{
  "best_global_step": 6000,
  "best_metric": 0.8987749218940735,
  "best_model_checkpoint": "/output/checkpoint-6000",
  "epoch": 1.0,
  "eval_steps": 1000,
  "global_step": 6875,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.9659422233700752,
      "epoch": 0.0014545454545454545,
      "grad_norm": 4.090943813323975,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 3.0094,
      "mean_token_accuracy": 0.5692892394959926,
      "num_tokens": 77676.0,
      "step": 10
    },
    {
      "entropy": 1.9524168580770493,
      "epoch": 0.002909090909090909,
      "grad_norm": 3.8272783756256104,
      "learning_rate": 6.333333333333333e-07,
      "loss": 3.0425,
      "mean_token_accuracy": 0.5691409226506948,
      "num_tokens": 152469.0,
      "step": 20
    },
    {
      "entropy": 1.9598139539361,
      "epoch": 0.004363636363636364,
      "grad_norm": 3.7499046325683594,
      "learning_rate": 9.666666666666668e-07,
      "loss": 2.9915,
      "mean_token_accuracy": 0.5720876976847649,
      "num_tokens": 229946.0,
      "step": 30
    },
    {
      "entropy": 1.9616472586989402,
      "epoch": 0.005818181818181818,
      "grad_norm": 3.472123622894287,
      "learning_rate": 1.3e-06,
      "loss": 2.9656,
      "mean_token_accuracy": 0.5695729449391365,
      "num_tokens": 306785.0,
      "step": 40
    },
    {
      "entropy": 1.983330498635769,
      "epoch": 0.007272727272727273,
      "grad_norm": 3.5334620475769043,
      "learning_rate": 1.6333333333333333e-06,
      "loss": 2.8909,
      "mean_token_accuracy": 0.569761898368597,
      "num_tokens": 384553.0,
      "step": 50
    },
    {
      "entropy": 1.9984836801886559,
      "epoch": 0.008727272727272728,
      "grad_norm": 3.185563087463379,
      "learning_rate": 1.9666666666666668e-06,
      "loss": 2.7998,
      "mean_token_accuracy": 0.5736399229615927,
      "num_tokens": 461373.0,
      "step": 60
    },
    {
      "entropy": 1.979162333905697,
      "epoch": 0.010181818181818183,
      "grad_norm": 2.248973846435547,
      "learning_rate": 2.3e-06,
      "loss": 2.6071,
      "mean_token_accuracy": 0.5999085880815983,
      "num_tokens": 538377.0,
      "step": 70
    },
    {
      "entropy": 1.9361985102295876,
      "epoch": 0.011636363636363636,
      "grad_norm": 1.7401518821716309,
      "learning_rate": 2.6333333333333337e-06,
      "loss": 2.4549,
      "mean_token_accuracy": 0.6177365817129612,
      "num_tokens": 615542.0,
      "step": 80
    },
    {
      "entropy": 1.8896187946200371,
      "epoch": 0.01309090909090909,
      "grad_norm": 1.7699965238571167,
      "learning_rate": 2.966666666666667e-06,
      "loss": 2.3215,
      "mean_token_accuracy": 0.6456378914415837,
      "num_tokens": 692214.0,
      "step": 90
    },
    {
      "entropy": 1.837496855854988,
      "epoch": 0.014545454545454545,
      "grad_norm": 1.7701467275619507,
      "learning_rate": 3.3e-06,
      "loss": 2.1539,
      "mean_token_accuracy": 0.6737071752548218,
      "num_tokens": 768454.0,
      "step": 100
    },
    {
      "entropy": 1.7528666377067565,
      "epoch": 0.016,
      "grad_norm": 1.9286140203475952,
      "learning_rate": 3.633333333333334e-06,
      "loss": 1.9427,
      "mean_token_accuracy": 0.7077549912035466,
      "num_tokens": 844867.0,
      "step": 110
    },
    {
      "entropy": 1.6664264917373657,
      "epoch": 0.017454545454545455,
      "grad_norm": 1.7424260377883911,
      "learning_rate": 3.966666666666667e-06,
      "loss": 1.7671,
      "mean_token_accuracy": 0.7324294425547123,
      "num_tokens": 922240.0,
      "step": 120
    },
    {
      "entropy": 1.4373633116483688,
      "epoch": 0.01890909090909091,
      "grad_norm": 1.6519197225570679,
      "learning_rate": 4.2999999999999995e-06,
      "loss": 1.5648,
      "mean_token_accuracy": 0.7576472572982311,
      "num_tokens": 995800.0,
      "step": 130
    },
    {
      "entropy": 1.2589309319853783,
      "epoch": 0.020363636363636365,
      "grad_norm": 0.7197794318199158,
      "learning_rate": 4.633333333333334e-06,
      "loss": 1.4726,
      "mean_token_accuracy": 0.7749086625874042,
      "num_tokens": 1071953.0,
      "step": 140
    },
    {
      "entropy": 1.1532729104161263,
      "epoch": 0.02181818181818182,
      "grad_norm": 0.5631716847419739,
      "learning_rate": 4.966666666666667e-06,
      "loss": 1.3865,
      "mean_token_accuracy": 0.7834423221647739,
      "num_tokens": 1148106.0,
      "step": 150
    },
    {
      "entropy": 1.1569199711084366,
      "epoch": 0.02327272727272727,
      "grad_norm": 0.6042912602424622,
      "learning_rate": 5.3e-06,
      "loss": 1.3816,
      "mean_token_accuracy": 0.7809078678488731,
      "num_tokens": 1226497.0,
      "step": 160
    },
    {
      "entropy": 1.135782204568386,
      "epoch": 0.024727272727272726,
      "grad_norm": 0.5481341481208801,
      "learning_rate": 5.633333333333333e-06,
      "loss": 1.3305,
      "mean_token_accuracy": 0.7899107336997986,
      "num_tokens": 1302834.0,
      "step": 170
    },
    {
      "entropy": 1.1304984226822854,
      "epoch": 0.02618181818181818,
      "grad_norm": 0.6024408936500549,
      "learning_rate": 5.9666666666666666e-06,
      "loss": 1.3141,
      "mean_token_accuracy": 0.791486781835556,
      "num_tokens": 1380407.0,
      "step": 180
    },
    {
      "entropy": 1.1483597584068774,
      "epoch": 0.027636363636363636,
      "grad_norm": 0.6185842752456665,
      "learning_rate": 6.300000000000001e-06,
      "loss": 1.3088,
      "mean_token_accuracy": 0.7890688091516495,
      "num_tokens": 1455335.0,
      "step": 190
    },
    {
      "entropy": 1.1483107447624206,
      "epoch": 0.02909090909090909,
      "grad_norm": 0.5758558511734009,
      "learning_rate": 6.633333333333333e-06,
      "loss": 1.3201,
      "mean_token_accuracy": 0.7872945092618465,
      "num_tokens": 1531470.0,
      "step": 200
    },
    {
      "entropy": 1.1816094659268856,
      "epoch": 0.030545454545454546,
      "grad_norm": 0.6477853655815125,
      "learning_rate": 6.966666666666667e-06,
      "loss": 1.3675,
      "mean_token_accuracy": 0.7840658474713564,
      "num_tokens": 1610425.0,
      "step": 210
    },
    {
      "entropy": 1.1367807120084763,
      "epoch": 0.032,
      "grad_norm": 0.6259286999702454,
      "learning_rate": 7.2999999999999996e-06,
      "loss": 1.2578,
      "mean_token_accuracy": 0.7919471129775048,
      "num_tokens": 1688549.0,
      "step": 220
    },
    {
      "entropy": 1.1455395191907882,
      "epoch": 0.03345454545454545,
      "grad_norm": 0.6879440546035767,
      "learning_rate": 7.633333333333334e-06,
      "loss": 1.2376,
      "mean_token_accuracy": 0.7930837780237198,
      "num_tokens": 1764440.0,
      "step": 230
    },
    {
      "entropy": 1.1699098110198975,
      "epoch": 0.03490909090909091,
      "grad_norm": 0.7712326645851135,
      "learning_rate": 7.966666666666666e-06,
      "loss": 1.2593,
      "mean_token_accuracy": 0.7901124738156795,
      "num_tokens": 1838579.0,
      "step": 240
    },
    {
      "entropy": 1.1891434825956821,
      "epoch": 0.03636363636363636,
      "grad_norm": 0.8381406664848328,
      "learning_rate": 8.3e-06,
      "loss": 1.1957,
      "mean_token_accuracy": 0.7920134104788303,
      "num_tokens": 1917117.0,
      "step": 250
    },
    {
      "entropy": 1.2279200583696366,
      "epoch": 0.03781818181818182,
      "grad_norm": 0.6233083605766296,
      "learning_rate": 8.633333333333334e-06,
      "loss": 1.1274,
      "mean_token_accuracy": 0.796196486800909,
      "num_tokens": 1992506.0,
      "step": 260
    },
    {
      "entropy": 1.2140305921435357,
      "epoch": 0.03927272727272727,
      "grad_norm": 0.5732085108757019,
      "learning_rate": 8.966666666666668e-06,
      "loss": 1.1105,
      "mean_token_accuracy": 0.8027118250727654,
      "num_tokens": 2069283.0,
      "step": 270
    },
    {
      "entropy": 1.2301497504115104,
      "epoch": 0.04072727272727273,
      "grad_norm": 0.6894718408584595,
      "learning_rate": 9.3e-06,
      "loss": 1.1371,
      "mean_token_accuracy": 0.7968612916767597,
      "num_tokens": 2146174.0,
      "step": 280
    },
    {
      "entropy": 1.1968476533889771,
      "epoch": 0.04218181818181818,
      "grad_norm": 0.583967387676239,
      "learning_rate": 9.633333333333335e-06,
      "loss": 1.0555,
      "mean_token_accuracy": 0.8039177753031254,
      "num_tokens": 2221425.0,
      "step": 290
    },
    {
      "entropy": 1.2118824765086174,
      "epoch": 0.04363636363636364,
      "grad_norm": 0.601884126663208,
      "learning_rate": 9.966666666666667e-06,
      "loss": 1.0708,
      "mean_token_accuracy": 0.799370014667511,
      "num_tokens": 2299095.0,
      "step": 300
    },
    {
      "entropy": 1.1940646573901177,
      "epoch": 0.04509090909090909,
      "grad_norm": 0.6982355713844299,
      "learning_rate": 1.03e-05,
      "loss": 1.0547,
      "mean_token_accuracy": 0.7977800592780113,
      "num_tokens": 2378075.0,
      "step": 310
    },
    {
      "entropy": 1.1546141751110555,
      "epoch": 0.04654545454545454,
      "grad_norm": 0.4953351616859436,
      "learning_rate": 1.0633333333333334e-05,
      "loss": 1.0335,
      "mean_token_accuracy": 0.7998595416545868,
      "num_tokens": 2455113.0,
      "step": 320
    },
    {
      "entropy": 1.1033454038202763,
      "epoch": 0.048,
      "grad_norm": 0.44239360094070435,
      "learning_rate": 1.0966666666666666e-05,
      "loss": 0.9985,
      "mean_token_accuracy": 0.8006901569664479,
      "num_tokens": 2531303.0,
      "step": 330
    },
    {
      "entropy": 1.0396368399262428,
      "epoch": 0.04945454545454545,
      "grad_norm": 0.42604953050613403,
      "learning_rate": 1.13e-05,
      "loss": 0.9764,
      "mean_token_accuracy": 0.8035364523530006,
      "num_tokens": 2608824.0,
      "step": 340
    },
    {
      "entropy": 1.0469321973621846,
      "epoch": 0.05090909090909091,
      "grad_norm": 0.49078768491744995,
      "learning_rate": 1.1633333333333334e-05,
      "loss": 1.019,
      "mean_token_accuracy": 0.7943111263215542,
      "num_tokens": 2686701.0,
      "step": 350
    },
    {
      "entropy": 1.0321501083672047,
      "epoch": 0.05236363636363636,
      "grad_norm": 0.5576013922691345,
      "learning_rate": 1.1966666666666668e-05,
      "loss": 1.005,
      "mean_token_accuracy": 0.7979445092380046,
      "num_tokens": 2764773.0,
      "step": 360
    },
    {
      "entropy": 0.990199426561594,
      "epoch": 0.05381818181818182,
      "grad_norm": 0.47645217180252075,
      "learning_rate": 1.23e-05,
      "loss": 0.9538,
      "mean_token_accuracy": 0.8067440554499626,
      "num_tokens": 2839781.0,
      "step": 370
    },
    {
      "entropy": 1.019500917941332,
      "epoch": 0.05527272727272727,
      "grad_norm": 0.4222368896007538,
      "learning_rate": 1.2633333333333333e-05,
      "loss": 0.9798,
      "mean_token_accuracy": 0.8025607392191887,
      "num_tokens": 2917887.0,
      "step": 380
    },
    {
      "entropy": 0.9895624630153179,
      "epoch": 0.05672727272727273,
      "grad_norm": 0.4202658236026764,
      "learning_rate": 1.2966666666666669e-05,
      "loss": 0.9658,
      "mean_token_accuracy": 0.8041856609284878,
      "num_tokens": 2996429.0,
      "step": 390
    },
    {
      "entropy": 1.0076496340334415,
      "epoch": 0.05818181818181818,
      "grad_norm": 0.42250972986221313,
      "learning_rate": 1.3300000000000001e-05,
      "loss": 0.9883,
      "mean_token_accuracy": 0.8030573949217796,
      "num_tokens": 3072279.0,
      "step": 400
    },
    {
      "entropy": 1.0123522110283374,
      "epoch": 0.05963636363636363,
      "grad_norm": 0.47815120220184326,
      "learning_rate": 1.3633333333333334e-05,
      "loss": 0.9774,
      "mean_token_accuracy": 0.8001051150262356,
      "num_tokens": 3151158.0,
      "step": 410
    },
    {
      "entropy": 1.0093131870031358,
      "epoch": 0.06109090909090909,
      "grad_norm": 0.5531169772148132,
      "learning_rate": 1.3966666666666666e-05,
      "loss": 0.9863,
      "mean_token_accuracy": 0.8017294406890869,
      "num_tokens": 3229828.0,
      "step": 420
    },
    {
      "entropy": 0.9861496143043041,
      "epoch": 0.06254545454545454,
      "grad_norm": 0.4163675308227539,
      "learning_rate": 1.43e-05,
      "loss": 0.9567,
      "mean_token_accuracy": 0.8050003960728646,
      "num_tokens": 3306024.0,
      "step": 430
    },
    {
      "entropy": 0.9772120468318463,
      "epoch": 0.064,
      "grad_norm": 0.46380943059921265,
      "learning_rate": 1.4633333333333334e-05,
      "loss": 0.9416,
      "mean_token_accuracy": 0.8080964677035809,
      "num_tokens": 3380727.0,
      "step": 440
    },
    {
      "entropy": 1.007182703912258,
      "epoch": 0.06545454545454546,
      "grad_norm": 0.5575771927833557,
      "learning_rate": 1.4966666666666668e-05,
      "loss": 0.9776,
      "mean_token_accuracy": 0.8016268402338028,
      "num_tokens": 3458256.0,
      "step": 450
    },
    {
      "entropy": 1.0141797795891763,
      "epoch": 0.0669090909090909,
      "grad_norm": 0.4321994483470917,
      "learning_rate": 1.53e-05,
      "loss": 0.9744,
      "mean_token_accuracy": 0.803185262531042,
      "num_tokens": 3535663.0,
      "step": 460
    },
    {
      "entropy": 0.9918124876916409,
      "epoch": 0.06836363636363636,
      "grad_norm": 0.435546338558197,
      "learning_rate": 1.563333333333333e-05,
      "loss": 0.967,
      "mean_token_accuracy": 0.8032719291746616,
      "num_tokens": 3614253.0,
      "step": 470
    },
    {
      "entropy": 0.9891813933849335,
      "epoch": 0.06981818181818182,
      "grad_norm": 0.4258497357368469,
      "learning_rate": 1.5966666666666667e-05,
      "loss": 0.9618,
      "mean_token_accuracy": 0.8035177089273929,
      "num_tokens": 3690297.0,
      "step": 480
    },
    {
      "entropy": 1.006697028875351,
