5214,
      "mean_token_accuracy": 0.838994313776493,
      "num_tokens": 178713062.0,
      "step": 148540
    },
    {
      "entropy": 1.7965051412582398,
      "epoch": 0.46049200261322626,
      "grad_norm": 7.688485622406006,
      "learning_rate": 3.7280617329086093e-06,
      "loss": 0.3361,
      "mean_token_accuracy": 0.8706543311476708,
      "num_tokens": 178726109.0,
      "step": 148550
    },
    {
      "entropy": 1.939668095111847,
      "epoch": 0.46052300173827593,
      "grad_norm": 8.922786712646484,
      "learning_rate": 3.7279362566846154e-06,
      "loss": 0.4764,
      "mean_token_accuracy": 0.8509280979633331,
      "num_tokens": 178737184.0,
      "step": 148560
    },
    {
      "entropy": 1.8518592774868012,
      "epoch": 0.46055400086332565,
      "grad_norm": 8.265620231628418,
      "learning_rate": 3.7278107931293138e-06,
      "loss": 0.387,
      "mean_token_accuracy": 0.8692864283919335,
      "num_tokens": 178749576.0,
      "step": 148570
    },
    {
      "entropy": 1.9392090559005737,
      "epoch": 0.4605849999883753,
      "grad_norm": 8.548821449279785,
      "learning_rate": 3.7276853422405733e-06,
      "loss": 0.5045,
      "mean_token_accuracy": 0.8460419103503227,
      "num_tokens": 178760264.0,
      "step": 148580
    },
    {
      "entropy": 1.9111363351345063,
      "epoch": 0.46061599911342505,
      "grad_norm": 7.530226230621338,
      "learning_rate": 3.7275599040162634e-06,
      "loss": 0.4283,
      "mean_token_accuracy": 0.8527784243226051,
      "num_tokens": 178771695.0,
      "step": 148590
    },
    {
      "entropy": 1.8285814180970192,
      "epoch": 0.4606469982384747,
      "grad_norm": 3.894634485244751,
      "learning_rate": 3.727434478454252e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8559230253100395,
      "num_tokens": 178784229.0,
      "step": 148600
    },
    {
      "entropy": 1.8875297084450722,
      "epoch": 0.46067799736352444,
      "grad_norm": 7.381993293762207,
      "learning_rate": 3.7273090655524108e-06,
      "loss": 0.4304,
      "mean_token_accuracy": 0.8570609837770462,
      "num_tokens": 178796492.0,
      "step": 148610
    },
    {
      "entropy": 1.8404451578855514,
      "epoch": 0.4607089964885741,
      "grad_norm": 7.309342384338379,
      "learning_rate": 3.7271836653086084e-06,
      "loss": 0.4021,
      "mean_token_accuracy": 0.8731897443532943,
      "num_tokens": 178808742.0,
      "step": 148620
    },
    {
      "entropy": 1.879277119040489,
      "epoch": 0.46073999561362383,
      "grad_norm": 9.954212188720703,
      "learning_rate": 3.7270582777207166e-06,
      "loss": 0.4689,
      "mean_token_accuracy": 0.8552733480930328,
      "num_tokens": 178820065.0,
      "step": 148630
    },
    {
      "entropy": 1.909400151669979,
      "epoch": 0.4607709947386735,
      "grad_norm": 8.673418045043945,
      "learning_rate": 3.7269329027866064e-06,
      "loss": 0.4915,
      "mean_token_accuracy": 0.8421248093247413,
      "num_tokens": 178832605.0,
      "step": 148640
    },
    {
      "entropy": 1.9523657143115998,
      "epoch": 0.46080199386372317,
      "grad_norm": 7.831915378570557,
      "learning_rate": 3.7268075405041496e-06,
      "loss": 0.4772,
      "mean_token_accuracy": 0.8484080314636231,
      "num_tokens": 178844397.0,
      "step": 148650
    },
    {
      "entropy": 1.7859274119138717,
      "epoch": 0.4608329929887729,
      "grad_norm": 8.42618465423584,
      "learning_rate": 3.7266821908712185e-06,
      "loss": 0.4208,
      "mean_token_accuracy": 0.8656513899564743,
      "num_tokens": 178856946.0,
      "step": 148660
    },
    {
      "entropy": 1.9320579081773759,
      "epoch": 0.46086399211382256,
      "grad_norm": 7.774157524108887,
      "learning_rate": 3.7265568538856866e-06,
      "loss": 0.4744,
      "mean_token_accuracy": 0.8523464262485504,
      "num_tokens": 178868317.0,
      "step": 148670
    },
    {
      "entropy": 1.8498662784695625,
      "epoch": 0.4608949912388723,
      "grad_norm": 8.220366477966309,
      "learning_rate": 3.726431529545426e-06,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8491094589233399,
      "num_tokens": 178880925.0,
      "step": 148680
    },
    {
      "entropy": 1.9150644198060036,
      "epoch": 0.46092599036392196,
      "grad_norm": 9.27624225616455,
      "learning_rate": 3.726306217848312e-06,
      "loss": 0.4911,
      "mean_token_accuracy": 0.8403692692518234,
      "num_tokens": 178892546.0,
      "step": 148690
    },
    {
      "entropy": 1.7894390970468521,
      "epoch": 0.4609569894889717,
      "grad_norm": 9.18440055847168,
      "learning_rate": 3.726180918792218e-06,
      "loss": 0.3835,
      "mean_token_accuracy": 0.8709027290344238,
      "num_tokens": 178904334.0,
      "step": 148700
    },
    {
      "entropy": 1.858834259212017,
      "epoch": 0.46098798861402135,
      "grad_norm": 3.9686553478240967,
      "learning_rate": 3.7260556323750185e-06,
      "loss": 0.3973,
      "mean_token_accuracy": 0.8583307415246964,
      "num_tokens": 178916596.0,
      "step": 148710
    },
    {
      "entropy": 1.8790298044681548,
      "epoch": 0.4610189877390711,
      "grad_norm": 8.098514556884766,
      "learning_rate": 3.72593035859459e-06,
      "loss": 0.4591,
      "mean_token_accuracy": 0.8534575864672661,
      "num_tokens": 178929270.0,
      "step": 148720