      "epoch": 0.07127272727272728,
      "grad_norm": 0.43594589829444885,
      "learning_rate": 1.63e-05,
      "loss": 0.9726,
      "mean_token_accuracy": 0.8012183904647827,
      "num_tokens": 3769149.0,
      "step": 490
    },
    {
      "entropy": 1.0048870131373406,
      "epoch": 0.07272727272727272,
      "grad_norm": 0.415245920419693,
      "learning_rate": 1.6633333333333336e-05,
      "loss": 0.9933,
      "mean_token_accuracy": 0.8001591399312019,
      "num_tokens": 3846637.0,
      "step": 500
    },
    {
      "entropy": 1.005825261026621,
      "epoch": 0.07418181818181818,
      "grad_norm": 0.4039178788661957,
      "learning_rate": 1.6966666666666668e-05,
      "loss": 0.9723,
      "mean_token_accuracy": 0.8034198030829429,
      "num_tokens": 3924639.0,
      "step": 510
    },
    {
      "entropy": 1.0154225043952465,
      "epoch": 0.07563636363636364,
      "grad_norm": 0.44846075773239136,
      "learning_rate": 1.73e-05,
      "loss": 0.9906,
      "mean_token_accuracy": 0.8011446699500084,
      "num_tokens": 4000254.0,
      "step": 520
    },
    {
      "entropy": 0.9866499230265617,
      "epoch": 0.07709090909090908,
      "grad_norm": 0.4658671021461487,
      "learning_rate": 1.7633333333333336e-05,
      "loss": 0.9539,
      "mean_token_accuracy": 0.8053110785782337,
      "num_tokens": 4075047.0,
      "step": 530
    },
    {
      "entropy": 0.9868831895291805,
      "epoch": 0.07854545454545454,
      "grad_norm": 0.5777063965797424,
      "learning_rate": 1.796666666666667e-05,
      "loss": 0.9586,
      "mean_token_accuracy": 0.8030114769935608,
      "num_tokens": 4150280.0,
      "step": 540
    },
    {
      "entropy": 0.9865054927766324,
      "epoch": 0.08,
      "grad_norm": 0.48241281509399414,
      "learning_rate": 1.83e-05,
      "loss": 0.9674,
      "mean_token_accuracy": 0.804360181838274,
      "num_tokens": 4227448.0,
      "step": 550
    },
    {
      "entropy": 0.9722688443958759,
      "epoch": 0.08145454545454546,
      "grad_norm": 0.41982826590538025,
      "learning_rate": 1.8633333333333333e-05,
      "loss": 0.9522,
      "mean_token_accuracy": 0.8053691528737545,
      "num_tokens": 4303006.0,
      "step": 560
    },
    {
      "entropy": 0.9883188255131244,
      "epoch": 0.0829090909090909,
      "grad_norm": 0.5174263119697571,
      "learning_rate": 1.896666666666667e-05,
      "loss": 0.96,
      "mean_token_accuracy": 0.8065396010875702,
      "num_tokens": 4377210.0,
      "step": 570
    },
    {
      "entropy": 0.9758467309176921,
      "epoch": 0.08436363636363636,
      "grad_norm": 0.4374963343143463,
      "learning_rate": 1.93e-05,
      "loss": 0.9476,
      "mean_token_accuracy": 0.8069633848965168,
      "num_tokens": 4452563.0,
      "step": 580
    },
    {
      "entropy": 0.9731172017753125,
      "epoch": 0.08581818181818182,
      "grad_norm": 0.4287698268890381,
      "learning_rate": 1.9633333333333334e-05,
      "loss": 0.957,
      "mean_token_accuracy": 0.8069525636732578,
      "num_tokens": 4529201.0,
      "step": 590
    },
    {
      "entropy": 1.0023035056889058,
      "epoch": 0.08727272727272728,
      "grad_norm": 0.4412376880645752,
      "learning_rate": 1.9966666666666666e-05,
      "loss": 0.9748,
      "mean_token_accuracy": 0.8018463432788849,
      "num_tokens": 4607606.0,
      "step": 600
    },
    {
      "entropy": 1.0000741586089135,
      "epoch": 0.08872727272727272,
      "grad_norm": 0.452994704246521,
      "learning_rate": 2.0300000000000002e-05,
      "loss": 0.9684,
      "mean_token_accuracy": 0.804426409304142,
      "num_tokens": 4684814.0,
      "step": 610
    },
    {
      "entropy": 1.003422600775957,
      "epoch": 0.09018181818181818,
      "grad_norm": 0.5886366367340088,
      "learning_rate": 2.0633333333333335e-05,
      "loss": 0.9772,
      "mean_token_accuracy": 0.8043254964053631,
      "num_tokens": 4762034.0,
      "step": 620
    },
    {
      "entropy": 0.971302555501461,
      "epoch": 0.09163636363636364,
      "grad_norm": 0.5860087871551514,
      "learning_rate": 2.0966666666666667e-05,
      "loss": 0.9572,
      "mean_token_accuracy": 0.8045364588499069,
      "num_tokens": 4839421.0,
      "step": 630
    },
    {
      "entropy": 1.0108731143176555,
      "epoch": 0.09309090909090909,
      "grad_norm": 0.47347405552864075,
      "learning_rate": 2.13e-05,
      "loss": 0.9706,
      "mean_token_accuracy": 0.8029872097074986,
      "num_tokens": 4915456.0,
      "step": 640
    },
    {
      "entropy": 1.0099310785531999,
      "epoch": 0.09454545454545454,
      "grad_norm": 0.5390433669090271,
      "learning_rate": 2.1633333333333332e-05,
      "loss": 0.9743,
      "mean_token_accuracy": 0.8021445274353027,
      "num_tokens": 4991780.0,
      "step": 650
    },
    {
      "entropy": 1.006123662739992,
      "epoch": 0.096,
      "grad_norm": 0.4339490532875061,
      "learning_rate": 2.1966666666666668e-05,
      "loss": 0.9833,
      "mean_token_accuracy": 0.8044153302907944,
      "num_tokens": 5066164.0,
      "step": 660
    },
    {
      "entropy": 0.9823346547782421,
      "epoch": 0.09745454545454546,
      "grad_norm": 0.4419339597225189,
      "learning_rate": 2.23e-05,
      "loss": 0.9593,
      "mean_token_accuracy": 0.8065159559249878,
      "num_tokens": 5141318.0,
      "step": 670
    },
    {
      "entropy": 0.9649036668241024,
      "epoch": 0.0989090909090909,
      "grad_norm": 0.5108367204666138,
      "learning_rate": 2.2633333333333336e-05,
      "loss": 0.9377,
      "mean_token_accuracy": 0.8067274354398251,
      "num_tokens": 5217622.0,
      "step": 680
    },
    {
      "entropy": 0.976323938369751,
      "epoch": 0.10036363636363636,
      "grad_norm": 0.4640137553215027,
      "learning_rate": 2.2966666666666668e-05,
      "loss": 0.9438,
      "mean_token_accuracy": 0.8051302477717399,
      "num_tokens": 5295472.0,
      "step": 690
    },
    {
      "entropy": 0.9709480322897435,
      "epoch": 0.10181818181818182,
      "grad_norm": 0.5125411152839661,
      "learning_rate": 2.3300000000000004e-05,
      "loss": 0.9338,
      "mean_token_accuracy": 0.8084210947155952,
      "num_tokens": 5370567.0,
      "step": 700
    },
    {
      "entropy": 0.9817489549517632,
      "epoch": 0.10327272727272727,
      "grad_norm": 0.43162453174591064,
      "learning_rate": 2.3633333333333336e-05,
      "loss": 0.9532,
      "mean_token_accuracy": 0.8055648073554039,
      "num_tokens": 5448305.0,
      "step": 710
    },
    {
      "entropy": 0.9898195452988148,
      "epoch": 0.10472727272727272,
      "grad_norm": 0.4311891496181488,
      "learning_rate": 2.396666666666667e-05,
      "loss": 0.9598,
      "mean_token_accuracy": 0.8042900271713733,
      "num_tokens": 5526145.0,
      "step": 720
    },
    {
      "entropy": 0.9635106429457665,
      "epoch": 0.10618181818181818,
      "grad_norm": 0.41602715849876404,
      "learning_rate": 2.43e-05,
      "loss": 0.9278,
      "mean_token_accuracy": 0.8106991343200207,
      "num_tokens": 5601880.0,
      "step": 730
    },
    {
      "entropy": 0.9739560626447201,
      "epoch": 0.10763636363636364,
      "grad_norm": 0.48207396268844604,
      "learning_rate": 2.4633333333333334e-05,
      "loss": 0.9408,
      "mean_token_accuracy": 0.8064304023981095,
      "num_tokens": 5678740.0,
      "step": 740
    },
    {
      "entropy": 0.9841963745653629,
      "epoch": 0.10909090909090909,
      "grad_norm": 0.4250541031360626,
      "learning_rate": 2.496666666666667e-05,
      "loss": 0.9615,
      "mean_token_accuracy": 0.8043922528624534,
      "num_tokens": 5757834.0,
      "step": 750
    },
    {
      "entropy": 0.9840492263436318,
      "epoch": 0.11054545454545454,
      "grad_norm": 0.42034247517585754,
      "learning_rate": 2.5300000000000002e-05,
      "loss": 0.9545,
      "mean_token_accuracy": 0.8051097564399242,
      "num_tokens": 5835143.0,
      "step": 760
    },
    {
      "entropy": 0.9977390602231025,
      "epoch": 0.112,
      "grad_norm": 0.4027940630912781,
      "learning_rate": 2.5633333333333338e-05,
      "loss": 0.9693,
      "mean_token_accuracy": 0.8035439133644104,
      "num_tokens": 5914748.0,
      "step": 770
    },
    {
      "entropy": 0.9808847881853581,
      "epoch": 0.11345454545454546,
      "grad_norm": 0.40575113892555237,
      "learning_rate": 2.5966666666666667e-05,
      "loss": 0.9457,
      "mean_token_accuracy": 0.8073798060417176,
      "num_tokens": 5989838.0,
      "step": 780
    },
    {
      "entropy": 0.9797362260520458,
      "epoch": 0.1149090909090909,
      "grad_norm": 0.4026121199131012,
      "learning_rate": 2.6300000000000002e-05,
      "loss": 0.9467,
      "mean_token_accuracy": 0.8068944476544857,
      "num_tokens": 6064792.0,
      "step": 790
    },
    {
      "entropy": 0.9880223348736763,
      "epoch": 0.11636363636363636,
      "grad_norm": 0.42550402879714966,
      "learning_rate": 2.663333333333333e-05,
      "loss": 0.971,
      "mean_token_accuracy": 0.8018131107091904,
      "num_tokens": 6143005.0,
      "step": 800
    },
    {
      "entropy": 0.9746092438697815,
      "epoch": 0.11781818181818182,
      "grad_norm": 0.3958792984485626,
      "learning_rate": 2.6966666666666667e-05,
      "loss": 0.9256,
      "mean_token_accuracy": 0.8095438420772553,
      "num_tokens": 6216947.0,
      "step": 810
    },
    {
      "entropy": 0.9785576783120632,
      "epoch": 0.11927272727272727,
      "grad_norm": 0.46700361371040344,
      "learning_rate": 2.7300000000000003e-05,
      "loss": 0.9592,
      "mean_token_accuracy": 0.8056280180811882,
      "num_tokens": 6292720.0,
      "step": 820
    },
    {
      "entropy": 0.9925856553018093,
      "epoch": 0.12072727272727272,
      "grad_norm": 0.4162468910217285,
      "learning_rate": 2.7633333333333332e-05,
      "loss": 0.9684,
      "mean_token_accuracy": 0.8065006777644157,
      "num_tokens": 6369195.0,
      "step": 830
    },
    {
      "entropy": 0.9815856114029884,
      "epoch": 0.12218181818181818,
      "grad_norm": 0.4176236391067505,
      "learning_rate": 2.7966666666666668e-05,
      "loss": 0.9417,
      "mean_token_accuracy": 0.806631051003933,
      "num_tokens": 6444477.0,
      "step": 840
    },
    {
      "entropy": 0.9838737137615681,
      "epoch": 0.12363636363636364,
      "grad_norm": 0.5555691123008728,
      "learning_rate": 2.83e-05,
      "loss": 0.9631,
      "mean_token_accuracy": 0.8037988044321537,
      "num_tokens": 6524429.0,
      "step": 850
    },
    {
      "entropy": 0.9618504159152508,
      "epoch": 0.12509090909090909,
      "grad_norm": 0.41715553402900696,
      "learning_rate": 2.8633333333333336e-05,
      "loss": 0.9111,
      "mean_token_accuracy": 0.8102175794541836,
      "num_tokens": 6598540.0,
      "step": 860
    },
    {
      "entropy": 0.9729882068932056,
      "epoch": 0.12654545454545454,
      "grad_norm": 0.43373847007751465,
      "learning_rate": 2.8966666666666668e-05,
      "loss": 0.9406,
      "mean_token_accuracy": 0.8060924611985684,
      "num_tokens": 6675725.0,
      "step": 870
    },
    {
      "entropy": 0.9947902150452137,
      "epoch": 0.128,
      "grad_norm": 0.43133389949798584,
      "learning_rate": 2.93e-05,
      "loss": 0.9728,
      "mean_token_accuracy": 0.8022032633423806,
      "num_tokens": 6754713.0,
      "step": 880
    },
    {
      "entropy": 0.995215217769146,
      "epoch": 0.12945454545454546,
      "grad_norm": 0.604076087474823,
      "learning_rate": 2.9633333333333336e-05,
      "loss": 0.9482,
      "mean_token_accuracy": 0.8048032231628894,
      "num_tokens": 6830667.0,
      "step": 890
    },
    {
      "entropy": 0.9664588801562786,
      "epoch": 0.13090909090909092,
      "grad_norm": 0.4081405997276306,
      "learning_rate": 2.9966666666666672e-05,
      "loss": 0.9472,
      "mean_token_accuracy": 0.8069540120661258,
      "num_tokens": 6905611.0,
      "step": 900
    },
    {
      "entropy": 0.994405622035265,
      "epoch": 0.13236363636363635,
      "grad_norm": 0.5211614966392517,
      "learning_rate": 3.03e-05,
      "loss": 0.9581,
      "mean_token_accuracy": 0.8033522762358188,
      "num_tokens": 6985193.0,
      "step": 910
    },
    {
      "entropy": 0.9925968185067177,
      "epoch": 0.1338181818181818,
      "grad_norm": 0.4536610543727875,
      "learning_rate": 3.063333333333334e-05,
      "loss": 0.99,
      "mean_token_accuracy": 0.8044977240264416,
      "num_tokens": 7061179.0,
      "step": 920
    },
    {
      "entropy": 0.9665735505521298,
      "epoch": 0.13527272727272727,
      "grad_norm": 0.4118134677410126,
      "learning_rate": 3.096666666666666e-05,
      "loss": 0.9331,
      "mean_token_accuracy": 0.8096266977488995,
      "num_tokens": 7136176.0,
      "step": 930
    },
    {
      "entropy": 0.9744052506983281,
      "epoch": 0.13672727272727273,
      "grad_norm": 0.5332080125808716,
      "learning_rate": 3.13e-05,
      "loss": 0.9528,
      "mean_token_accuracy": 0.8053979344666005,
      "num_tokens": 7213493.0,
      "step": 940
    },
    {
      "entropy": 0.9805201821029186,
      "epoch": 0.13818181818181818,
      "grad_norm": 0.4272365868091583,
      "learning_rate": 3.1633333333333334e-05,
      "loss": 0.9382,
      "mean_token_accuracy": 0.8079843915998935,
      "num_tokens": 7289917.0,
      "step": 950
    },
    {
      "entropy": 0.9595341965556144,
      "epoch": 0.13963636363636364,
      "grad_norm": 0.4467701017856598,
      "learning_rate": 3.196666666666667e-05,
      "loss": 0.9312,
      "mean_token_accuracy": 0.8084880888462067,
      "num_tokens": 7367477.0,
      "step": 960
    },
    {
      "entropy": 0.9859929539263248,
      "epoch": 0.1410909090909091,
      "grad_norm": 0.462103933095932,
      "learning_rate": 3.2300000000000006e-05,
      "loss": 0.9524,
      "mean_token_accuracy": 0.804368419200182,
      "num_tokens": 7444578.0,
      "step": 970