    },
    {
      "entropy": 1.8725609213113785,
      "epoch": 0.46104998686412074,
      "grad_norm": 8.010581016540527,
      "learning_rate": 3.725805097448807e-06,
      "loss": 0.4111,
      "mean_token_accuracy": 0.8627361282706261,
      "num_tokens": 178941517.0,
      "step": 148730
    },
    {
      "entropy": 1.9305458396673203,
      "epoch": 0.46108098598917047,
      "grad_norm": 8.154109954833984,
      "learning_rate": 3.7256798489355474e-06,
      "loss": 0.4493,
      "mean_token_accuracy": 0.8501837208867074,
      "num_tokens": 178953101.0,
      "step": 148740
    },
    {
      "entropy": 1.9796837285161017,
      "epoch": 0.46111198511422014,
      "grad_norm": 9.379487991333008,
      "learning_rate": 3.7255546130526867e-06,
      "loss": 0.508,
      "mean_token_accuracy": 0.8350688204169273,
      "num_tokens": 178964485.0,
      "step": 148750
    },
    {
      "entropy": 1.951874351501465,
      "epoch": 0.46114298423926986,
      "grad_norm": 7.641496658325195,
      "learning_rate": 3.7254293897981025e-06,
      "loss": 0.4815,
      "mean_token_accuracy": 0.843595664203167,
      "num_tokens": 178975573.0,
      "step": 148760
    },
    {
      "entropy": 1.8258162140846252,
      "epoch": 0.46117398336431953,
      "grad_norm": 2.5984456539154053,
      "learning_rate": 3.725304179169673e-06,
      "loss": 0.4267,
      "mean_token_accuracy": 0.8542465507984162,
      "num_tokens": 178988605.0,
      "step": 148770
    },
    {
      "entropy": 1.8872524976730347,
      "epoch": 0.46120498248936925,
      "grad_norm": 7.967230319976807,
      "learning_rate": 3.7251789811652763e-06,
      "loss": 0.4397,
      "mean_token_accuracy": 0.8548190101981163,
      "num_tokens": 179001072.0,
      "step": 148780
    },
    {
      "entropy": 1.9689855933189393,
      "epoch": 0.4612359816144189,
      "grad_norm": 4.094521522521973,
      "learning_rate": 3.7250537957827913e-06,
      "loss": 0.5026,
      "mean_token_accuracy": 0.8418999388813972,
      "num_tokens": 179012080.0,
      "step": 148790
    },
    {
      "entropy": 1.9371369108557701,
      "epoch": 0.46126698073946865,
      "grad_norm": 7.492170333862305,
      "learning_rate": 3.7249286230200974e-06,
      "loss": 0.4804,
      "mean_token_accuracy": 0.854214908182621,
      "num_tokens": 179023118.0,
      "step": 148800
    },
    {
      "entropy": 1.941811603307724,
      "epoch": 0.4612979798645183,
      "grad_norm": 7.5748491287231445,
      "learning_rate": 3.7248034628750744e-06,
      "loss": 0.4423,
      "mean_token_accuracy": 0.85891852080822,
      "num_tokens": 179034894.0,
      "step": 148810
    },
    {
      "entropy": 1.8954156935214996,
      "epoch": 0.46132897898956804,
      "grad_norm": 3.9654111862182617,
      "learning_rate": 3.7246783153456033e-06,
      "loss": 0.4425,
      "mean_token_accuracy": 0.8546635210514069,
      "num_tokens": 179047196.0,
      "step": 148820
    },
    {
      "entropy": 1.9124370649456979,
      "epoch": 0.4613599781146177,
      "grad_norm": 8.43816089630127,
      "learning_rate": 3.7245531804295637e-06,
      "loss": 0.4774,
      "mean_token_accuracy": 0.8510984659194947,
      "num_tokens": 179058196.0,
      "step": 148830
    },
    {
      "entropy": 1.949829702079296,
      "epoch": 0.46139097723966743,
      "grad_norm": 9.092519760131836,
      "learning_rate": 3.724428058124837e-06,
      "loss": 0.4881,
      "mean_token_accuracy": 0.8465055644512176,
      "num_tokens": 179069758.0,
      "step": 148840
    },
    {
      "entropy": 1.9349031627178193,
      "epoch": 0.4614219763647171,
      "grad_norm": 7.756466865539551,
      "learning_rate": 3.7243029484293057e-06,
      "loss": 0.4625,
      "mean_token_accuracy": 0.8493330150842666,
      "num_tokens": 179081712.0,
      "step": 148850
    },
    {
      "entropy": 1.8835980251431466,
      "epoch": 0.4614529754897668,
      "grad_norm": 5.772395610809326,
      "learning_rate": 3.724177851340852e-06,
      "loss": 0.5075,
      "mean_token_accuracy": 0.8446671605110169,
      "num_tokens": 179095186.0,
      "step": 148860
    },
    {
      "entropy": 1.9193114623427392,
      "epoch": 0.4614839746148165,
      "grad_norm": 7.486071586608887,
      "learning_rate": 3.724052766857359e-06,
      "loss": 0.4239,
      "mean_token_accuracy": 0.8581682324409485,
      "num_tokens": 179107120.0,
      "step": 148870
    },
    {
      "entropy": 1.9004424124956132,
      "epoch": 0.4615149737398662,
      "grad_norm": 4.124865531921387,
      "learning_rate": 3.7239276949767094e-06,
      "loss": 0.4304,
      "mean_token_accuracy": 0.8557851612567902,
      "num_tokens": 179119606.0,
      "step": 148880
    },
    {
      "entropy": 1.8963897615671157,
      "epoch": 0.4615459728649159,
      "grad_norm": 3.557903289794922,
      "learning_rate": 3.7238026356967873e-06,
      "loss": 0.4233,
      "mean_token_accuracy": 0.8588035330176353,
      "num_tokens": 179132126.0,
      "step": 148890
    },
    {
      "entropy": 1.9891429126262665,
      "epoch": 0.46157697198996556,
      "grad_norm": 8.919794082641602,
      "learning_rate": 3.7236775890154765e-06,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8473231881856919,
      "num_tokens": 179143132.0,
      "step": 148900
    },
    {
      "entropy": 1.9493281915783882,
      "epoch": 0.4616079711150153,
      "grad_norm": 9.03589153289795,
      "learning_rate": 3.723552554930663e-06,
      "loss": 0.5008,
      "mean_token_accuracy": 0.8437724366784096,
      "num_tokens": 179154839.0,
      "step": 148910
    },
    {
      "entropy": 1.879402995109558,
      "epoch": 0.46163897024006495,
      "grad_norm": 7.5339741706848145,
      "learning_rate": 3.7234275334402314e-06,
      "loss": 0.4215,
      "mean_token_accuracy": 0.8598219409585,
      "num_tokens": 179167503.0,
      "step": 148920
    },
    {
      "entropy": 1.8404528819024564,
      "epoch": 0.4616699693651147,
      "grad_norm": 5.325761795043945,
      "learning_rate": 3.7233025245420666e-06,
      "loss": 0.4685,
      "mean_token_accuracy": 0.8565814971923829,
      "num_tokens": 179180748.0,
      "step": 148930
    },
    {
      "entropy": 1.8871916115283967,
      "epoch": 0.46170096849016434,
      "grad_norm": 7.421195983886719,
      "learning_rate": 3.7231775282340564e-06,
      "loss": 0.4556,
      "mean_token_accuracy": 0.8600098505616188,
      "num_tokens": 179192693.0,
      "step": 148940
    },
    {
      "entropy": 1.9641156986355781,
      "epoch": 0.46173196761521407,
      "grad_norm": 7.706519603729248,
      "learning_rate": 3.7230525445140874e-06,
      "loss": 0.4721,
      "mean_token_accuracy": 0.8449445277452469,
      "num_tokens": 179203723.0,
      "step": 148950
    },
    {
      "entropy": 1.9411569505929946,
      "epoch": 0.46176296674026374,
      "grad_norm": 9.283172607421875,
      "learning_rate": 3.7229275733800462e-06,
      "loss": 0.4603,
      "mean_token_accuracy": 0.8541247725486756,
      "num_tokens": 179215213.0,
      "step": 148960
    },
    {
      "entropy": 2.000536176562309,
      "epoch": 0.46179396586531346,
      "grad_norm": 7.330437183380127,
      "learning_rate": 3.722802614829821e-06,
      "loss": 0.4743,
      "mean_token_accuracy": 0.8524016052484512,
      "num_tokens": 179225916.0,
      "step": 148970
    },
    {
      "entropy": 1.841599515080452,
      "epoch": 0.46182496499036313,
      "grad_norm": 8.324941635131836,
      "learning_rate": 3.7226776688612994e-06,
      "loss": 0.4567,
      "mean_token_accuracy": 0.8502395913004875,
      "num_tokens": 179238822.0,
      "step": 148980
    },
    {
      "entropy": 1.8596518978476524,
      "epoch": 0.46185596411541285,
      "grad_norm": 9.439504623413086,
      "learning_rate": 3.722552735472371e-06,
      "loss": 0.3836,
      "mean_token_accuracy": 0.8619119867682457,
      "num_tokens": 179251678.0,
      "step": 148990
    },
    {
      "entropy": 1.9302584543824195,
      "epoch": 0.4618869632404625,
      "grad_norm": 9.211128234863281,
      "learning_rate": 3.722427814660926e-06,
      "loss": 0.4525,
      "mean_token_accuracy": 0.8613484084606171,
      "num_tokens": 179263263.0,
      "step": 149000
    },
    {
      "entropy": 1.840495379269123,
      "epoch": 0.46191796236551225,
      "grad_norm": 7.943533897399902,
      "learning_rate": 3.722302906424852e-06,
      "loss": 0.3887,
      "mean_token_accuracy": 0.8663021042943001,
      "num_tokens": 179275505.0,
      "step": 149010
    },
    {
      "entropy": 1.9889112055301665,
      "epoch": 0.4619489614905619,
      "grad_norm": 8.111519813537598,
      "learning_rate": 3.7221780107620397e-06,
      "loss": 0.4651,
      "mean_token_accuracy": 0.8498691335320473,
      "num_tokens": 179286521.0,
      "step": 149020
    },
    {
      "entropy": 1.9183252349495887,
      "epoch": 0.46197996061561164,
      "grad_norm": 7.588090419769287,
      "learning_rate": 3.7220531276703815e-06,
      "loss": 0.4665,
      "mean_token_accuracy": 0.8486698940396309,
      "num_tokens": 179297674.0,
      "step": 149030
    },
    {
      "entropy": 1.8954298749566079,
      "epoch": 0.4620109597406613,
      "grad_norm": 7.996798515319824,
      "learning_rate": 3.7219282571477677e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.852338932454586,
      "num_tokens": 179309717.0,
      "step": 149040
    },
    {
      "entropy": 1.902578841149807,
      "epoch": 0.46204195886571103,
      "grad_norm": 7.982056140899658,
      "learning_rate": 3.7218033991920895e-06,
      "loss": 0.4607,
      "mean_token_accuracy": 0.8511950179934502,
      "num_tokens": 179321639.0,
      "step": 149050
    },
    {
      "entropy": 1.8236243352293968,
      "epoch": 0.4620729579907607,
      "grad_norm": 8.151361465454102,
      "learning_rate": 3.7216785538012397e-06,
      "loss": 0.3998,
      "mean_token_accuracy": 0.8600339099764824,
      "num_tokens": 179334706.0,
      "step": 149060
    },
    {
      "entropy": 1.9069883227348328,
      "epoch": 0.4621039571158104,
      "grad_norm": 8.232033729553223,
      "learning_rate": 3.7215537209731107e-06,
      "loss": 0.4365,
      "mean_token_accuracy": 0.860856780409813,
      "num_tokens": 179346482.0,
      "step": 149070
    },
    {
      "entropy": 1.8350558295845985,
      "epoch": 0.4621349562408601,
      "grad_norm": 8.054594039916992,
      "learning_rate": 3.7214289007055965e-06,
      "loss": 0.3866,
      "mean_token_accuracy": 0.8655603528022766,
      "num_tokens": 179359849.0,
      "step": 149080
    },
    {
      "entropy": 1.972979885339737,
      "epoch": 0.4621659553659098,
      "grad_norm": 10.824267387390137,
      "learning_rate": 3.7213040929965905e-06,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8447522640228271,