    },
    {
      "entropy": 0.9547855578362942,
      "epoch": 0.14254545454545456,
      "grad_norm": 0.432309091091156,
      "learning_rate": 3.263333333333333e-05,
      "loss": 0.9258,
      "mean_token_accuracy": 0.8119194991886616,
      "num_tokens": 7517378.0,
      "step": 980
    },
    {
      "entropy": 0.9911278739571572,
      "epoch": 0.144,
      "grad_norm": 0.40271005034446716,
      "learning_rate": 3.296666666666667e-05,
      "loss": 0.9613,
      "mean_token_accuracy": 0.8036044925451279,
      "num_tokens": 7595008.0,
      "step": 990
    },
    {
      "entropy": 0.9437328346073628,
      "epoch": 0.14545454545454545,
      "grad_norm": 0.4583914875984192,
      "learning_rate": 3.33e-05,
      "loss": 0.9148,
      "mean_token_accuracy": 0.8127484254539012,
      "num_tokens": 7669624.0,
      "step": 1000
    },
    {
      "epoch": 0.14545454545454545,
      "eval_entropy": 0.9819960155487061,
      "eval_loss": 0.930533230304718,
      "eval_mean_token_accuracy": 0.8069609030723571,
      "eval_num_tokens": 7669624.0,
      "eval_runtime": 77.1409,
      "eval_samples_per_second": 64.816,
      "eval_steps_per_second": 8.102,
      "step": 1000
    },
    {
      "entropy": 0.9614314928650856,
      "epoch": 0.1469090909090909,
      "grad_norm": 0.41151049733161926,
      "learning_rate": 3.3633333333333335e-05,
      "loss": 0.928,
      "mean_token_accuracy": 0.8098596774041653,
      "num_tokens": 7745451.0,
      "step": 1010
    },
    {
      "entropy": 0.9788591407239438,
      "epoch": 0.14836363636363636,
      "grad_norm": 0.46174564957618713,
      "learning_rate": 3.396666666666667e-05,
      "loss": 0.9353,
      "mean_token_accuracy": 0.8068431109189987,
      "num_tokens": 7824306.0,
      "step": 1020
    },
    {
      "entropy": 0.9697593882679939,
      "epoch": 0.14981818181818182,
      "grad_norm": 0.41812098026275635,
      "learning_rate": 3.430000000000001e-05,
      "loss": 0.9549,
      "mean_token_accuracy": 0.8075274929404259,
      "num_tokens": 7900392.0,
      "step": 1030
    },
    {
      "entropy": 0.983103808760643,
      "epoch": 0.15127272727272728,
      "grad_norm": 0.4310772120952606,
      "learning_rate": 3.463333333333333e-05,
      "loss": 0.9581,
      "mean_token_accuracy": 0.8044828698039055,
      "num_tokens": 7977471.0,
      "step": 1040
    },
    {
      "entropy": 0.9643499620258809,
      "epoch": 0.15272727272727274,
      "grad_norm": 0.4172477126121521,
      "learning_rate": 3.496666666666667e-05,
      "loss": 0.9328,
      "mean_token_accuracy": 0.8084025174379349,
      "num_tokens": 8054025.0,
      "step": 1050
    },
    {
      "entropy": 0.9781779557466507,
      "epoch": 0.15418181818181817,
      "grad_norm": 0.4708545506000519,
      "learning_rate": 3.53e-05,
      "loss": 0.9375,
      "mean_token_accuracy": 0.8080290637910366,
      "num_tokens": 8129830.0,
      "step": 1060
    },
    {
      "entropy": 0.97263238504529,
      "epoch": 0.15563636363636363,
      "grad_norm": 0.4220433533191681,
      "learning_rate": 3.563333333333334e-05,
      "loss": 0.9523,
      "mean_token_accuracy": 0.8069098271429539,
      "num_tokens": 8208871.0,
      "step": 1070
    },
    {
      "entropy": 0.9988827370107174,
      "epoch": 0.1570909090909091,
      "grad_norm": 0.4833963215351105,
      "learning_rate": 3.596666666666667e-05,
      "loss": 0.9781,
      "mean_token_accuracy": 0.8028389126062393,
      "num_tokens": 8285421.0,
      "step": 1080
    },
    {
      "entropy": 0.981606388837099,
      "epoch": 0.15854545454545454,
      "grad_norm": 0.4167567491531372,
      "learning_rate": 3.63e-05,
      "loss": 0.9592,
      "mean_token_accuracy": 0.8043723694980145,
      "num_tokens": 8364166.0,
      "step": 1090
    },
    {
      "entropy": 0.972727719694376,
      "epoch": 0.16,
      "grad_norm": 0.41988062858581543,
      "learning_rate": 3.6633333333333334e-05,
      "loss": 0.9582,
      "mean_token_accuracy": 0.8055802926421165,
      "num_tokens": 8441165.0,
      "step": 1100
    },
    {
      "entropy": 0.9736937612295151,
      "epoch": 0.16145454545454546,
      "grad_norm": 0.3738587498664856,
      "learning_rate": 3.6966666666666666e-05,
      "loss": 0.9332,
      "mean_token_accuracy": 0.8084321849048137,
      "num_tokens": 8517287.0,
      "step": 1110
    },
    {
      "entropy": 0.9818037033081055,
      "epoch": 0.16290909090909092,
      "grad_norm": 0.39459410309791565,
      "learning_rate": 3.73e-05,
      "loss": 0.9535,
      "mean_token_accuracy": 0.8051852092146874,
      "num_tokens": 8594226.0,
      "step": 1120
    },
    {
      "entropy": 0.9724020875990391,
      "epoch": 0.16436363636363635,
      "grad_norm": 0.3696468770503998,
      "learning_rate": 3.763333333333334e-05,
      "loss": 0.9481,
      "mean_token_accuracy": 0.8077612824738025,
      "num_tokens": 8670364.0,
      "step": 1130
    },
    {
      "entropy": 0.9997089162468911,
      "epoch": 0.1658181818181818,
      "grad_norm": 0.3834758400917053,
      "learning_rate": 3.796666666666667e-05,
      "loss": 0.9674,
      "mean_token_accuracy": 0.8061671704053879,
      "num_tokens": 8746754.0,
      "step": 1140
    },
    {
      "entropy": 0.9531174071133137,
      "epoch": 0.16727272727272727,
      "grad_norm": 0.41454288363456726,
      "learning_rate": 3.83e-05,
      "loss": 0.9294,
      "mean_token_accuracy": 0.8093575298786163,
      "num_tokens": 8822823.0,
      "step": 1150
    },
    {
      "entropy": 0.9577366463840008,
      "epoch": 0.16872727272727273,
      "grad_norm": 0.392678439617157,
      "learning_rate": 3.8633333333333335e-05,
      "loss": 0.9257,
      "mean_token_accuracy": 0.8099003352224827,
      "num_tokens": 8899652.0,
      "step": 1160
    },
    {
      "entropy": 0.9818715266883373,
      "epoch": 0.17018181818181818,
      "grad_norm": 0.42063838243484497,
      "learning_rate": 3.896666666666667e-05,
      "loss": 0.9707,
      "mean_token_accuracy": 0.8053286999464035,
      "num_tokens": 8973918.0,
      "step": 1170
    },
    {
      "entropy": 0.9495494857430458,
      "epoch": 0.17163636363636364,
      "grad_norm": 0.4945516586303711,
      "learning_rate": 3.9300000000000007e-05,
      "loss": 0.9105,
      "mean_token_accuracy": 0.8125501699745655,
      "num_tokens": 9047774.0,
      "step": 1180
    },
    {
      "entropy": 0.9826383985579014,
      "epoch": 0.1730909090909091,
      "grad_norm": 0.4486415386199951,
      "learning_rate": 3.963333333333333e-05,
      "loss": 0.9821,
      "mean_token_accuracy": 0.8069185055792332,
      "num_tokens": 9124562.0,
      "step": 1190
    },
    {
      "entropy": 0.9851367242634297,
      "epoch": 0.17454545454545456,
      "grad_norm": 0.40190815925598145,
      "learning_rate": 3.996666666666667e-05,
      "loss": 0.962,
      "mean_token_accuracy": 0.8029212132096291,
      "num_tokens": 9204050.0,
      "step": 1200
    },
    {
      "entropy": 0.9761835239827633,
      "epoch": 0.176,
      "grad_norm": 0.3759288489818573,
      "learning_rate": 4.0300000000000004e-05,
      "loss": 0.9456,
      "mean_token_accuracy": 0.8066259518265724,
      "num_tokens": 9279622.0,
      "step": 1210
    },
    {
      "entropy": 0.9822940185666085,
      "epoch": 0.17745454545454545,
      "grad_norm": 0.4097426235675812,
      "learning_rate": 4.0633333333333336e-05,
      "loss": 0.9627,
      "mean_token_accuracy": 0.8040494129061699,
      "num_tokens": 9356777.0,
      "step": 1220
    },
    {
      "entropy": 0.9587647147476673,
      "epoch": 0.1789090909090909,
      "grad_norm": 0.4142778813838959,
      "learning_rate": 4.096666666666667e-05,
      "loss": 0.9318,
      "mean_token_accuracy": 0.8084065370261669,
      "num_tokens": 9436024.0,
      "step": 1230
    },
    {
      "entropy": 0.9789021521806717,
      "epoch": 0.18036363636363636,
      "grad_norm": 0.3709389269351959,
      "learning_rate": 4.13e-05,
      "loss": 0.9486,
      "mean_token_accuracy": 0.8077368602156639,
      "num_tokens": 9511885.0,
      "step": 1240
    },
    {
      "entropy": 0.9523784570395947,
      "epoch": 0.18181818181818182,
      "grad_norm": 0.400603324174881,
      "learning_rate": 4.1633333333333333e-05,
      "loss": 0.9218,
      "mean_token_accuracy": 0.8094192072749138,
      "num_tokens": 9587513.0,
      "step": 1250
    },
    {
      "entropy": 0.9448994480073452,
      "epoch": 0.18327272727272728,
      "grad_norm": 0.4005239009857178,
      "learning_rate": 4.196666666666667e-05,
      "loss": 0.9146,
      "mean_token_accuracy": 0.811667163670063,
      "num_tokens": 9661338.0,
      "step": 1260
    },
    {
      "entropy": 0.9539985828101635,
      "epoch": 0.18472727272727274,
      "grad_norm": 0.38217607140541077,
      "learning_rate": 4.23e-05,
      "loss": 0.9269,
      "mean_token_accuracy": 0.811439736187458,
      "num_tokens": 9737921.0,
      "step": 1270
    },
    {
      "entropy": 0.9687586560845375,
      "epoch": 0.18618181818181817,
      "grad_norm": 0.41379278898239136,
      "learning_rate": 4.263333333333334e-05,
      "loss": 0.9489,
      "mean_token_accuracy": 0.8053133375942707,
      "num_tokens": 9815704.0,
      "step": 1280
    },
    {
      "entropy": 0.9803968369960785,
      "epoch": 0.18763636363636363,
      "grad_norm": 0.40967634320259094,
      "learning_rate": 4.296666666666666e-05,
      "loss": 0.9468,
      "mean_token_accuracy": 0.8058573313057422,
      "num_tokens": 9894635.0,
      "step": 1290
    },
    {
      "entropy": 0.9675418928265571,
      "epoch": 0.1890909090909091,
      "grad_norm": 0.398711621761322,
      "learning_rate": 4.33e-05,
      "loss": 0.9453,
      "mean_token_accuracy": 0.8056280307471753,
      "num_tokens": 9972662.0,
      "step": 1300
    },
    {
      "entropy": 0.9511002562940121,
      "epoch": 0.19054545454545455,
      "grad_norm": 0.3831891417503357,
      "learning_rate": 4.3633333333333335e-05,
      "loss": 0.9275,
      "mean_token_accuracy": 0.8103599421679973,
      "num_tokens": 10049277.0,
      "step": 1310
    },
    {
      "entropy": 0.9432598523795604,
      "epoch": 0.192,
      "grad_norm": 0.3834156095981598,
      "learning_rate": 4.396666666666667e-05,
      "loss": 0.9216,
      "mean_token_accuracy": 0.811466021835804,
      "num_tokens": 10125383.0,
      "step": 1320
    },
    {
      "entropy": 0.9695509068667889,
      "epoch": 0.19345454545454546,
      "grad_norm": 0.35917171835899353,
      "learning_rate": 4.43e-05,
      "loss": 0.9382,
      "mean_token_accuracy": 0.8076730266213417,
      "num_tokens": 10201809.0,
      "step": 1330
    },
    {
      "entropy": 0.9703869722783566,
      "epoch": 0.19490909090909092,
      "grad_norm": 0.423606812953949,
      "learning_rate": 4.463333333333334e-05,
      "loss": 0.9389,
      "mean_token_accuracy": 0.8067005030810833,
      "num_tokens": 10278795.0,
      "step": 1340
    },
    {
      "entropy": 0.9651624113321304,
      "epoch": 0.19636363636363635,
      "grad_norm": 0.3861379325389862,
      "learning_rate": 4.496666666666667e-05,
      "loss": 0.947,
      "mean_token_accuracy": 0.806987015902996,
      "num_tokens": 10356816.0,
      "step": 1350
    },
    {
      "entropy": 0.9701198406517506,
      "epoch": 0.1978181818181818,
      "grad_norm": 0.3669825792312622,
      "learning_rate": 4.53e-05,
      "loss": 0.9555,
      "mean_token_accuracy": 0.8069166608154774,
      "num_tokens": 10433476.0,
      "step": 1360
    },
    {
      "entropy": 0.9480470046401024,
      "epoch": 0.19927272727272727,
      "grad_norm": 0.5006280541419983,
      "learning_rate": 4.5633333333333336e-05,
      "loss": 0.9147,
      "mean_token_accuracy": 0.8121759042143821,
      "num_tokens": 10509912.0,
      "step": 1370
    },
    {
      "entropy": 0.9768481604754925,
      "epoch": 0.20072727272727273,
      "grad_norm": 0.35172003507614136,
      "learning_rate": 4.596666666666667e-05,
      "loss": 0.9552,
      "mean_token_accuracy": 0.8053648866713047,
      "num_tokens": 10588749.0,
      "step": 1380
    },
    {
      "entropy": 0.9502091735601426,
      "epoch": 0.20218181818181818,
      "grad_norm": 0.4971792697906494,
      "learning_rate": 4.630000000000001e-05,
      "loss": 0.9283,
      "mean_token_accuracy": 0.8122193835675716,
      "num_tokens": 10664262.0,
      "step": 1390
    },
    {
      "entropy": 0.9593991063535213,
      "epoch": 0.20363636363636364,
      "grad_norm": 0.3790259063243866,
      "learning_rate": 4.663333333333333e-05,
      "loss": 0.9229,
      "mean_token_accuracy": 0.8104354314506054,
      "num_tokens": 10739818.0,
      "step": 1400
    },
    {
      "entropy": 0.9550940081477165,
      "epoch": 0.2050909090909091,
      "grad_norm": 0.4480496644973755,
      "learning_rate": 4.696666666666667e-05,
      "loss": 0.937,
      "mean_token_accuracy": 0.8094347164034843,
      "num_tokens": 10815065.0,
      "step": 1410
    },
    {
      "entropy": 0.952402526885271,
      "epoch": 0.20654545454545453,
      "grad_norm": 0.3347347378730774,
      "learning_rate": 4.73e-05,
      "loss": 0.9277,
      "mean_token_accuracy": 0.8107926152646542,
      "num_tokens": 10892502.0,
      "step": 1420
    },
    {
      "entropy": 0.9966687709093094,
      "epoch": 0.208,
      "grad_norm": 0.39030221104621887,
      "learning_rate": 4.763333333333334e-05,
      "loss": 0.9658,
      "mean_token_accuracy": 0.8048013903200626,
      "num_tokens": 10969083.0,
      "step": 1430
    },
    {
      "entropy": 0.9894838415086269,
      "epoch": 0.20945454545454545,
      "grad_norm": 0.3712361454963684,
      "learning_rate": 4.796666666666667e-05,
      "loss": 0.9615,
      "mean_token_accuracy": 0.8025919504463672,
      "num_tokens": 11044803.0,
      "step": 1440
    },
    {
      "entropy": 0.9651565335690975,
      "epoch": 0.2109090909090909,
      "grad_norm": 0.38604411482810974,