      "num_tokens": 179371308.0,
      "step": 149090
    },
    {
      "entropy": 1.8467707887291909,
      "epoch": 0.4621969544909595,
      "grad_norm": 8.265110969543457,
      "learning_rate": 3.721179297843986e-06,
      "loss": 0.4606,
      "mean_token_accuracy": 0.8495004117488861,
      "num_tokens": 179384138.0,
      "step": 149100
    },
    {
      "entropy": 1.9475996479392053,
      "epoch": 0.4622279536160092,
      "grad_norm": 7.331584453582764,
      "learning_rate": 3.721054515245679e-06,
      "loss": 0.5038,
      "mean_token_accuracy": 0.848460578918457,
      "num_tokens": 179395588.0,
      "step": 149110
    },
    {
      "entropy": 1.879881000518799,
      "epoch": 0.4622589527410589,
      "grad_norm": 7.338623523712158,
      "learning_rate": 3.720929745199565e-06,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8511856839060783,
      "num_tokens": 179407656.0,
      "step": 149120
    },
    {
      "entropy": 1.9637582659721375,
      "epoch": 0.4622899518661086,
      "grad_norm": 9.248175621032715,
      "learning_rate": 3.720804987703538e-06,
      "loss": 0.4885,
      "mean_token_accuracy": 0.8401120364665985,
      "num_tokens": 179418775.0,
      "step": 149130
    },
    {
      "entropy": 1.9078269064426423,
      "epoch": 0.4623209509911583,
      "grad_norm": 8.055317878723145,
      "learning_rate": 3.7206802427554957e-06,
      "loss": 0.4194,
      "mean_token_accuracy": 0.856049720942974,
      "num_tokens": 179431246.0,
      "step": 149140
    },
    {
      "entropy": 1.880311058461666,
      "epoch": 0.46235195011620794,
      "grad_norm": 3.6950716972351074,
      "learning_rate": 3.7205555103533336e-06,
      "loss": 0.4157,
      "mean_token_accuracy": 0.862914027273655,
      "num_tokens": 179443865.0,
      "step": 149150
    },
    {
      "entropy": 1.9153502866625787,
      "epoch": 0.46238294924125767,
      "grad_norm": 3.731424331665039,
      "learning_rate": 3.7204307904949507e-06,
      "loss": 0.4726,
      "mean_token_accuracy": 0.849963067471981,
      "num_tokens": 179455278.0,
      "step": 149160
    },
    {
      "entropy": 1.9533527433872222,
      "epoch": 0.46241394836630734,
      "grad_norm": 8.858798027038574,
      "learning_rate": 3.7203060831782423e-06,
      "loss": 0.4618,
      "mean_token_accuracy": 0.8534232169389725,
      "num_tokens": 179467185.0,
      "step": 149170
    },
    {
      "entropy": 1.9469057649374009,
      "epoch": 0.46244494749135706,
      "grad_norm": 7.934726715087891,
      "learning_rate": 3.7201813884011084e-06,
      "loss": 0.4779,
      "mean_token_accuracy": 0.8529333204030991,
      "num_tokens": 179478428.0,
      "step": 149180
    },
    {
      "entropy": 1.8080351307988167,
      "epoch": 0.46247594661640673,
      "grad_norm": 7.475351810455322,
      "learning_rate": 3.7200567061614475e-06,
      "loss": 0.4169,
      "mean_token_accuracy": 0.8639614179730415,
      "num_tokens": 179491614.0,
      "step": 149190
    },
    {
      "entropy": 1.9022365421056748,
      "epoch": 0.46250694574145645,
      "grad_norm": 8.010424613952637,
      "learning_rate": 3.719932036457157e-06,
      "loss": 0.4343,
      "mean_token_accuracy": 0.8514123097062111,
      "num_tokens": 179502942.0,
      "step": 149200
    },
    {
      "entropy": 2.004026171565056,
      "epoch": 0.4625379448665061,
      "grad_norm": 12.545236587524414,
      "learning_rate": 3.71980737928614e-06,
      "loss": 0.5794,
      "mean_token_accuracy": 0.830876411497593,
      "num_tokens": 179513755.0,
      "step": 149210
    },
    {
      "entropy": 1.8916426077485085,
      "epoch": 0.46256894399155585,
      "grad_norm": 5.487705230712891,
      "learning_rate": 3.719682734646293e-06,
      "loss": 0.4147,
      "mean_token_accuracy": 0.8689695119857788,
      "num_tokens": 179525126.0,
      "step": 149220
    },
    {
      "entropy": 1.8946526229381562,
      "epoch": 0.4625999431166055,
      "grad_norm": 9.722243309020996,
      "learning_rate": 3.7195581025355194e-06,
      "loss": 0.4581,
      "mean_token_accuracy": 0.8481919586658477,
      "num_tokens": 179536892.0,
      "step": 149230
    },
    {
      "entropy": 1.9386965110898018,
      "epoch": 0.46263094224165524,
      "grad_norm": 7.750062942504883,
      "learning_rate": 3.719433482951718e-06,
      "loss": 0.4443,
      "mean_token_accuracy": 0.8568326279520988,
      "num_tokens": 179548377.0,
      "step": 149240
    },
    {
      "entropy": 1.8785508409142495,
      "epoch": 0.4626619413667049,
      "grad_norm": 4.4261393547058105,
      "learning_rate": 3.7193088758927924e-06,
      "loss": 0.3965,
      "mean_token_accuracy": 0.8603580877184868,
      "num_tokens": 179560882.0,
      "step": 149250
    },
    {
      "entropy": 1.863745103776455,
      "epoch": 0.46269294049175463,
      "grad_norm": 7.138288974761963,
      "learning_rate": 3.7191842813566436e-06,
      "loss": 0.4085,
      "mean_token_accuracy": 0.8604954317212105,
      "num_tokens": 179573767.0,
      "step": 149260
    },
    {
      "entropy": 1.848945789039135,
      "epoch": 0.4627239396168043,
      "grad_norm": 5.103929042816162,
      "learning_rate": 3.7190596993411744e-06,
      "loss": 0.4499,
      "mean_token_accuracy": 0.8594519913196563,
      "num_tokens": 179587556.0,
      "step": 149270
    },
    {
      "entropy": 1.8771667987108231,
      "epoch": 0.462754938741854,
      "grad_norm": 9.511100769042969,
      "learning_rate": 3.7189351298442885e-06,
      "loss": 0.4375,
      "mean_token_accuracy": 0.8482055693864823,
      "num_tokens": 179599698.0,
      "step": 149280
    },
    {
      "entropy": 1.8984940245747566,
      "epoch": 0.4627859378669037,
      "grad_norm": 3.7801976203918457,
      "learning_rate": 3.718810572863889e-06,
      "loss": 0.4823,
      "mean_token_accuracy": 0.8494816675782204,
      "num_tokens": 179611306.0,
      "step": 149290
    },
    {
      "entropy": 1.8457211047410964,
      "epoch": 0.4628169369919534,
      "grad_norm": 8.371335983276367,
      "learning_rate": 3.718686028397879e-06,
      "loss": 0.425,
      "mean_token_accuracy": 0.8612516298890114,
      "num_tokens": 179623572.0,
      "step": 149300
    },
    {
      "entropy": 1.8669260188937187,
      "epoch": 0.4628479361170031,
      "grad_norm": 4.057291507720947,
      "learning_rate": 3.718561496444166e-06,
      "loss": 0.4648,
      "mean_token_accuracy": 0.8506672963500023,
      "num_tokens": 179635689.0,
      "step": 149310
    },
    {
      "entropy": 2.0177447497844696,
      "epoch": 0.4628789352420528,
      "grad_norm": 7.82882022857666,
      "learning_rate": 3.718436977000652e-06,
      "loss": 0.5247,
      "mean_token_accuracy": 0.8370867878198623,
      "num_tokens": 179646327.0,
      "step": 149320
    },
    {
      "entropy": 1.9195616245269775,
      "epoch": 0.4629099343671025,
      "grad_norm": 4.400636196136475,
      "learning_rate": 3.7183124700652433e-06,
      "loss": 0.4436,
      "mean_token_accuracy": 0.8610096588730812,
      "num_tokens": 179658283.0,
      "step": 149330
    },
    {
      "entropy": 1.9372329905629158,
      "epoch": 0.4629409334921522,
      "grad_norm": 7.58730936050415,
      "learning_rate": 3.7181879756358472e-06,
      "loss": 0.4956,
      "mean_token_accuracy": 0.8437316343188286,
      "num_tokens": 179670274.0,
      "step": 149340
    },
    {
      "entropy": 1.8574361830949784,
      "epoch": 0.4629719326172019,
      "grad_norm": 6.498873233795166,
      "learning_rate": 3.7180634937103697e-06,
      "loss": 0.4185,
      "mean_token_accuracy": 0.8578757628798485,
      "num_tokens": 179682351.0,
      "step": 149350
    },
    {
      "entropy": 1.9330038219690322,
      "epoch": 0.4630029317422516,
      "grad_norm": 8.56723690032959,
      "learning_rate": 3.717939024286717e-06,
      "loss": 0.4655,
      "mean_token_accuracy": 0.8528059482574463,
      "num_tokens": 179693716.0,
      "step": 149360
    },
    {
      "entropy": 1.9517970651388168,
      "epoch": 0.46303393086730127,
      "grad_norm": 7.569751262664795,
      "learning_rate": 3.7178145673627976e-06,
      "loss": 0.4461,
      "mean_token_accuracy": 0.8599812477827072,
      "num_tokens": 179705054.0,
      "step": 149370
    },
    {
      "entropy": 1.8608211636543275,
      "epoch": 0.463064929992351,
      "grad_norm": 4.166712760925293,
      "learning_rate": 3.717690122936518e-06,
      "loss": 0.4196,
      "mean_token_accuracy": 0.8539978817105294,
      "num_tokens": 179718580.0,
      "step": 149380
    },
    {
      "entropy": 1.797726885974407,
      "epoch": 0.46309592911740066,
      "grad_norm": 8.749974250793457,
      "learning_rate": 3.7175656910057894e-06,
      "loss": 0.4225,
      "mean_token_accuracy": 0.8570615857839584,
      "num_tokens": 179732381.0,
      "step": 149390
    },
    {
      "entropy": 1.863544289022684,
      "epoch": 0.46312692824245033,
      "grad_norm": 6.951138496398926,
      "learning_rate": 3.717441271568518e-06,
      "loss": 0.4032,
      "mean_token_accuracy": 0.8617607593536377,
      "num_tokens": 179745055.0,
      "step": 149400
    },
    {
      "entropy": 1.8385921865701675,
      "epoch": 0.46315792736750006,
      "grad_norm": 3.46244215965271,
      "learning_rate": 3.7173168646226155e-06,
      "loss": 0.3779,
      "mean_token_accuracy": 0.8700759828090667,
      "num_tokens": 179757818.0,
      "step": 149410
    },
    {
      "entropy": 1.88816240131855,
      "epoch": 0.4631889264925497,
      "grad_norm": 6.927772521972656,
      "learning_rate": 3.7171924701659907e-06,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8400983780622482,
      "num_tokens": 179770718.0,
      "step": 149420
    },
    {
      "entropy": 1.9157932326197624,
      "epoch": 0.46321992561759945,
      "grad_norm": 6.942224025726318,
      "learning_rate": 3.717068088196554e-06,
      "loss": 0.4132,
      "mean_token_accuracy": 0.8620764970779419,
      "num_tokens": 179782310.0,
      "step": 149430
    },
    {
      "entropy": 1.925161102414131,
      "epoch": 0.4632509247426491,
      "grad_norm": 6.985175609588623,
      "learning_rate": 3.7169437187122166e-06,
      "loss": 0.4419,
      "mean_token_accuracy": 0.8568544924259186,
      "num_tokens": 179793811.0,
      "step": 149440
    },
    {
      "entropy": 1.8122023433446883,
      "epoch": 0.46328192386769884,
      "grad_norm": 8.966583251953125,
      "learning_rate": 3.71681936171089e-06,
      "loss": 0.4477,
      "mean_token_accuracy": 0.8570965170860291,
      "num_tokens": 179807530.0,
      "step": 149450
    },
    {
      "entropy": 1.8448025688529015,
      "epoch": 0.4633129229927485,
      "grad_norm": 8.287588119506836,