      "learning_rate": 4.83e-05,
      "loss": 0.9367,
      "mean_token_accuracy": 0.8086418233811855,
      "num_tokens": 11118299.0,
      "step": 1450
    },
    {
      "entropy": 0.981209859997034,
      "epoch": 0.21236363636363637,
      "grad_norm": 0.37849947810173035,
      "learning_rate": 4.8633333333333334e-05,
      "loss": 0.9573,
      "mean_token_accuracy": 0.8040041103959084,
      "num_tokens": 11196762.0,
      "step": 1460
    },
    {
      "entropy": 0.9578036166727543,
      "epoch": 0.21381818181818182,
      "grad_norm": 0.4023977518081665,
      "learning_rate": 4.8966666666666667e-05,
      "loss": 0.9327,
      "mean_token_accuracy": 0.809338016808033,
      "num_tokens": 11271543.0,
      "step": 1470
    },
    {
      "entropy": 0.9572117276489734,
      "epoch": 0.21527272727272728,
      "grad_norm": 0.5094729661941528,
      "learning_rate": 4.93e-05,
      "loss": 0.9195,
      "mean_token_accuracy": 0.8115039594471455,
      "num_tokens": 11346972.0,
      "step": 1480
    },
    {
      "entropy": 0.9531163774430752,
      "epoch": 0.21672727272727274,
      "grad_norm": 0.37371397018432617,
      "learning_rate": 4.963333333333334e-05,
      "loss": 0.9373,
      "mean_token_accuracy": 0.8089435353875161,
      "num_tokens": 11424829.0,
      "step": 1490
    },
    {
      "entropy": 0.9597573406994343,
      "epoch": 0.21818181818181817,
      "grad_norm": 0.3477155864238739,
      "learning_rate": 4.996666666666667e-05,
      "loss": 0.9251,
      "mean_token_accuracy": 0.810747179389,
      "num_tokens": 11501054.0,
      "step": 1500
    },
    {
      "entropy": 0.9775046378374099,
      "epoch": 0.21963636363636363,
      "grad_norm": 0.402798056602478,
      "learning_rate": 4.999965480270814e-05,
      "loss": 0.9481,
      "mean_token_accuracy": 0.8051461532711983,
      "num_tokens": 11580117.0,
      "step": 1510
    },
    {
      "entropy": 0.9834331803023815,
      "epoch": 0.2210909090909091,
      "grad_norm": 0.368505597114563,
      "learning_rate": 4.9998461540382306e-05,
      "loss": 0.9489,
      "mean_token_accuracy": 0.8037358179688454,
      "num_tokens": 11657739.0,
      "step": 1520
    },
    {
      "entropy": 0.9874706588685512,
      "epoch": 0.22254545454545455,
      "grad_norm": 0.3849066495895386,
      "learning_rate": 4.999641599208191e-05,
      "loss": 0.9668,
      "mean_token_accuracy": 0.8038628540933133,
      "num_tokens": 11737453.0,
      "step": 1530
    },
    {
      "entropy": 0.9514814570546151,
      "epoch": 0.224,
      "grad_norm": 0.34791556000709534,
      "learning_rate": 4.9993518227686675e-05,
      "loss": 0.9257,
      "mean_token_accuracy": 0.8099611721932888,
      "num_tokens": 11813293.0,
      "step": 1540
    },
    {
      "entropy": 0.9615619413554668,
      "epoch": 0.22545454545454546,
      "grad_norm": 0.387060284614563,
      "learning_rate": 4.998976834618965e-05,
      "loss": 0.9317,
      "mean_token_accuracy": 0.8071263432502747,
      "num_tokens": 11889932.0,
      "step": 1550
    },
    {
      "entropy": 0.9558730646967888,
      "epoch": 0.22690909090909092,
      "grad_norm": 0.47426292300224304,
      "learning_rate": 4.998516647569379e-05,
      "loss": 0.9249,
      "mean_token_accuracy": 0.8095598690211773,
      "num_tokens": 11968281.0,
      "step": 1560
    },
    {
      "entropy": 0.9359629921615124,
      "epoch": 0.22836363636363635,
      "grad_norm": 0.35277876257896423,
      "learning_rate": 4.9979712773407554e-05,
      "loss": 0.9262,
      "mean_token_accuracy": 0.8103468790650368,
      "num_tokens": 12044028.0,
      "step": 1570
    },
    {
      "entropy": 0.9580157190561295,
      "epoch": 0.2298181818181818,
      "grad_norm": 0.37145012617111206,
      "learning_rate": 4.9973407425639596e-05,
      "loss": 0.9232,
      "mean_token_accuracy": 0.8107329845428467,
      "num_tokens": 12116626.0,
      "step": 1580
    },
    {
      "entropy": 0.9677141040563584,
      "epoch": 0.23127272727272727,
      "grad_norm": 0.3671353757381439,
      "learning_rate": 4.996625064779235e-05,
      "loss": 0.946,
      "mean_token_accuracy": 0.8072254911065102,
      "num_tokens": 12193513.0,
      "step": 1590
    },
    {
      "entropy": 0.9686196386814118,
      "epoch": 0.23272727272727273,
      "grad_norm": 0.35971471667289734,
      "learning_rate": 4.99582426843547e-05,
      "loss": 0.9481,
      "mean_token_accuracy": 0.8071218743920326,
      "num_tokens": 12270596.0,
      "step": 1600
    },
    {
      "entropy": 0.9497562602162362,
      "epoch": 0.23418181818181819,
      "grad_norm": 0.3752904534339905,
      "learning_rate": 4.994938380889361e-05,
      "loss": 0.9153,
      "mean_token_accuracy": 0.8122536353766918,
      "num_tokens": 12345573.0,
      "step": 1610
    },
    {
      "entropy": 0.9516720153391361,
      "epoch": 0.23563636363636364,
      "grad_norm": 0.39580070972442627,
      "learning_rate": 4.9939674324044774e-05,
      "loss": 0.9286,
      "mean_token_accuracy": 0.8101835794746876,
      "num_tokens": 12420887.0,
      "step": 1620
    },
    {
      "entropy": 0.9540682218968868,
      "epoch": 0.2370909090909091,
      "grad_norm": 0.38636210560798645,
      "learning_rate": 4.9929114561502316e-05,
      "loss": 0.9416,
      "mean_token_accuracy": 0.8087294958531857,
      "num_tokens": 12497505.0,
      "step": 1630
    },
    {
      "entropy": 0.974821624904871,
      "epoch": 0.23854545454545453,
      "grad_norm": 0.39139118790626526,
      "learning_rate": 4.991770488200743e-05,
      "loss": 0.941,
      "mean_token_accuracy": 0.8072389960289001,
      "num_tokens": 12573715.0,
      "step": 1640
    },
    {
      "entropy": 0.9520889095962047,
      "epoch": 0.24,
      "grad_norm": 0.3420006334781647,
      "learning_rate": 4.990544567533601e-05,
      "loss": 0.9301,
      "mean_token_accuracy": 0.8097614191472531,
      "num_tokens": 12650432.0,
      "step": 1650
    },
    {
      "entropy": 0.9789438024163246,
      "epoch": 0.24145454545454545,
      "grad_norm": 0.3620925545692444,
      "learning_rate": 4.9892337360285455e-05,
      "loss": 0.9684,
      "mean_token_accuracy": 0.8048036694526672,
      "num_tokens": 12729021.0,
      "step": 1660
    },
    {
      "entropy": 0.9667664721608162,
      "epoch": 0.2429090909090909,
      "grad_norm": 0.3239378035068512,
      "learning_rate": 4.987838038466024e-05,
      "loss": 0.9406,
      "mean_token_accuracy": 0.8075184918940067,
      "num_tokens": 12805720.0,
      "step": 1670
    },
    {
      "entropy": 0.9742927253246307,
      "epoch": 0.24436363636363637,
      "grad_norm": 0.3850766718387604,
      "learning_rate": 4.9863575225256676e-05,
      "loss": 0.9454,
      "mean_token_accuracy": 0.8060403868556023,
      "num_tokens": 12883299.0,
      "step": 1680
    },
    {
      "entropy": 0.9463119588792324,
      "epoch": 0.24581818181818182,
      "grad_norm": 0.37080949544906616,
      "learning_rate": 4.984792238784663e-05,
      "loss": 0.9237,
      "mean_token_accuracy": 0.8102991424500943,
      "num_tokens": 12959320.0,
      "step": 1690
    },
    {
      "entropy": 0.9709216229617595,
      "epoch": 0.24727272727272728,
      "grad_norm": 0.3962990343570709,
      "learning_rate": 4.983142240716021e-05,
      "loss": 0.9365,
      "mean_token_accuracy": 0.8095202818512917,
      "num_tokens": 13035585.0,
      "step": 1700
    },
    {
      "entropy": 0.9721660941839219,
      "epoch": 0.2487272727272727,
      "grad_norm": 0.4943894147872925,
      "learning_rate": 4.9814075846867545e-05,
      "loss": 0.9522,
      "mean_token_accuracy": 0.8075993709266186,
      "num_tokens": 13112084.0,
      "step": 1710
    },
    {
      "entropy": 0.9542104244232178,
      "epoch": 0.25018181818181817,
      "grad_norm": 0.35401037335395813,
      "learning_rate": 4.9795883299559465e-05,
      "loss": 0.9274,
      "mean_token_accuracy": 0.8096915259957314,
      "num_tokens": 13188680.0,
      "step": 1720
    },
    {
      "entropy": 0.9506775386631489,
      "epoch": 0.25163636363636366,
      "grad_norm": 0.3483401834964752,
      "learning_rate": 4.977684538672732e-05,
      "loss": 0.9339,
      "mean_token_accuracy": 0.8106581576168537,
      "num_tokens": 13264460.0,
      "step": 1730
    },
    {
      "entropy": 0.9490490764379501,
      "epoch": 0.2530909090909091,
      "grad_norm": 0.3306885063648224,
      "learning_rate": 4.975696275874173e-05,
      "loss": 0.9208,
      "mean_token_accuracy": 0.8103131666779518,
      "num_tokens": 13340847.0,
      "step": 1740
    },
    {
      "entropy": 0.9423711940646171,
      "epoch": 0.2545454545454545,
      "grad_norm": 0.31907522678375244,
      "learning_rate": 4.973623609483034e-05,
      "loss": 0.9281,
      "mean_token_accuracy": 0.8098199367523193,
      "num_tokens": 13416747.0,
      "step": 1750
    },
    {
      "entropy": 0.9571038953959942,
      "epoch": 0.256,
      "grad_norm": 0.3257800042629242,
      "learning_rate": 4.971466610305464e-05,
      "loss": 0.9321,
      "mean_token_accuracy": 0.808480229228735,
      "num_tokens": 13495344.0,
      "step": 1760
    },
    {
      "entropy": 0.9668020732700825,
      "epoch": 0.25745454545454544,
      "grad_norm": 0.38670650124549866,
      "learning_rate": 4.9692253520285776e-05,
      "loss": 0.9388,
      "mean_token_accuracy": 0.8080372959375381,
      "num_tokens": 13572767.0,
      "step": 1770
    },
    {
      "entropy": 0.9665816195309163,
      "epoch": 0.2589090909090909,
      "grad_norm": 0.3654588758945465,
      "learning_rate": 4.966899911217938e-05,
      "loss": 0.9538,
      "mean_token_accuracy": 0.8079737424850464,
      "num_tokens": 13648876.0,
      "step": 1780
    },
    {
      "entropy": 0.9419120058417321,
      "epoch": 0.26036363636363635,
      "grad_norm": 0.3396640419960022,
      "learning_rate": 4.9644903673149414e-05,
      "loss": 0.9216,
      "mean_token_accuracy": 0.809430719166994,
      "num_tokens": 13728286.0,
      "step": 1790
    },
    {
      "entropy": 0.9564610227942467,
      "epoch": 0.26181818181818184,
      "grad_norm": 0.3665502667427063,
      "learning_rate": 4.9619968026340985e-05,
      "loss": 0.9273,
      "mean_token_accuracy": 0.8096649095416069,
      "num_tokens": 13804065.0,
      "step": 1800
    },
    {
      "entropy": 0.9598980963230133,
      "epoch": 0.26327272727272727,
      "grad_norm": 0.3303588330745697,
      "learning_rate": 4.959419302360231e-05,
      "loss": 0.9411,
      "mean_token_accuracy": 0.8076228819787502,
      "num_tokens": 13880540.0,
      "step": 1810
    },
    {
      "entropy": 1.006081961095333,
      "epoch": 0.2647272727272727,
      "grad_norm": 0.3581242859363556,
      "learning_rate": 4.9567579545455546e-05,
      "loss": 0.9989,
      "mean_token_accuracy": 0.8009657979011535,
      "num_tokens": 13960083.0,
      "step": 1820
    },
    {
      "entropy": 0.9628503814339637,
      "epoch": 0.2661818181818182,
      "grad_norm": 0.32779112458229065,
      "learning_rate": 4.9540128501066745e-05,
      "loss": 0.9338,
      "mean_token_accuracy": 0.8094890132546425,
      "num_tokens": 14037453.0,
      "step": 1830
    },
    {
      "entropy": 0.9735031522810459,
      "epoch": 0.2676363636363636,
      "grad_norm": 0.37286221981048584,
      "learning_rate": 4.951184082821478e-05,
      "loss": 0.9717,
      "mean_token_accuracy": 0.80675984993577,
      "num_tokens": 14114568.0,
      "step": 1840
    },
    {
      "entropy": 0.9624778412282466,
      "epoch": 0.2690909090909091,
      "grad_norm": 0.33689644932746887,
      "learning_rate": 4.948271749325931e-05,
      "loss": 0.9329,
      "mean_token_accuracy": 0.8100909695029259,
      "num_tokens": 14191206.0,
      "step": 1850
    },
    {
      "entropy": 0.9402816690504551,
      "epoch": 0.27054545454545453,
      "grad_norm": 0.43606722354888916,
      "learning_rate": 4.945275949110776e-05,
      "loss": 0.9176,
      "mean_token_accuracy": 0.8115701250731945,
      "num_tokens": 14267843.0,
      "step": 1860
    },
    {
      "entropy": 0.9536763995885849,
      "epoch": 0.272,
      "grad_norm": 0.36859753727912903,
      "learning_rate": 4.942196784518137e-05,
      "loss": 0.9325,
      "mean_token_accuracy": 0.8102965012192727,
      "num_tokens": 14343772.0,
      "step": 1870
    },
    {
      "entropy": 0.9411172643303871,
      "epoch": 0.27345454545454545,
      "grad_norm": 0.31749650835990906,
      "learning_rate": 4.939034360738019e-05,
      "loss": 0.9409,
      "mean_token_accuracy": 0.8114774063229561,
      "num_tokens": 14420089.0,
      "step": 1880
    },
    {
      "entropy": 0.9489653743803501,
      "epoch": 0.27490909090909094,
      "grad_norm": 0.32358062267303467,
      "learning_rate": 4.935788785804715e-05,
      "loss": 0.9194,
      "mean_token_accuracy": 0.8125798895955085,
      "num_tokens": 14495945.0,
      "step": 1890
    },
    {
      "entropy": 0.9492652073502541,
      "epoch": 0.27636363636363637,
      "grad_norm": 0.34433141350746155,
      "learning_rate": 4.932460170593119e-05,
      "loss": 0.9282,
      "mean_token_accuracy": 0.8116157680749894,
      "num_tokens": 14570699.0,
      "step": 1900
    },
    {
      "entropy": 0.9683300837874412,
      "epoch": 0.2778181818181818,
      "grad_norm": 0.3886536657810211,
      "learning_rate": 4.929048628814933e-05,
      "loss": 0.9391,
      "mean_token_accuracy": 0.8086844593286514,
      "num_tokens": 14646642.0,
      "step": 1910
    },
    {
      "entropy": 0.9709325604140758,
      "epoch": 0.2792727272727273,
      "grad_norm": 0.39230605959892273,
      "learning_rate": 4.925554277014788e-05,
      "loss": 0.9506,
      "mean_token_accuracy": 0.8078415438532829,
      "num_tokens": 14723663.0,
      "step": 1920
    },
    {
      "entropy": 0.9794384993612766,
      "epoch": 0.2807272727272727,
      "grad_norm": 0.3203677833080292,
      "learning_rate": 4.9219772345662575e-05,