      "learning_rate": 3.716695017190486e-06,
      "loss": 0.3767,
      "mean_token_accuracy": 0.8626005664467812,
      "num_tokens": 179820550.0,
      "step": 149460
    },
    {
      "entropy": 1.963685867190361,
      "epoch": 0.46334392211779823,
      "grad_norm": 7.667242050170898,
      "learning_rate": 3.716570685148917e-06,
      "loss": 0.4761,
      "mean_token_accuracy": 0.85397337526083,
      "num_tokens": 179830999.0,
      "step": 149470
    },
    {
      "entropy": 1.8665792480111123,
      "epoch": 0.4633749212428479,
      "grad_norm": 8.30051040649414,
      "learning_rate": 3.716446365584096e-06,
      "loss": 0.3937,
      "mean_token_accuracy": 0.8599151030182839,
      "num_tokens": 179843005.0,
      "step": 149480
    },
    {
      "entropy": 1.8342208310961723,
      "epoch": 0.46340592036789763,
      "grad_norm": 3.5341451168060303,
      "learning_rate": 3.7163220584939363e-06,
      "loss": 0.402,
      "mean_token_accuracy": 0.8594845876097679,
      "num_tokens": 179855385.0,
      "step": 149490
    },
    {
      "entropy": 1.9173177301883697,
      "epoch": 0.4634369194929473,
      "grad_norm": 8.4180326461792,
      "learning_rate": 3.7161977638763523e-06,
      "loss": 0.4628,
      "mean_token_accuracy": 0.8482234224677085,
      "num_tokens": 179867801.0,
      "step": 149500
    },
    {
      "entropy": 1.9616335391998292,
      "epoch": 0.463467918617997,
      "grad_norm": 6.9208221435546875,
      "learning_rate": 3.716073481729258e-06,
      "loss": 0.5128,
      "mean_token_accuracy": 0.8467522487044334,
      "num_tokens": 179879398.0,
      "step": 149510
    },
    {
      "entropy": 1.8291648626327515,
      "epoch": 0.4634989177430467,
      "grad_norm": 3.6951959133148193,
      "learning_rate": 3.715949212050568e-06,
      "loss": 0.4187,
      "mean_token_accuracy": 0.8521149665117264,
      "num_tokens": 179892903.0,
      "step": 149520
    },
    {
      "entropy": 1.9397112756967545,
      "epoch": 0.4635299168680964,
      "grad_norm": 7.79329252243042,
      "learning_rate": 3.715824954838198e-06,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8460395872592926,
      "num_tokens": 179903789.0,
      "step": 149530
    },
    {
      "entropy": 1.8958303049206733,
      "epoch": 0.4635609159931461,
      "grad_norm": 9.17414665222168,
      "learning_rate": 3.715700710090064e-06,
      "loss": 0.4351,
      "mean_token_accuracy": 0.8583477824926377,
      "num_tokens": 179915857.0,
      "step": 149540
    },
    {
      "entropy": 1.934301419556141,
      "epoch": 0.4635919151181958,
      "grad_norm": 11.262062072753906,
      "learning_rate": 3.7155764778040813e-06,
      "loss": 0.4945,
      "mean_token_accuracy": 0.8361470863223076,
      "num_tokens": 179927748.0,
      "step": 149550
    },
    {
      "entropy": 2.001458024978638,
      "epoch": 0.4636229142432455,
      "grad_norm": 9.550740242004395,
      "learning_rate": 3.7154522579781682e-06,
      "loss": 0.531,
      "mean_token_accuracy": 0.8417502701282501,
      "num_tokens": 179938659.0,
      "step": 149560
    },
    {
      "entropy": 1.838100890815258,
      "epoch": 0.4636539133682952,
      "grad_norm": 7.261234760284424,
      "learning_rate": 3.715328050610241e-06,
      "loss": 0.3895,
      "mean_token_accuracy": 0.8724177911877632,
      "num_tokens": 179951437.0,
      "step": 149570
    },
    {
      "entropy": 1.8873604014515877,
      "epoch": 0.46368491249334487,
      "grad_norm": 4.315876007080078,
      "learning_rate": 3.715203855698218e-06,
      "loss": 0.4199,
      "mean_token_accuracy": 0.854177550971508,
      "num_tokens": 179964006.0,
      "step": 149580
    },
    {
      "entropy": 1.938150754570961,
      "epoch": 0.4637159116183946,
      "grad_norm": 8.612800598144531,
      "learning_rate": 3.715079673240017e-06,
      "loss": 0.4698,
      "mean_token_accuracy": 0.8499852314591407,
      "num_tokens": 179975104.0,
      "step": 149590
    },
    {
      "entropy": 1.9041712909936905,
      "epoch": 0.46374691074344426,
      "grad_norm": 3.8847134113311768,
      "learning_rate": 3.714955503233557e-06,
      "loss": 0.4501,
      "mean_token_accuracy": 0.8547717839479446,
      "num_tokens": 179987110.0,
      "step": 149600
    },
    {
      "entropy": 1.951756013929844,
      "epoch": 0.463777909868494,
      "grad_norm": 6.442052364349365,
      "learning_rate": 3.714831345676757e-06,
      "loss": 0.4659,
      "mean_token_accuracy": 0.8537251800298691,
      "num_tokens": 179998453.0,
      "step": 149610
    },
    {
      "entropy": 1.8288054898381234,
      "epoch": 0.46380890899354366,
      "grad_norm": 6.619274139404297,
      "learning_rate": 3.714707200567537e-06,
      "loss": 0.3848,
      "mean_token_accuracy": 0.8710165083408355,
      "num_tokens": 180010976.0,
      "step": 149620
    },
    {
      "entropy": 1.9353380858898164,
      "epoch": 0.4638399081185934,
      "grad_norm": 7.440566062927246,
      "learning_rate": 3.7145830679038177e-06,
      "loss": 0.495,
      "mean_token_accuracy": 0.8452064722776413,
      "num_tokens": 180022973.0,
      "step": 149630
    },
    {
      "entropy": 1.857330885529518,
      "epoch": 0.46387090724364305,
      "grad_norm": 5.837486267089844,
      "learning_rate": 3.714458947683519e-06,
      "loss": 0.3873,
      "mean_token_accuracy": 0.8586672931909561,