      "loss": 0.9504,
      "mean_token_accuracy": 0.8059118509292602,
      "num_tokens": 14801605.0,
      "step": 1930
    },
    {
      "entropy": 0.9421239554882049,
      "epoch": 0.2821818181818182,
      "grad_norm": 0.3160009980201721,
      "learning_rate": 4.918317623667783e-05,
      "loss": 0.9169,
      "mean_token_accuracy": 0.8127878844738007,
      "num_tokens": 14877660.0,
      "step": 1940
    },
    {
      "entropy": 0.9400995798408985,
      "epoch": 0.28363636363636363,
      "grad_norm": 0.3499312400817871,
      "learning_rate": 4.9145755693384975e-05,
      "loss": 0.9175,
      "mean_token_accuracy": 0.8117764830589295,
      "num_tokens": 14955540.0,
      "step": 1950
    },
    {
      "entropy": 0.9481630109250545,
      "epoch": 0.2850909090909091,
      "grad_norm": 0.36080819368362427,
      "learning_rate": 4.910751199413956e-05,
      "loss": 0.9239,
      "mean_token_accuracy": 0.8113817647099495,
      "num_tokens": 15032507.0,
      "step": 1960
    },
    {
      "entropy": 0.9711731873452664,
      "epoch": 0.28654545454545455,
      "grad_norm": 0.34929078817367554,
      "learning_rate": 4.9068446445417674e-05,
      "loss": 0.9517,
      "mean_token_accuracy": 0.8062169618904591,
      "num_tokens": 15111260.0,
      "step": 1970
    },
    {
      "entropy": 0.9310999117791653,
      "epoch": 0.288,
      "grad_norm": 0.3529198169708252,
      "learning_rate": 4.902856038177132e-05,
      "loss": 0.8995,
      "mean_token_accuracy": 0.8154916487634182,
      "num_tokens": 15187148.0,
      "step": 1980
    },
    {
      "entropy": 0.9400709681212902,
      "epoch": 0.28945454545454546,
      "grad_norm": 0.38182997703552246,
      "learning_rate": 4.898785516578282e-05,
      "loss": 0.9329,
      "mean_token_accuracy": 0.8103081896901131,
      "num_tokens": 15262960.0,
      "step": 1990
    },
    {
      "entropy": 0.9691479153931141,
      "epoch": 0.2909090909090909,
      "grad_norm": 0.4837636947631836,
      "learning_rate": 4.8946332188018265e-05,
      "loss": 0.94,
      "mean_token_accuracy": 0.8090752199292183,
      "num_tokens": 15340217.0,
      "step": 2000
    },
    {
      "epoch": 0.2909090909090909,
      "eval_entropy": 0.9552994007110596,
      "eval_loss": 0.9162204265594482,
      "eval_mean_token_accuracy": 0.8091625803947449,
      "eval_num_tokens": 15340217.0,
      "eval_runtime": 77.5949,
      "eval_samples_per_second": 64.437,
      "eval_steps_per_second": 8.055,
      "step": 2000
    },
    {
      "entropy": 0.9568299718201161,
      "epoch": 0.2923636363636364,
      "grad_norm": 0.45410582423210144,
      "learning_rate": 4.890399286698002e-05,
      "loss": 0.958,
      "mean_token_accuracy": 0.8088321089744568,
      "num_tokens": 15417075.0,
      "step": 2010
    },
    {
      "entropy": 0.9500932864844799,
      "epoch": 0.2938181818181818,
      "grad_norm": 0.3521057665348053,
      "learning_rate": 4.8860838649058236e-05,
      "loss": 0.9156,
      "mean_token_accuracy": 0.8119934909045696,
      "num_tokens": 15492696.0,
      "step": 2020
    },
    {
      "entropy": 0.9271357983350754,
      "epoch": 0.2952727272727273,
      "grad_norm": 0.34188908338546753,
      "learning_rate": 4.8816871008481494e-05,
      "loss": 0.9133,
      "mean_token_accuracy": 0.8133960500359535,
      "num_tokens": 15569060.0,
      "step": 2030
    },
    {
      "entropy": 0.9351197473704815,
      "epoch": 0.29672727272727273,
      "grad_norm": 0.3230991065502167,
      "learning_rate": 4.87720914472664e-05,
      "loss": 0.9083,
      "mean_token_accuracy": 0.8133096255362033,
      "num_tokens": 15645189.0,
      "step": 2040
    },
    {
      "entropy": 0.9560852214694023,
      "epoch": 0.29818181818181816,
      "grad_norm": 0.42373934388160706,
      "learning_rate": 4.872650149516626e-05,
      "loss": 0.9272,
      "mean_token_accuracy": 0.8098875015974045,
      "num_tokens": 15722425.0,
      "step": 2050
    },
    {
      "entropy": 0.9704909332096576,
      "epoch": 0.29963636363636365,
      "grad_norm": 0.2958741784095764,
      "learning_rate": 4.8680102709618874e-05,
      "loss": 0.974,
      "mean_token_accuracy": 0.8073359347879887,
      "num_tokens": 15800614.0,
      "step": 2060
    },
    {
      "entropy": 0.9467493839561939,
      "epoch": 0.3010909090909091,
      "grad_norm": 0.3630009591579437,
      "learning_rate": 4.863289667569328e-05,
      "loss": 0.9285,
      "mean_token_accuracy": 0.8092270664870739,
      "num_tokens": 15877885.0,
      "step": 2070
    },
    {
      "entropy": 0.9440895996987819,
      "epoch": 0.30254545454545456,
      "grad_norm": 0.3074013590812683,
      "learning_rate": 4.8584885006035654e-05,
      "loss": 0.9179,
      "mean_token_accuracy": 0.8130018509924412,
      "num_tokens": 15954449.0,
      "step": 2080
    },
    {
      "entropy": 0.9428293339908123,
      "epoch": 0.304,
      "grad_norm": 0.32661521434783936,
      "learning_rate": 4.853606934081415e-05,
      "loss": 0.9303,
      "mean_token_accuracy": 0.8101019099354744,
      "num_tokens": 16032223.0,
      "step": 2090
    },
    {
      "entropy": 0.9483884163200855,
      "epoch": 0.3054545454545455,
      "grad_norm": 0.3456633388996124,
      "learning_rate": 4.8486451347662955e-05,
      "loss": 0.9278,
      "mean_token_accuracy": 0.8105225287377834,
      "num_tokens": 16109007.0,
      "step": 2100
    },
    {
      "entropy": 0.9410559803247451,
      "epoch": 0.3069090909090909,
      "grad_norm": 0.3802877366542816,
      "learning_rate": 4.843603272162522e-05,
      "loss": 0.9153,
      "mean_token_accuracy": 0.8141518786549569,
      "num_tokens": 16186233.0,
      "step": 2110
    },
    {
      "entropy": 0.9398015670478344,
      "epoch": 0.30836363636363634,
      "grad_norm": 0.370561808347702,
      "learning_rate": 4.838481518509527e-05,
      "loss": 0.9145,
      "mean_token_accuracy": 0.8143410183489322,
      "num_tokens": 16262660.0,
      "step": 2120
    },
    {
      "entropy": 0.9372091844677926,
      "epoch": 0.3098181818181818,
      "grad_norm": 0.35077038407325745,
      "learning_rate": 4.8332800487759666e-05,
      "loss": 0.9266,
      "mean_token_accuracy": 0.8107892982661724,
      "num_tokens": 16338493.0,
      "step": 2130
    },
    {
      "entropy": 0.923904549330473,
      "epoch": 0.31127272727272726,
      "grad_norm": 0.3715773820877075,
      "learning_rate": 4.8279990406537466e-05,
      "loss": 0.9024,
      "mean_token_accuracy": 0.8144333988428116,
      "num_tokens": 16413438.0,
      "step": 2140
    },
    {
      "entropy": 0.9547001197934151,
      "epoch": 0.31272727272727274,
      "grad_norm": 0.33229219913482666,
      "learning_rate": 4.8226386745519536e-05,
      "loss": 0.9287,
      "mean_token_accuracy": 0.8108532302081585,
      "num_tokens": 16489851.0,
      "step": 2150
    },
    {
      "entropy": 0.9539036385715007,
      "epoch": 0.3141818181818182,
      "grad_norm": 0.4329231083393097,
      "learning_rate": 4.8171991335906915e-05,
      "loss": 0.9358,
      "mean_token_accuracy": 0.8108682915568352,
      "num_tokens": 16566105.0,
      "step": 2160
    },
    {
      "entropy": 0.9734240360558033,
      "epoch": 0.31563636363636366,
      "grad_norm": 0.38508152961730957,
      "learning_rate": 4.8116806035948224e-05,
      "loss": 0.9546,
      "mean_token_accuracy": 0.8067123219370842,
      "num_tokens": 16644769.0,
      "step": 2170
    },
    {
      "entropy": 0.949046066403389,
      "epoch": 0.3170909090909091,
      "grad_norm": 0.35903850197792053,
      "learning_rate": 4.8060832730876256e-05,
      "loss": 0.9279,
      "mean_token_accuracy": 0.810236643254757,
      "num_tokens": 16720786.0,
      "step": 2180
    },
    {
      "entropy": 0.9580598846077919,
      "epoch": 0.3185454545454545,
      "grad_norm": 0.3273390829563141,
      "learning_rate": 4.800407333284348e-05,
      "loss": 0.9408,
      "mean_token_accuracy": 0.8102932535111904,
      "num_tokens": 16796483.0,
      "step": 2190
    },
    {
      "entropy": 0.9703283801674842,
      "epoch": 0.32,
      "grad_norm": 0.3284286856651306,
      "learning_rate": 4.794652978085679e-05,
      "loss": 0.9662,
      "mean_token_accuracy": 0.8069152064621449,
      "num_tokens": 16875079.0,
      "step": 2200
    },
    {
      "entropy": 0.9524148769676686,
      "epoch": 0.32145454545454544,
      "grad_norm": 0.31296366453170776,
      "learning_rate": 4.788820404071125e-05,
      "loss": 0.9355,
      "mean_token_accuracy": 0.8084697395563125,
      "num_tokens": 16953792.0,
      "step": 2210
    },
    {
      "entropy": 0.9396099261939526,
      "epoch": 0.3229090909090909,
      "grad_norm": 0.34159210324287415,
      "learning_rate": 4.782909810492291e-05,
      "loss": 0.9103,
      "mean_token_accuracy": 0.8125396214425564,
      "num_tokens": 17029630.0,
      "step": 2220
    },
    {
      "entropy": 0.9276422686874867,
      "epoch": 0.32436363636363635,
      "grad_norm": 0.3081703186035156,
      "learning_rate": 4.776921399266076e-05,
      "loss": 0.9259,
      "mean_token_accuracy": 0.8100886262953282,
      "num_tokens": 17105834.0,
      "step": 2230
    },
    {
      "entropy": 0.965707752853632,
      "epoch": 0.32581818181818184,
      "grad_norm": 0.5014197826385498,
      "learning_rate": 4.770855374967778e-05,
      "loss": 0.9398,
      "mean_token_accuracy": 0.8066871888935566,
      "num_tokens": 17185155.0,
      "step": 2240
    },
    {
      "entropy": 0.9296950966119766,
      "epoch": 0.32727272727272727,
      "grad_norm": 0.3334687352180481,
      "learning_rate": 4.7647119448240984e-05,
      "loss": 0.9069,
      "mean_token_accuracy": 0.8139915406703949,
      "num_tokens": 17261055.0,
      "step": 2250
    },
    {
      "entropy": 0.9424406073987484,
      "epoch": 0.3287272727272727,
      "grad_norm": 0.30235975980758667,
      "learning_rate": 4.7584913187060695e-05,
      "loss": 0.9335,
      "mean_token_accuracy": 0.8098909392952919,
      "num_tokens": 17338769.0,
      "step": 2260
    },
    {
      "entropy": 0.958009808510542,
      "epoch": 0.3301818181818182,
      "grad_norm": 0.41258761286735535,
      "learning_rate": 4.752193709121879e-05,
      "loss": 0.9745,
      "mean_token_accuracy": 0.8083416305482387,
      "num_tokens": 17414642.0,
      "step": 2270
    },
    {
      "entropy": 0.9475978955626487,
      "epoch": 0.3316363636363636,
      "grad_norm": 0.39095503091812134,
      "learning_rate": 4.745819331209617e-05,
      "loss": 0.9313,
      "mean_token_accuracy": 0.8099818773567676,
      "num_tokens": 17493365.0,
      "step": 2280
    },
    {
      "entropy": 0.9432621687650681,
      "epoch": 0.3330909090909091,
      "grad_norm": 0.3390606939792633,
      "learning_rate": 4.73936840272992e-05,
      "loss": 0.9171,
      "mean_token_accuracy": 0.8123437277972698,
      "num_tokens": 17569404.0,
      "step": 2290
    },
    {
      "entropy": 0.958767655491829,
      "epoch": 0.33454545454545453,
      "grad_norm": 0.41404950618743896,
      "learning_rate": 4.7328411440585356e-05,
      "loss": 0.938,
      "mean_token_accuracy": 0.8085070058703423,
      "num_tokens": 17648132.0,
      "step": 2300
    },
    {
      "entropy": 0.9349784903228283,
      "epoch": 0.336,
      "grad_norm": 0.43275147676467896,
      "learning_rate": 4.726237778178794e-05,
      "loss": 0.9089,
      "mean_token_accuracy": 0.8132195100188255,
      "num_tokens": 17723536.0,
      "step": 2310
    },
    {
      "entropy": 0.9718450345098972,
      "epoch": 0.33745454545454545,
      "grad_norm": 0.31931039690971375,
      "learning_rate": 4.7195585306739875e-05,
      "loss": 0.9637,
      "mean_token_accuracy": 0.8050491996109486,
      "num_tokens": 17801793.0,
      "step": 2320
    },
    {
      "entropy": 0.9561341628432274,
      "epoch": 0.3389090909090909,
      "grad_norm": 0.3197164237499237,
      "learning_rate": 4.71280362971967e-05,
      "loss": 0.9233,
      "mean_token_accuracy": 0.8104301512241363,
      "num_tokens": 17880055.0,
      "step": 2330
    },
    {
      "entropy": 0.9318038932979107,
      "epoch": 0.34036363636363637,
      "grad_norm": 0.3325251042842865,
      "learning_rate": 4.7059733060758544e-05,
      "loss": 0.9091,
      "mean_token_accuracy": 0.811470115184784,
      "num_tokens": 17956799.0,
      "step": 2340
    },
    {
      "entropy": 0.978094021230936,
      "epoch": 0.3418181818181818,
      "grad_norm": 0.30171552300453186,
      "learning_rate": 4.6990677930791334e-05,
      "loss": 0.97,
      "mean_token_accuracy": 0.8068195424973965,
      "num_tokens": 18034021.0,
      "step": 2350
    },
    {
      "entropy": 0.9303769364953041,
      "epoch": 0.3432727272727273,
      "grad_norm": 0.359419584274292,
      "learning_rate": 4.692087326634711e-05,
      "loss": 0.9088,
      "mean_token_accuracy": 0.814723651856184,
      "num_tokens": 18109871.0,
      "step": 2360
    },
    {
      "entropy": 0.9502622775733471,
      "epoch": 0.3447272727272727,
      "grad_norm": 0.580029308795929,
      "learning_rate": 4.6850321452083404e-05,
      "loss": 0.9298,
      "mean_token_accuracy": 0.8113038286566734,
      "num_tokens": 18184703.0,
      "step": 2370
    },
    {
      "entropy": 0.9524484679102898,
      "epoch": 0.3461818181818182,
      "grad_norm": 0.37079447507858276,
      "learning_rate": 4.677902489818175e-05,
      "loss": 0.938,
      "mean_token_accuracy": 0.8087799564003945,
      "num_tokens": 18261506.0,
      "step": 2380
    },
    {
      "entropy": 0.9275636874139309,
      "epoch": 0.34763636363636363,
      "grad_norm": 0.38261520862579346,
      "learning_rate": 4.67069860402654e-05,
      "loss": 0.9015,
      "mean_token_accuracy": 0.8157583817839622,
      "num_tokens": 18335664.0,
      "step": 2390
    },
    {
      "entropy": 0.9545292399823666,
      "epoch": 0.3490909090909091,
      "grad_norm": 0.32832345366477966,
      "learning_rate": 4.66342073393161e-05,
      "loss": 0.9329,
      "mean_token_accuracy": 0.8100977629423142,
      "num_tokens": 18412936.0,
      "step": 2400
    },
    {
      "entropy": 0.9629996709525586,
      "epoch": 0.35054545454545455,
      "grad_norm": 0.3800794780254364,
      "learning_rate": 4.6560691281589996e-05,
      "loss": 0.9491,
      "mean_token_accuracy": 0.8075961828231811,
      "num_tokens": 18492236.0,
      "step": 2410
    },
    {
      "entropy": 0.9659391663968563,
      "epoch": 0.352,
      "grad_norm": 0.5112242698669434,
      "learning_rate": 4.6486440378532734e-05,
      "loss": 0.9387,
      "mean_token_accuracy": 0.8073771148920059,
      "num_tokens": 18569666.0,
      "step": 2420
    },
    {
      "entropy": 0.9221311785280705,
      "epoch": 0.35345454545454547,
      "grad_norm": 0.40865597128868103,
      "learning_rate": 4.641145716669363e-05,
      "loss": 0.8972,
      "mean_token_accuracy": 0.8154647730290889,
      "num_tokens": 18644378.0,
      "step": 2430
    },
    {
      "entropy": 0.9529199428856373,
      "epoch": 0.3549090909090909,
      "grad_norm": 0.4366351068019867,
      "learning_rate": 4.6335744207639046e-05,
      "loss": 0.9296,
      "mean_token_accuracy": 0.8110525794327259,
      "num_tokens": 18720972.0,
      "step": 2440
    },
    {
      "entropy": 0.938295865058899,
      "epoch": 0.3563636363636364,
      "grad_norm": 0.3744116425514221,
      "learning_rate": 4.6259304087864866e-05,
      "loss": 0.9209,
      "mean_token_accuracy": 0.8117184430360794,
      "num_tokens": 18798212.0,
      "step": 2450
    },
    {
      "entropy": 0.9459369130432606,
      "epoch": 0.3578181818181818,
      "grad_norm": 0.3553162217140198,
      "learning_rate": 4.618213941870815e-05,
      "loss": 0.9292,
      "mean_token_accuracy": 0.8086876280605793,
      "num_tokens": 18877752.0,
      "step": 2460
    },
    {
      "entropy": 0.9362581737339497,
      "epoch": 0.3592727272727273,
      "grad_norm": 0.349174827337265,
      "learning_rate": 4.6104252836257896e-05,
      "loss": 0.9158,
      "mean_token_accuracy": 0.8127870723605156,
      "num_tokens": 18953448.0,
      "step": 2470
    },
    {
      "entropy": 0.9538705930113792,
      "epoch": 0.36072727272727273,
      "grad_norm": 0.32221290469169617,
      "learning_rate": 4.602564700126502e-05,
      "loss": 0.9232,
      "mean_token_accuracy": 0.8129294320940972,
      "num_tokens": 19030937.0,
      "step": 2480
    },
    {
      "entropy": 0.9399419300258159,
      "epoch": 0.36218181818181816,
      "grad_norm": 0.3328811824321747,
      "learning_rate": 4.5946324599051445e-05,
      "loss": 0.9293,
      "mean_token_accuracy": 0.8096531957387925,
      "num_tokens": 19109004.0,
      "step": 2490
    },
    {
      "entropy": 0.9541197009384632,
      "epoch": 0.36363636363636365,
      "grad_norm": 0.2997928261756897,
      "learning_rate": 4.586628833941835e-05,
      "loss": 0.9206,
      "mean_token_accuracy": 0.8126299031078815,
      "num_tokens": 19184351.0,
      "step": 2500
    },
    {
      "entropy": 0.944844014197588,
      "epoch": 0.3650909090909091,
      "grad_norm": 0.38323697447776794,
      "learning_rate": 4.578554095655366e-05,
      "loss": 0.9363,
      "mean_token_accuracy": 0.809072308242321,
      "num_tokens": 19261014.0,
      "step": 2510
    },
    {
      "entropy": 0.9703416049480438,
      "epoch": 0.36654545454545456,
      "grad_norm": 0.31747499108314514,
      "learning_rate": 4.570408520893854e-05,
      "loss": 0.9407,
      "mean_token_accuracy": 0.8093077220022679,
      "num_tokens": 19337470.0,
      "step": 2520
    },
    {
      "entropy": 0.9435131579637528,
      "epoch": 0.368,
      "grad_norm": 0.3520627021789551,
      "learning_rate": 4.562192387925326e-05,
      "loss": 0.9226,
      "mean_token_accuracy": 0.8121565252542495,
      "num_tokens": 19413364.0,
      "step": 2530
    },
    {
      "entropy": 0.9352571688592434,
      "epoch": 0.3694545454545455,
      "grad_norm": 0.34164953231811523,
      "learning_rate": 4.5539059774282053e-05,
      "loss": 0.9121,
      "mean_token_accuracy": 0.8150348752737046,
      "num_tokens": 19488149.0,
      "step": 2540
    },
    {
      "entropy": 0.9180795967578887,
      "epoch": 0.3709090909090909,
      "grad_norm": 0.30217695236206055,
      "learning_rate": 4.5455495724817306e-05,
      "loss": 0.8977,
      "mean_token_accuracy": 0.8154868699610234,
      "num_tokens": 19562468.0,
      "step": 2550
    },
    {
      "entropy": 0.9273349486291409,
      "epoch": 0.37236363636363634,
      "grad_norm": 0.3009410500526428,
      "learning_rate": 4.537123458556277e-05,
      "loss": 0.9118,
      "mean_token_accuracy": 0.8138667806982994,
      "num_tokens": 19639234.0,
      "step": 2560
    },
    {
      "entropy": 0.9390451267361641,
      "epoch": 0.3738181818181818,
      "grad_norm": 0.34551864862442017,
      "learning_rate": 4.5286279235036135e-05,
      "loss": 0.9104,
      "mean_token_accuracy": 0.8139432102441788,
      "num_tokens": 19715023.0,
      "step": 2570
    },
    {
      "entropy": 0.9297771051526069,
      "epoch": 0.37527272727272726,
      "grad_norm": 0.3966526985168457,
      "learning_rate": 4.520063257547058e-05,
      "loss": 0.9146,
      "mean_token_accuracy": 0.8128842279314995,
      "num_tokens": 19789823.0,
      "step": 2580
    },
    {
      "entropy": 0.9290238834917546,
      "epoch": 0.37672727272727274,
      "grad_norm": 0.5094491839408875,
      "learning_rate": 4.511429753271575e-05,
      "loss": 0.9127,
      "mean_token_accuracy": 0.8136251755058765,
      "num_tokens": 19866350.0,
      "step": 2590
    },
    {
      "entropy": 0.9276151649653912,
      "epoch": 0.3781818181818182,
      "grad_norm": 0.3241227865219116,
      "learning_rate": 4.502727705613773e-05,
      "loss": 0.9165,
      "mean_token_accuracy": 0.8133880697190762,
      "num_tokens": 19940099.0,
      "step": 2600
    },
    {
      "entropy": 0.9553582802414894,
      "epoch": 0.37963636363636366,
      "grad_norm": 0.4422711133956909,
      "learning_rate": 4.4939574118518254e-05,
      "loss": 0.9313,
      "mean_token_accuracy": 0.8098856434226036,
      "num_tokens": 20017376.0,
      "step": 2610
    },
    {
      "entropy": 0.966649767756462,
      "epoch": 0.3810909090909091,
      "grad_norm": 0.35644596815109253,
      "learning_rate": 4.485119171595326e-05,
      "loss": 0.9559,
      "mean_token_accuracy": 0.807817342877388,
      "num_tokens": 20096134.0,
      "step": 2620
    },
    {
      "entropy": 0.9594257563352585,
      "epoch": 0.3825454545454545,
      "grad_norm": 0.3995525538921356,
      "learning_rate": 4.476213286775044e-05,
      "loss": 0.9286,
      "mean_token_accuracy": 0.8108187094330788,
      "num_tokens": 20175183.0,
      "step": 2630
    },
    {
      "entropy": 0.9350139319896698,
      "epoch": 0.384,
      "grad_norm": 0.4301507771015167,
      "learning_rate": 4.4672400616326166e-05,
      "loss": 0.9323,
      "mean_token_accuracy": 0.8119172587990761,
      "num_tokens": 20252274.0,
      "step": 2640
    },
    {
      "entropy": 0.9751557812094689,
      "epoch": 0.38545454545454544,
      "grad_norm": 0.3629772961139679,
      "learning_rate": 4.4581998027101465e-05,
      "loss": 0.9579,
      "mean_token_accuracy": 0.8065298214554787,
      "num_tokens": 20331794.0,
      "step": 2650
    },
    {
      "entropy": 0.9497589774429798,
      "epoch": 0.3869090909090909,
      "grad_norm": 0.5881274938583374,
      "learning_rate": 4.4490928188397395e-05,
      "loss": 0.9266,
      "mean_token_accuracy": 0.811337610334158,
      "num_tokens": 20409223.0,
      "step": 2660
    },
    {
      "entropy": 0.9513171821832657,
      "epoch": 0.38836363636363636,
      "grad_norm": 0.40529772639274597,
      "learning_rate": 4.43991942113295e-05,
      "loss": 0.9326,
      "mean_token_accuracy": 0.8086440213024616,
      "num_tokens": 20488388.0,
      "step": 2670
    },
    {
      "entropy": 0.931061378121376,
      "epoch": 0.38981818181818184,
      "grad_norm": 0.3018703758716583,
      "learning_rate": 4.4306799229701496e-05,
      "loss": 0.9021,
      "mean_token_accuracy": 0.8149745658040046,
      "num_tokens": 20563678.0,
      "step": 2680
    },
    {
      "entropy": 0.9159508943557739,
      "epoch": 0.39127272727272727,
      "grad_norm": 0.31962108612060547,
      "learning_rate": 4.421374639989829e-05,
      "loss": 0.8959,
      "mean_token_accuracy": 0.8160949237644672,
      "num_tokens": 20638007.0,
      "step": 2690
    },
    {
      "entropy": 0.951894699037075,
      "epoch": 0.3927272727272727,
      "grad_norm": 0.6005260944366455,
      "learning_rate": 4.412003890077808e-05,
      "loss": 0.9379,
      "mean_token_accuracy": 0.8094079531729221,
      "num_tokens": 20714574.0,
      "step": 2700
    },
    {
      "entropy": 0.9519496828317642,
      "epoch": 0.3941818181818182,
      "grad_norm": 0.29477086663246155,
      "learning_rate": 4.40256799335638e-05,
      "loss": 0.9234,
      "mean_token_accuracy": 0.8126239985227585,
      "num_tokens": 20790166.0,
      "step": 2710
    },
    {
      "entropy": 0.9489260002970695,
      "epoch": 0.3956363636363636,
      "grad_norm": 0.3043905794620514,
      "learning_rate": 4.393067272173375e-05,
      "loss": 0.9302,
      "mean_token_accuracy": 0.8110754653811455,
      "num_tokens": 20867119.0,
      "step": 2720
    },
    {
      "entropy": 0.9309591136872768,
      "epoch": 0.3970909090909091,
      "grad_norm": 0.37266799807548523,
      "learning_rate": 4.3835020510911474e-05,
      "loss": 0.8999,
      "mean_token_accuracy": 0.8145073987543583,
      "num_tokens": 20941673.0,
      "step": 2730
    },
    {
      "entropy": 0.9514899291098118,
      "epoch": 0.39854545454545454,
      "grad_norm": 0.33977237343788147,
      "learning_rate": 4.3738726568754875e-05,
      "loss": 0.9485,
      "mean_token_accuracy": 0.8062153734266758,
      "num_tokens": 21021527.0,
      "step": 2740
    },
    {
      "entropy": 0.9509587049484253,
      "epoch": 0.4,
      "grad_norm": 0.34360271692276,
      "learning_rate": 4.364179418484461e-05,
      "loss": 0.9284,
      "mean_token_accuracy": 0.8108667440712451,
      "num_tokens": 21100206.0,
      "step": 2750
    },
    {
      "entropy": 0.9193227373063564,
      "epoch": 0.40145454545454545,
      "grad_norm": 0.3559633195400238,
      "learning_rate": 4.354422667057169e-05,
      "loss": 0.8994,
      "mean_token_accuracy": 0.8140523225069046,
      "num_tokens": 21175327.0,
      "step": 2760
    },
    {
      "entropy": 0.9564199469983578,
      "epoch": 0.4029090909090909,
      "grad_norm": 0.4009409546852112,
      "learning_rate": 4.344602735902436e-05,
      "loss": 0.9339,
      "mean_token_accuracy": 0.8101021826267243,
      "num_tokens": 21252732.0,
      "step": 2770
    },
    {
      "entropy": 0.9387026049196721,
      "epoch": 0.40436363636363637,
      "grad_norm": 0.32919296622276306,
      "learning_rate": 4.334719960487425e-05,
      "loss": 0.9224,
      "mean_token_accuracy": 0.8106619887053966,
      "num_tokens": 21329936.0,
      "step": 2780
    },
    {
      "entropy": 0.9394312173128128,
      "epoch": 0.4058181818181818,
      "grad_norm": 0.3113020360469818,
      "learning_rate": 4.3247746784261736e-05,
      "loss": 0.916,
      "mean_token_accuracy": 0.8107848934829235,
      "num_tokens": 21406701.0,
      "step": 2790
    },
    {
      "entropy": 0.9433960065245628,
      "epoch": 0.4072727272727273,
      "grad_norm": 0.3280106484889984,
      "learning_rate": 4.314767229468067e-05,
      "loss": 0.9233,
      "mean_token_accuracy": 0.8092322021722793,
      "num_tokens": 21483787.0,
      "step": 2800
    },
    {
      "entropy": 0.9473354913294315,
      "epoch": 0.4087272727272727,
      "grad_norm": 0.3314054012298584,
      "learning_rate": 4.3046979554862246e-05,
      "loss": 0.9265,
      "mean_token_accuracy": 0.813962434977293,
      "num_tokens": 21561130.0,
      "step": 2810
    },
    {
      "entropy": 0.9354553163051605,
      "epoch": 0.4101818181818182,
      "grad_norm": 0.3385145664215088,
      "learning_rate": 4.2945672004658235e-05,
      "loss": 0.9178,
      "mean_token_accuracy": 0.8126618385314941,
      "num_tokens": 21637505.0,
      "step": 2820
    },
    {
      "entropy": 0.9360568217933178,
      "epoch": 0.41163636363636363,
      "grad_norm": 0.3262713551521301,
      "learning_rate": 4.28437531049235e-05,
      "loss": 0.9107,
      "mean_token_accuracy": 0.8125479564070701,
      "num_tokens": 21715070.0,
      "step": 2830
    },
    {
      "entropy": 0.9339744850993157,
      "epoch": 0.41309090909090906,
      "grad_norm": 0.37787461280822754,
      "learning_rate": 4.274122633739772e-05,
      "loss": 0.9051,
      "mean_token_accuracy": 0.8136063523590564,
      "num_tokens": 21790607.0,
      "step": 2840
    },
    {
      "entropy": 0.9209459543228149,
      "epoch": 0.41454545454545455,
      "grad_norm": 0.3270346522331238,
      "learning_rate": 4.263809520458651e-05,
      "loss": 0.9028,
      "mean_token_accuracy": 0.8154242433607578,
      "num_tokens": 21865036.0,
      "step": 2850
    },
    {
      "entropy": 0.9485324196517467,
      "epoch": 0.416,
      "grad_norm": 0.35406219959259033,
      "learning_rate": 4.253436322964167e-05,
      "loss": 0.9377,
      "mean_token_accuracy": 0.8116884425282478,
      "num_tokens": 21941174.0,
      "step": 2860
    },
    {
      "entropy": 0.9417855083942414,
      "epoch": 0.41745454545454547,
      "grad_norm": 0.32201579213142395,
      "learning_rate": 4.243003395624094e-05,
      "loss": 0.914,
      "mean_token_accuracy": 0.8134081222116947,
      "num_tokens": 22015943.0,
      "step": 2870
    },
    {
      "entropy": 0.9514484673738479,
      "epoch": 0.4189090909090909,
      "grad_norm": 0.30800074338912964,