      "num_tokens": 180035450.0,
      "step": 149640
    },
    {
      "entropy": 1.9173126935958862,
      "epoch": 0.4639019063686927,
      "grad_norm": 9.367176055908203,
      "learning_rate": 3.714334839904563e-06,
      "loss": 0.4133,
      "mean_token_accuracy": 0.8661222815513611,
      "num_tokens": 180046863.0,
      "step": 149650
    },
    {
      "entropy": 1.8740798771381377,
      "epoch": 0.46393290549374244,
      "grad_norm": 8.433083534240723,
      "learning_rate": 3.7142107445648706e-06,
      "loss": 0.4236,
      "mean_token_accuracy": 0.8610092878341675,
      "num_tokens": 180059176.0,
      "step": 149660
    },
    {
      "entropy": 1.9137323945760727,
      "epoch": 0.4639639046187921,
      "grad_norm": 8.501178741455078,
      "learning_rate": 3.714086661662364e-06,
      "loss": 0.443,
      "mean_token_accuracy": 0.8547899782657623,
      "num_tokens": 180071303.0,
      "step": 149670
    },
    {
      "entropy": 1.8415597707033158,
      "epoch": 0.46399490374384184,
      "grad_norm": 9.41646957397461,
      "learning_rate": 3.713962591194966e-06,
      "loss": 0.4168,
      "mean_token_accuracy": 0.8580908700823784,
      "num_tokens": 180084425.0,
      "step": 149680
    },
    {
      "entropy": 1.8512480854988098,
      "epoch": 0.4640259028688915,
      "grad_norm": 4.0245442390441895,
      "learning_rate": 3.7138385331606e-06,
      "loss": 0.3873,
      "mean_token_accuracy": 0.8679956078529358,
      "num_tokens": 180097243.0,
      "step": 149690
    },
    {
      "entropy": 1.908530332148075,
      "epoch": 0.46405690199394123,
      "grad_norm": 8.557743072509766,
      "learning_rate": 3.713714487557189e-06,
      "loss": 0.4529,
      "mean_token_accuracy": 0.852016007900238,
      "num_tokens": 180108962.0,
      "step": 149700
    },
    {
      "entropy": 1.8267366781830787,
      "epoch": 0.4640879011189909,
      "grad_norm": 4.43588924407959,
      "learning_rate": 3.713590454382658e-06,
      "loss": 0.4163,
      "mean_token_accuracy": 0.8512490093708038,
      "num_tokens": 180121542.0,
      "step": 149710
    },
    {
      "entropy": 1.8044231168925762,
      "epoch": 0.4641189002440406,
      "grad_norm": 8.573354721069336,
      "learning_rate": 3.71346643363493e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.8628219202160835,
      "num_tokens": 180134671.0,
      "step": 149720
    },
    {
      "entropy": 1.9123155683279038,
      "epoch": 0.4641498993690903,
      "grad_norm": 3.89583158493042,
      "learning_rate": 3.7133424253119323e-06,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8472074523568154,
      "num_tokens": 180146781.0,
      "step": 149730
    },
    {
      "entropy": 1.9491783201694488,
      "epoch": 0.46418089849414,
      "grad_norm": 7.883488655090332,
      "learning_rate": 3.713218429411589e-06,
      "loss": 0.5038,
      "mean_token_accuracy": 0.8498224407434464,
      "num_tokens": 180158375.0,
      "step": 149740
    },
    {
      "entropy": 1.8938617929816246,
      "epoch": 0.4642118976191897,
      "grad_norm": 7.663883209228516,
      "learning_rate": 3.713094445931827e-06,
      "loss": 0.4595,
      "mean_token_accuracy": 0.8473029389977456,
      "num_tokens": 180170450.0,
      "step": 149750
    },
    {
      "entropy": 1.8819366082549096,
      "epoch": 0.4642428967442394,
      "grad_norm": 8.43736457824707,
      "learning_rate": 3.712970474870572e-06,
      "loss": 0.4776,
      "mean_token_accuracy": 0.8485167771577835,
      "num_tokens": 180182097.0,
      "step": 149760
    },
    {
      "entropy": 1.8832198321819305,
      "epoch": 0.4642738958692891,
      "grad_norm": 6.826776504516602,
      "learning_rate": 3.7128465162257517e-06,
      "loss": 0.4203,
      "mean_token_accuracy": 0.8600822359323501,
      "num_tokens": 180194184.0,
      "step": 149770
    },
    {
      "entropy": 1.9311166375875473,
      "epoch": 0.4643048949943388,
      "grad_norm": 8.055139541625977,
      "learning_rate": 3.712722569995293e-06,
      "loss": 0.4818,
      "mean_token_accuracy": 0.850020831823349,
      "num_tokens": 180205786.0,
      "step": 149780
    },
    {
      "entropy": 1.9121816590428353,
      "epoch": 0.46433589411938847,
      "grad_norm": 8.332880973815918,
      "learning_rate": 3.712598636177124e-06,
      "loss": 0.4429,
      "mean_token_accuracy": 0.8503255888819694,
      "num_tokens": 180218001.0,
      "step": 149790
    },
    {
      "entropy": 1.9609209045767784,
      "epoch": 0.4643668932444382,
      "grad_norm": 4.948374271392822,
      "learning_rate": 3.7124747147691733e-06,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8423975124955178,
      "num_tokens": 180229921.0,
      "step": 149800
    },
    {
      "entropy": 1.9410631626844406,
      "epoch": 0.46439789236948786,
      "grad_norm": 8.004473686218262,
      "learning_rate": 3.71235080576937e-06,
      "loss": 0.4916,
      "mean_token_accuracy": 0.8438965693116188,
      "num_tokens": 180242344.0,
      "step": 149810
    },
    {
      "entropy": 1.8821017548441887,
      "epoch": 0.4644288914945376,
      "grad_norm": 5.065506935119629,
      "learning_rate": 3.7122269091756436e-06,
      "loss": 0.4566,
      "mean_token_accuracy": 0.8540616199374199,
      "num_tokens": 180254354.0,
      "step": 149820
    },
    {
      "entropy": 1.9584839269518852,
      "epoch": 0.46445989061958726,