      "learning_rate": 4.232511094846687e-05,
      "loss": 0.9274,
      "mean_token_accuracy": 0.8104920752346516,
      "num_tokens": 22096053.0,
      "step": 2880
    },
    {
      "entropy": 0.9354054771363736,
      "epoch": 0.4203636363636364,
      "grad_norm": 0.3503027856349945,
      "learning_rate": 4.221959779068509e-05,
      "loss": 0.9115,
      "mean_token_accuracy": 0.8127908557653427,
      "num_tokens": 22171394.0,
      "step": 2890
    },
    {
      "entropy": 0.9206963747739791,
      "epoch": 0.4218181818181818,
      "grad_norm": 0.2915414571762085,
      "learning_rate": 4.211349808742184e-05,
      "loss": 0.9021,
      "mean_token_accuracy": 0.8156040996313095,
      "num_tokens": 22247511.0,
      "step": 2900
    },
    {
      "entropy": 0.9469660073518753,
      "epoch": 0.4232727272727273,
      "grad_norm": 0.30320265889167786,
      "learning_rate": 4.200681546324087e-05,
      "loss": 0.9361,
      "mean_token_accuracy": 0.8108836710453033,
      "num_tokens": 22323687.0,
      "step": 2910
    },
    {
      "entropy": 0.9925447411835193,
      "epoch": 0.42472727272727273,
      "grad_norm": 0.287954717874527,
      "learning_rate": 4.189955356261956e-05,
      "loss": 0.9683,
      "mean_token_accuracy": 0.8058648020029068,
      "num_tokens": 22402528.0,
      "step": 2920
    },
    {
      "entropy": 0.9799452804028987,
      "epoch": 0.42618181818181816,
      "grad_norm": 0.5119282007217407,
      "learning_rate": 4.1791716049824496e-05,
      "loss": 0.992,
      "mean_token_accuracy": 0.8043662630021572,
      "num_tokens": 22481118.0,
      "step": 2930
    },
    {
      "entropy": 0.9636691868305206,
      "epoch": 0.42763636363636365,
      "grad_norm": 0.30673807859420776,
      "learning_rate": 4.168330660878623e-05,
      "loss": 0.9516,
      "mean_token_accuracy": 0.8084825851023197,
      "num_tokens": 22558803.0,
      "step": 2940
    },
    {
      "entropy": 0.9445181660354137,
      "epoch": 0.4290909090909091,
      "grad_norm": 0.42131251096725464,
      "learning_rate": 4.1574328942973445e-05,
      "loss": 0.9266,
      "mean_token_accuracy": 0.810728045552969,
      "num_tokens": 22635681.0,
      "step": 2950
    },
    {
      "entropy": 0.9450175166130066,
      "epoch": 0.43054545454545456,
      "grad_norm": 0.30848971009254456,
      "learning_rate": 4.146478677526644e-05,
      "loss": 0.9439,
      "mean_token_accuracy": 0.8095552891492843,
      "num_tokens": 22711102.0,
      "step": 2960
    },
    {
      "entropy": 0.9384726583957672,
      "epoch": 0.432,
      "grad_norm": 0.321064293384552,
      "learning_rate": 4.135468384782996e-05,
      "loss": 0.9135,
      "mean_token_accuracy": 0.8127228446304798,
      "num_tokens": 22787605.0,
      "step": 2970
    },
    {
      "entropy": 0.9518445454537868,
      "epoch": 0.4334545454545455,
      "grad_norm": 0.3238827586174011,
      "learning_rate": 4.1244023921985344e-05,
      "loss": 0.9378,
      "mean_token_accuracy": 0.8076546937227249,
      "num_tokens": 22865947.0,
      "step": 2980
    },
    {
      "entropy": 0.943280752748251,
      "epoch": 0.4349090909090909,
      "grad_norm": 0.330661416053772,
      "learning_rate": 4.1132810778082054e-05,
      "loss": 0.9578,
      "mean_token_accuracy": 0.8113200690597295,
      "num_tokens": 22943810.0,
      "step": 2990
    },
    {
      "entropy": 0.9362864792346954,
      "epoch": 0.43636363636363634,
      "grad_norm": 0.3080872595310211,
      "learning_rate": 4.102104821536848e-05,
      "loss": 0.92,
      "mean_token_accuracy": 0.8104663982987403,
      "num_tokens": 23021690.0,
      "step": 3000
    },
    {
      "epoch": 0.43636363636363634,
      "eval_entropy": 0.9390125914573669,
      "eval_loss": 0.9096689224243164,
      "eval_mean_token_accuracy": 0.8104583250045776,
      "eval_num_tokens": 23021690.0,
      "eval_runtime": 77.2691,
      "eval_samples_per_second": 64.709,
      "eval_steps_per_second": 8.089,
      "step": 3000
    },
    {
      "entropy": 0.949031674861908,
      "epoch": 0.43781818181818183,
      "grad_norm": 0.35485225915908813,
      "learning_rate": 4.0908740051862194e-05,
      "loss": 0.9337,
      "mean_token_accuracy": 0.8110528267920017,
      "num_tokens": 23098757.0,
      "step": 3010
    },
    {
      "entropy": 0.9277644000947476,
      "epoch": 0.43927272727272726,
      "grad_norm": 0.3260344862937927,
      "learning_rate": 4.079589012421953e-05,
      "loss": 0.9086,
      "mean_token_accuracy": 0.8140561677515507,
      "num_tokens": 23174179.0,
      "step": 3020
    },
    {
      "entropy": 0.9273935034871101,
      "epoch": 0.44072727272727275,
      "grad_norm": 0.30524784326553345,
      "learning_rate": 4.068250228760447e-05,
      "loss": 0.9336,
      "mean_token_accuracy": 0.8133032880723476,
      "num_tokens": 23248956.0,
      "step": 3030
    },
    {
      "entropy": 0.9435071803629398,
      "epoch": 0.4421818181818182,
      "grad_norm": 0.37396615743637085,
      "learning_rate": 4.056858041555698e-05,
      "loss": 0.9155,
      "mean_token_accuracy": 0.8119289055466652,
      "num_tokens": 23326368.0,
      "step": 3040
    },
    {
      "entropy": 0.9278220035135746,
      "epoch": 0.44363636363636366,
      "grad_norm": 0.3182052969932556,
      "learning_rate": 4.045412839986067e-05,
      "loss": 0.9144,
      "mean_token_accuracy": 0.8117367446422576,
      "num_tokens": 23401650.0,
      "step": 3050
    },
    {
      "entropy": 0.9288572505116462,
      "epoch": 0.4450909090909091,
      "grad_norm": 0.39195388555526733,
      "learning_rate": 4.033915015040984e-05,
      "loss": 0.9185,
      "mean_token_accuracy": 0.8136234790086746,
      "num_tokens": 23477886.0,
      "step": 3060
    },
    {
      "entropy": 0.9454001143574715,
      "epoch": 0.4465454545454545,
      "grad_norm": 0.3858128786087036,
      "learning_rate": 4.022364959507595e-05,
      "loss": 0.9246,
      "mean_token_accuracy": 0.8114547491073608,
      "num_tokens": 23554655.0,
      "step": 3070
    },
    {
      "entropy": 0.9290557987987995,
      "epoch": 0.448,
      "grad_norm": 0.32388290762901306,
      "learning_rate": 4.0107630679573345e-05,
      "loss": 0.9083,
      "mean_token_accuracy": 0.8145299278199672,
      "num_tokens": 23631947.0,
      "step": 3080
    },
    {
      "entropy": 0.9243955463171005,
      "epoch": 0.44945454545454544,
      "grad_norm": 0.37549296021461487,
      "learning_rate": 3.999109736732456e-05,
      "loss": 0.9141,
      "mean_token_accuracy": 0.8141500823199749,
      "num_tokens": 23707907.0,
      "step": 3090
    },
    {
      "entropy": 0.9421398542821408,
      "epoch": 0.4509090909090909,
      "grad_norm": 0.34024038910865784,
      "learning_rate": 3.987405363932488e-05,
      "loss": 0.9153,
      "mean_token_accuracy": 0.8122914619743824,
      "num_tokens": 23784020.0,
      "step": 3100
    },
    {
      "entropy": 0.9259913168847561,
      "epoch": 0.45236363636363636,
      "grad_norm": 0.3002021014690399,
      "learning_rate": 3.9756503494006345e-05,
      "loss": 0.9104,
      "mean_token_accuracy": 0.8137310676276683,
      "num_tokens": 23861070.0,
      "step": 3110
    },
    {
      "entropy": 0.9394382186233997,
      "epoch": 0.45381818181818184,
      "grad_norm": 0.32609429955482483,
      "learning_rate": 3.963845094710113e-05,
      "loss": 0.9334,
      "mean_token_accuracy": 0.8116750314831733,
      "num_tokens": 23937344.0,
      "step": 3120
    },
    {
      "entropy": 0.9637476593255997,
      "epoch": 0.4552727272727273,
      "grad_norm": 0.3102467954158783,
      "learning_rate": 3.95199000315044e-05,
      "loss": 0.9483,
      "mean_token_accuracy": 0.807917469739914,
      "num_tokens": 24016141.0,
      "step": 3130
    },
    {
      "entropy": 0.9454035952687263,
      "epoch": 0.4567272727272727,
      "grad_norm": 0.291983425617218,
      "learning_rate": 3.940085479713655e-05,
      "loss": 0.9478,
      "mean_token_accuracy": 0.812795314192772,
      "num_tokens": 24091590.0,
      "step": 3140
    },
    {
      "entropy": 0.9189323402941227,
      "epoch": 0.4581818181818182,
      "grad_norm": 0.3117421865463257,
      "learning_rate": 3.928131931080479e-05,
      "loss": 0.9037,
      "mean_token_accuracy": 0.8151950173079967,
      "num_tokens": 24167519.0,
      "step": 3150
    },
    {
      "entropy": 0.9363495454192161,
      "epoch": 0.4596363636363636,
      "grad_norm": 0.34000906348228455,
      "learning_rate": 3.9161297656064266e-05,
      "loss": 0.9246,
      "mean_token_accuracy": 0.8120167836546898,
      "num_tokens": 24246228.0,
      "step": 3160
    },
    {
      "entropy": 0.9339604489505291,
      "epoch": 0.4610909090909091,
      "grad_norm": 0.3630552589893341,
      "learning_rate": 3.9040793933078526e-05,
      "loss": 0.9156,
      "mean_token_accuracy": 0.8135395906865597,
      "num_tokens": 24322463.0,
      "step": 3170
    },
    {
      "entropy": 0.9174147740006446,
      "epoch": 0.46254545454545454,
      "grad_norm": 0.3118632733821869,
      "learning_rate": 3.89198122584795e-05,
      "loss": 0.9055,
      "mean_token_accuracy": 0.8144961826503276,
      "num_tokens": 24396716.0,
      "step": 3180
    },
    {
      "entropy": 0.9205701947212219,
      "epoch": 0.464,
      "grad_norm": 0.36817336082458496,
      "learning_rate": 3.8798356765226835e-05,
      "loss": 0.9009,
      "mean_token_accuracy": 0.8156480133533478,
      "num_tokens": 24473065.0,
      "step": 3190
    },
    {
      "entropy": 0.937190480530262,
      "epoch": 0.46545454545454545,
      "grad_norm": 0.31383636593818665,
      "learning_rate": 3.8676431602466664e-05,
      "loss": 0.9317,
      "mean_token_accuracy": 0.8110474929213524,
      "num_tokens": 24552959.0,
      "step": 3200
    },
    {
      "entropy": 0.9277811951935291,
      "epoch": 0.4669090909090909,
      "grad_norm": 0.37277960777282715,
      "learning_rate": 3.855404093538994e-05,
      "loss": 0.9132,
      "mean_token_accuracy": 0.8147149369120598,
      "num_tokens": 24629697.0,
      "step": 3210
    },
    {
      "entropy": 0.939825527369976,
      "epoch": 0.46836363636363637,
      "grad_norm": 0.37860503792762756,
      "learning_rate": 3.843118894509012e-05,
      "loss": 0.9191,
      "mean_token_accuracy": 0.8115852467715741,
      "num_tokens": 24707230.0,
      "step": 3220
    },
    {
      "entropy": 0.9355527043342591,
      "epoch": 0.4698181818181818,
      "grad_norm": 0.40541505813598633,
      "learning_rate": 3.83078798284203e-05,
      "loss": 0.9193,
      "mean_token_accuracy": 0.8121265314519406,
      "num_tokens": 24784950.0,
      "step": 3230
    },
    {
      "entropy": 0.9393999978899956,
      "epoch": 0.4712727272727273,
      "grad_norm": 0.31179094314575195,
      "learning_rate": 3.818411779784988e-05,
      "loss": 0.927,
      "mean_token_accuracy": 0.8106306716799736,
      "num_tokens": 24861205.0,
      "step": 3240
    },
    {
      "entropy": 0.9308439895510674,
      "epoch": 0.4727272727272727,
      "grad_norm": 0.3248290419578552,
      "learning_rate": 3.805990708132064e-05,
      "loss": 0.9101,
      "mean_token_accuracy": 0.8138051718473435,
      "num_tokens": 24937478.0,
      "step": 3250
    },
    {
      "entropy": 0.9559439294040203,
      "epoch": 0.4741818181818182,
      "grad_norm": 0.3539363145828247,
      "learning_rate": 3.7935251922102294e-05,
      "loss": 0.9429,
      "mean_token_accuracy": 0.8100346833467483,
      "num_tokens": 25014666.0,
      "step": 3260
    },
    {
      "entropy": 0.9356847994029522,
      "epoch": 0.47563636363636363,
      "grad_norm": 0.3429943323135376,
      "learning_rate": 3.781015657864757e-05,
      "loss": 0.9387,
      "mean_token_accuracy": 0.8129906997084617,
      "num_tokens": 25090445.0,
      "step": 3270
    },
    {
      "entropy": 0.9545519553124905,
      "epoch": 0.47709090909090907,
      "grad_norm": 0.34916821122169495,
      "learning_rate": 3.768462532444669e-05,
      "loss": 0.9261,
      "mean_token_accuracy": 0.8117854639887809,
      "num_tokens": 25167108.0,
      "step": 3280
    },
    {
      "entropy": 0.9233941897749901,
      "epoch": 0.47854545454545455,
      "grad_norm": 0.35967811942100525,
      "learning_rate": 3.755866244788143e-05,
      "loss": 0.9086,
      "mean_token_accuracy": 0.8152607150375843,
      "num_tokens": 25240549.0,
      "step": 3290
    },
    {
      "entropy": 0.9435656048357487,
      "epoch": 0.48,
      "grad_norm": 0.34135791659355164,
      "learning_rate": 3.743227225207855e-05,
      "loss": 0.9281,
      "mean_token_accuracy": 0.8106845133006573,
      "num_tokens": 25318449.0,
      "step": 3300
    },
    {
      "entropy": 0.9465586185455322,
      "epoch": 0.48145454545454547,
      "grad_norm": 0.3166083097457886,
      "learning_rate": 3.730545905476286e-05,
      "loss": 0.9275,
      "mean_token_accuracy": 0.8122962072491646,
      "num_tokens": 25397039.0,
      "step": 3310
    },
    {
      "entropy": 0.937959486246109,
      "epoch": 0.4829090909090909,
      "grad_norm": 0.3031077980995178,
      "learning_rate": 3.71782271881097e-05,
      "loss": 0.9299,
      "mean_token_accuracy": 0.8110368713736534,
      "num_tokens": 25473863.0,
      "step": 3320
    },
    {
      "entropy": 0.9503463953733444,
      "epoch": 0.4843636363636364,
      "grad_norm": 0.3423995077610016,
      "learning_rate": 3.70505809985969e-05,
      "loss": 0.9321,
      "mean_token_accuracy": 0.810620927810669,
      "num_tokens": 25550661.0,
      "step": 3330
    },
    {
      "entropy": 0.9261522769927979,
      "epoch": 0.4858181818181818,
      "grad_norm": 0.3627229332923889,
      "learning_rate": 3.692252484685637e-05,
      "loss": 0.9108,
      "mean_token_accuracy": 0.8140110865235328,
      "num_tokens": 25626970.0,
      "step": 3340
    },
    {
      "entropy": 0.9232843138277531,
      "epoch": 0.48727272727272725,