      "grad_norm": 7.649283409118652,
      "learning_rate": 3.7121030249859243e-06,
      "loss": 0.512,
      "mean_token_accuracy": 0.8388520136475563,
      "num_tokens": 180265796.0,
      "step": 149830
    },
    {
      "entropy": 1.8087536610662938,
      "epoch": 0.464490889744637,
      "grad_norm": 4.455826282501221,
      "learning_rate": 3.7119791531981408e-06,
      "loss": 0.3759,
      "mean_token_accuracy": 0.8636887416243553,
      "num_tokens": 180279611.0,
      "step": 149840
    },
    {
      "entropy": 1.9264666765928269,
      "epoch": 0.46452188886968665,
      "grad_norm": 7.5778985023498535,
      "learning_rate": 3.711855293810227e-06,
      "loss": 0.479,
      "mean_token_accuracy": 0.8526265174150467,
      "num_tokens": 180290829.0,
      "step": 149850
    },
    {
      "entropy": 1.9072339527308941,
      "epoch": 0.4645528879947364,
      "grad_norm": 6.993530750274658,
      "learning_rate": 3.711731446820111e-06,
      "loss": 0.4546,
      "mean_token_accuracy": 0.8573938056826591,
      "num_tokens": 180303038.0,
      "step": 149860
    },
    {
      "entropy": 1.9101138547062875,
      "epoch": 0.46458388711978604,
      "grad_norm": 8.588576316833496,
      "learning_rate": 3.7116076122257273e-06,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8481775879859924,
      "num_tokens": 180314807.0,
      "step": 149870
    },
    {
      "entropy": 1.7950350888073445,
      "epoch": 0.4646148862448357,
      "grad_norm": 4.491353511810303,
      "learning_rate": 3.7114837900250068e-06,
      "loss": 0.4137,
      "mean_token_accuracy": 0.8682404264807702,
      "num_tokens": 180328419.0,
      "step": 149880
    },
    {
      "entropy": 1.9566629469394683,
      "epoch": 0.46464588536988544,
      "grad_norm": 7.9985198974609375,
      "learning_rate": 3.7113599802158823e-06,
      "loss": 0.4799,
      "mean_token_accuracy": 0.8541330322623253,
      "num_tokens": 180339648.0,
      "step": 149890
    },
    {
      "entropy": 1.9067189604043961,
      "epoch": 0.4646768844949351,
      "grad_norm": 7.05324649810791,
      "learning_rate": 3.711236182796288e-06,
      "loss": 0.427,
      "mean_token_accuracy": 0.8619942396879197,
      "num_tokens": 180351473.0,
      "step": 149900
    },
    {
      "entropy": 1.9476832181215287,
      "epoch": 0.46470788361998483,
      "grad_norm": 3.14349365234375,
      "learning_rate": 3.711112397764157e-06,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8488188147544861,
      "num_tokens": 180362723.0,
      "step": 149910
    },
    {
      "entropy": 1.7743832789361478,
      "epoch": 0.4647388827450345,
      "grad_norm": 7.153512001037598,
      "learning_rate": 3.7109886251174236e-06,
      "loss": 0.3569,
      "mean_token_accuracy": 0.8677819207310676,
      "num_tokens": 180376546.0,
      "step": 149920
    },
    {
      "entropy": 1.8927229553461076,
      "epoch": 0.4647698818700842,
      "grad_norm": 2.545341968536377,
      "learning_rate": 3.710864864854023e-06,
      "loss": 0.4332,
      "mean_token_accuracy": 0.8583212435245514,
      "num_tokens": 180388347.0,
      "step": 149930
    },
    {
      "entropy": 1.8385560512542725,
      "epoch": 0.4648008809951339,
      "grad_norm": 7.90435791015625,
      "learning_rate": 3.71074111697189e-06,
      "loss": 0.4113,
      "mean_token_accuracy": 0.8564158409833909,
      "num_tokens": 180400665.0,
      "step": 149940
    },
    {
      "entropy": 1.8350412741303443,
      "epoch": 0.4648318801201836,
      "grad_norm": 4.03927755355835,
      "learning_rate": 3.7106173814689606e-06,
      "loss": 0.4009,
      "mean_token_accuracy": 0.8569675713777543,
      "num_tokens": 180412723.0,
      "step": 149950
    },
    {
      "entropy": 1.8664705529808998,
      "epoch": 0.4648628792452333,
      "grad_norm": 2.9268198013305664,
      "learning_rate": 3.710493658343171e-06,
      "loss": 0.5036,
      "mean_token_accuracy": 0.8416856169700623,
      "num_tokens": 180426131.0,
      "step": 149960
    },
    {
      "entropy": 1.922463881969452,
      "epoch": 0.464893878370283,
      "grad_norm": 7.96071195602417,
      "learning_rate": 3.7103699475924576e-06,
      "loss": 0.4839,
      "mean_token_accuracy": 0.844486691057682,
      "num_tokens": 180437621.0,
      "step": 149970
    },
    {
      "entropy": 1.8738661885261536,
      "epoch": 0.4649248774953327,
      "grad_norm": 7.6994404792785645,
      "learning_rate": 3.710246249214757e-06,
      "loss": 0.4296,
      "mean_token_accuracy": 0.859145550429821,
      "num_tokens": 180449886.0,
      "step": 149980
    },
    {
      "entropy": 1.854253427684307,
      "epoch": 0.4649558766203824,
      "grad_norm": 3.478299856185913,
      "learning_rate": 3.7101225632080085e-06,
      "loss": 0.4088,
      "mean_token_accuracy": 0.8661620482802391,
      "num_tokens": 180461919.0,
      "step": 149990
    },
    {
      "entropy": 1.9357280641794206,
      "epoch": 0.46498687574543207,
      "grad_norm": 10.878789901733398,
      "learning_rate": 3.709998889570149e-06,
      "loss": 0.464,
      "mean_token_accuracy": 0.8454708248376847,
      "num_tokens": 180473902.0,
      "step": 150000
    }
  ],
  "logging_steps": 10,
  "max_steps": 258072,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.432957984332841e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}