      "grad_norm": 0.35451003909111023,
      "learning_rate": 3.679406310752506e-05,
      "loss": 0.912,
      "mean_token_accuracy": 0.8130640156567097,
      "num_tokens": 25702592.0,
      "step": 3350
    },
    {
      "entropy": 0.9272889249026776,
      "epoch": 0.48872727272727273,
      "grad_norm": 0.31371548771858215,
      "learning_rate": 3.6665200169095555e-05,
      "loss": 0.9111,
      "mean_token_accuracy": 0.8153925858438015,
      "num_tokens": 25778382.0,
      "step": 3360
    },
    {
      "entropy": 0.9283372424542904,
      "epoch": 0.49018181818181816,
      "grad_norm": 0.35514751076698303,
      "learning_rate": 3.653594043376618e-05,
      "loss": 0.9305,
      "mean_token_accuracy": 0.811019615828991,
      "num_tokens": 25856942.0,
      "step": 3370
    },
    {
      "entropy": 0.9131379060447216,
      "epoch": 0.49163636363636365,
      "grad_norm": 0.4405730962753296,
      "learning_rate": 3.640628831729055e-05,
      "loss": 0.9011,
      "mean_token_accuracy": 0.8163982145488262,
      "num_tokens": 25932344.0,
      "step": 3380
    },
    {
      "entropy": 0.9400516137480736,
      "epoch": 0.4930909090909091,
      "grad_norm": 0.2962123453617096,
      "learning_rate": 3.627624824882676e-05,
      "loss": 0.9219,
      "mean_token_accuracy": 0.8115826070308685,
      "num_tokens": 26009940.0,
      "step": 3390
    },
    {
      "entropy": 0.9201727285981178,
      "epoch": 0.49454545454545457,
      "grad_norm": 0.3848011791706085,
      "learning_rate": 3.6145824670786094e-05,
      "loss": 0.9069,
      "mean_token_accuracy": 0.813917699456215,
      "num_tokens": 26087646.0,
      "step": 3400
    },
    {
      "entropy": 0.9200171507894993,
      "epoch": 0.496,
      "grad_norm": 0.36265838146209717,
      "learning_rate": 3.6015022038681176e-05,
      "loss": 0.9159,
      "mean_token_accuracy": 0.8137690626084805,
      "num_tokens": 26164658.0,
      "step": 3410
    },
    {
      "entropy": 0.9186624906957149,
      "epoch": 0.4974545454545454,
      "grad_norm": 0.3266400098800659,
      "learning_rate": 3.588384482097389e-05,
      "loss": 0.9082,
      "mean_token_accuracy": 0.8157230131328106,
      "num_tokens": 26239949.0,
      "step": 3420
    },
    {
      "entropy": 0.9403476119041443,
      "epoch": 0.4989090909090909,
      "grad_norm": 0.3292923867702484,
      "learning_rate": 3.575229749892263e-05,
      "loss": 0.9266,
      "mean_token_accuracy": 0.8107123725116253,
      "num_tokens": 26318213.0,
      "step": 3430
    },
    {
      "entropy": 0.9599842190742492,
      "epoch": 0.5003636363636363,
      "grad_norm": 0.30845844745635986,
      "learning_rate": 3.5620384566429246e-05,
      "loss": 0.9916,
      "mean_token_accuracy": 0.8076838113367557,
      "num_tokens": 26394324.0,
      "step": 3440
    },
    {
      "entropy": 0.9396763563156127,
      "epoch": 0.5018181818181818,
      "grad_norm": 0.30212074518203735,
      "learning_rate": 3.5488110529885533e-05,
      "loss": 0.9253,
      "mean_token_accuracy": 0.8118289895355701,
      "num_tokens": 26470681.0,
      "step": 3450
    },
    {
      "entropy": 0.930367311090231,
      "epoch": 0.5032727272727273,
      "grad_norm": 0.2892606854438782,
      "learning_rate": 3.5355479908019265e-05,
      "loss": 0.9076,
      "mean_token_accuracy": 0.8147512383759021,
      "num_tokens": 26548509.0,
      "step": 3460
    },
    {
      "entropy": 0.9388787277042866,
      "epoch": 0.5047272727272727,
      "grad_norm": 0.3451310396194458,
      "learning_rate": 3.5222497231739834e-05,
      "loss": 0.927,
      "mean_token_accuracy": 0.8111469961702824,
      "num_tokens": 26627444.0,
      "step": 3470
    },
    {
      "entropy": 0.9242687962949276,
      "epoch": 0.5061818181818182,
      "grad_norm": 0.3954116702079773,
      "learning_rate": 3.508916704398348e-05,
      "loss": 0.9081,
      "mean_token_accuracy": 0.8125782743096351,
      "num_tokens": 26704031.0,
      "step": 3480
    },
    {
      "entropy": 0.913933489471674,
      "epoch": 0.5076363636363637,
      "grad_norm": 0.33013808727264404,
      "learning_rate": 3.495549389955806e-05,
      "loss": 0.9086,
      "mean_token_accuracy": 0.8151891089975833,
      "num_tokens": 26780389.0,
      "step": 3490
    },
    {
      "entropy": 0.9393781751394272,
      "epoch": 0.509090909090909,
      "grad_norm": 0.3132812976837158,
      "learning_rate": 3.482148236498751e-05,
      "loss": 0.9255,
      "mean_token_accuracy": 0.8121484272181988,
      "num_tokens": 26856489.0,
      "step": 3500
    },
    {
      "entropy": 0.9314693167805672,
      "epoch": 0.5105454545454545,
      "grad_norm": 0.36915311217308044,
      "learning_rate": 3.468713701835575e-05,
      "loss": 0.9101,
      "mean_token_accuracy": 0.8132783479988575,
      "num_tokens": 26930832.0,
      "step": 3510
    },
    {
      "entropy": 0.915811663120985,
      "epoch": 0.512,
      "grad_norm": 0.33095115423202515,
      "learning_rate": 3.455246244915039e-05,
      "loss": 0.9013,
      "mean_token_accuracy": 0.8143532752990723,
      "num_tokens": 27005645.0,
      "step": 3520
    },
    {
      "entropy": 0.9287936054170132,
      "epoch": 0.5134545454545455,
      "grad_norm": 0.3176046311855316,
      "learning_rate": 3.441746325810587e-05,
      "loss": 0.9234,
      "mean_token_accuracy": 0.8127094700932502,
      "num_tokens": 27082589.0,
      "step": 3530
    },
    {
      "entropy": 0.9292538709938526,
      "epoch": 0.5149090909090909,
      "grad_norm": 0.4348084628582001,
      "learning_rate": 3.428214405704634e-05,
      "loss": 0.9212,
      "mean_token_accuracy": 0.8118156298995018,
      "num_tokens": 27158602.0,
      "step": 3540
    },
    {
      "entropy": 0.9403552688658238,
      "epoch": 0.5163636363636364,
      "grad_norm": 0.29724040627479553,
      "learning_rate": 3.414650946872807e-05,
      "loss": 0.9223,
      "mean_token_accuracy": 0.8112016417086124,
      "num_tokens": 27234799.0,
      "step": 3550
    },
    {
      "entropy": 0.9113413311541081,
      "epoch": 0.5178181818181818,
      "grad_norm": 0.3209291994571686,
      "learning_rate": 3.4010564126681554e-05,
      "loss": 0.8956,
      "mean_token_accuracy": 0.8150275580585002,
      "num_tokens": 27310411.0,
      "step": 3560
    },
    {
      "entropy": 0.9307005263864994,
      "epoch": 0.5192727272727272,
      "grad_norm": 0.33619439601898193,
      "learning_rate": 3.38743126750532e-05,
      "loss": 0.9278,
      "mean_token_accuracy": 0.8099561981856823,
      "num_tokens": 27387609.0,
      "step": 3570
    },
    {
      "entropy": 0.9498361259698868,
      "epoch": 0.5207272727272727,
      "grad_norm": 0.3329075574874878,
      "learning_rate": 3.3737759768446716e-05,
      "loss": 0.9246,
      "mean_token_accuracy": 0.8113810978829861,
      "num_tokens": 27464690.0,
      "step": 3580
    },
    {
      "entropy": 0.9319490067660808,
      "epoch": 0.5221818181818182,
      "grad_norm": 0.2935169041156769,
      "learning_rate": 3.360091007176405e-05,
      "loss": 0.9231,
      "mean_token_accuracy": 0.8126781105995178,
      "num_tokens": 27542426.0,
      "step": 3590
    },
    {
      "entropy": 0.9247090421617031,
      "epoch": 0.5236363636363637,
      "grad_norm": 0.3140055537223816,
      "learning_rate": 3.346376826004607e-05,
      "loss": 0.9169,
      "mean_token_accuracy": 0.8137174978852272,
      "num_tokens": 27618489.0,
      "step": 3600
    },
    {
      "entropy": 0.9323613822460175,
      "epoch": 0.525090909090909,
      "grad_norm": 0.3542056083679199,
      "learning_rate": 3.3326339018312825e-05,
      "loss": 0.9169,
      "mean_token_accuracy": 0.8120187312364578,
      "num_tokens": 27693540.0,
      "step": 3610
    },
    {
      "entropy": 0.9273739039897919,
      "epoch": 0.5265454545454545,
      "grad_norm": 0.3772445023059845,
      "learning_rate": 3.318862704140352e-05,
      "loss": 0.9018,
      "mean_token_accuracy": 0.8152659721672535,
      "num_tokens": 27768991.0,
      "step": 3620
    },
    {
      "entropy": 0.9225555047392845,
      "epoch": 0.528,
      "grad_norm": 0.30556240677833557,
      "learning_rate": 3.305063703381611e-05,
      "loss": 0.914,
      "mean_token_accuracy": 0.8134266242384911,
      "num_tokens": 27847456.0,
      "step": 3630
    },
    {
      "entropy": 0.9254167668521405,
      "epoch": 0.5294545454545454,
      "grad_norm": 0.3305240273475647,
      "learning_rate": 3.291237370954661e-05,
      "loss": 0.9133,
      "mean_token_accuracy": 0.8138591662049294,
      "num_tokens": 27922870.0,
      "step": 3640
    },
    {
      "entropy": 0.9333068393170834,
      "epoch": 0.5309090909090909,
      "grad_norm": 0.4308319389820099,
      "learning_rate": 3.2773841791928026e-05,
      "loss": 0.9195,
      "mean_token_accuracy": 0.8125985987484455,
      "num_tokens": 27999197.0,
      "step": 3650
    },
    {
      "entropy": 0.9453266486525536,
      "epoch": 0.5323636363636364,
      "grad_norm": 0.35205429792404175,
      "learning_rate": 3.263504601346904e-05,
      "loss": 0.9364,
      "mean_token_accuracy": 0.8111186057329178,
      "num_tokens": 28075063.0,
      "step": 3660
    },
    {
      "entropy": 0.9272417090833187,
      "epoch": 0.5338181818181819,
      "grad_norm": 0.2925795912742615,
      "learning_rate": 3.249599111569227e-05,
      "loss": 0.915,
      "mean_token_accuracy": 0.8154182583093643,
      "num_tokens": 28151467.0,
      "step": 3670
    },
    {
      "entropy": 0.9294666975736618,
      "epoch": 0.5352727272727272,
      "grad_norm": 0.3383394777774811,
      "learning_rate": 3.2356681848972384e-05,
      "loss": 0.9129,
      "mean_token_accuracy": 0.8128080390393734,
      "num_tokens": 28228643.0,
      "step": 3680
    },
    {
      "entropy": 0.9178059481084346,
      "epoch": 0.5367272727272727,
      "grad_norm": 0.4347728192806244,
      "learning_rate": 3.221712297237372e-05,
      "loss": 0.9097,
      "mean_token_accuracy": 0.8125108242034912,
      "num_tokens": 28304443.0,
      "step": 3690
    },
    {
      "entropy": 0.9356210947036743,
      "epoch": 0.5381818181818182,
      "grad_norm": 0.3118355870246887,
      "learning_rate": 3.207731925348779e-05,
      "loss": 0.9301,
      "mean_token_accuracy": 0.81150324344635,
      "num_tokens": 28381759.0,
      "step": 3700
    },
    {
      "entropy": 0.9528975121676921,
      "epoch": 0.5396363636363637,
      "grad_norm": 0.30240732431411743,
      "learning_rate": 3.193727546827037e-05,
      "loss": 0.9445,
      "mean_token_accuracy": 0.8085363268852234,
      "num_tokens": 28460963.0,
      "step": 3710
    },
    {
      "entropy": 0.9330151177942753,
      "epoch": 0.5410909090909091,
      "grad_norm": 0.3076750338077545,
      "learning_rate": 3.1796996400878315e-05,
      "loss": 0.9237,
      "mean_token_accuracy": 0.812045356631279,
      "num_tokens": 28538470.0,
      "step": 3720
    },
    {
      "entropy": 0.9240581601858139,
      "epoch": 0.5425454545454546,
      "grad_norm": 0.3159777522087097,
      "learning_rate": 3.165648684350621e-05,
      "loss": 0.9348,
      "mean_token_accuracy": 0.8137698240578175,
      "num_tokens": 28613263.0,
      "step": 3730
    },
    {
      "entropy": 0.9310328140854836,
      "epoch": 0.544,
      "grad_norm": 0.3062320053577423,
      "learning_rate": 3.151575159622257e-05,
      "loss": 0.9294,
      "mean_token_accuracy": 0.8146270386874676,
      "num_tokens": 28687810.0,
      "step": 3740
    },
    {
      "entropy": 0.9495957009494305,
      "epoch": 0.5454545454545454,
      "grad_norm": 0.33293402194976807,
      "learning_rate": 3.137479546680592e-05,
      "loss": 0.9326,
      "mean_token_accuracy": 0.8113441608846188,
      "num_tokens": 28764721.0,
      "step": 3750
    },
    {
      "entropy": 0.9448008291423321,
      "epoch": 0.5469090909090909,
      "grad_norm": 0.29420945048332214,
      "learning_rate": 3.123362327058051e-05,
      "loss": 0.93,
      "mean_token_accuracy": 0.8090423189103604,
      "num_tokens": 28841296.0,
      "step": 3760
    },
    {
      "entropy": 0.9217709936201572,
      "epoch": 0.5483636363636364,
      "grad_norm": 0.3123224377632141,
      "learning_rate": 3.109223983025184e-05,
      "loss": 0.9071,
      "mean_token_accuracy": 0.8159178838133812,
      "num_tokens": 28915919.0,
      "step": 3770
    },
    {
      "entropy": 0.9265149988234043,
      "epoch": 0.5498181818181819,
      "grad_norm": 0.34357714653015137,
      "learning_rate": 3.095064997574194e-05,
      "loss": 0.9121,
      "mean_token_accuracy": 0.8129103809595108,
      "num_tokens": 28993630.0,
      "step": 3780
    },
    {
      "entropy": 0.9530934929847718,
      "epoch": 0.5512727272727272,
      "grad_norm": 0.3064063489437103,
      "learning_rate": 3.080885854402426e-05,
      "loss": 0.9261,
      "mean_token_accuracy": 0.8121055580675602,
      "num_tokens": 29070598.0,
      "step": 3790
    },
    {
      "entropy": 0.9290414094924927,
      "epoch": 0.5527272727272727,
      "grad_norm": 0.30586880445480347,
      "learning_rate": 3.066687037895858e-05,
      "loss": 0.9143,
      "mean_token_accuracy": 0.8123991221189499,
      "num_tokens": 29149139.0,
      "step": 3800
    },
    {
      "entropy": 0.9213745802640915,
      "epoch": 0.5541818181818182,
      "grad_norm": 0.31518247723579407,
      "learning_rate": 3.052469033112541e-05,
      "loss": 0.9126,
      "mean_token_accuracy": 0.8130135543644428,
      "num_tokens": 29224329.0,
      "step": 3810
    },
    {
      "entropy": 0.9490518577396869,
      "epoch": 0.5556363636363636,
      "grad_norm": 0.38361936807632446,
      "learning_rate": 3.0382323257660343e-05,
      "loss": 0.9322,
      "mean_token_accuracy": 0.8099796429276467,
      "num_tokens": 29298869.0,
      "step": 3820
    },
    {
      "entropy": 0.9164307072758675,
      "epoch": 0.5570909090909091,
      "grad_norm": 0.3322400748729706,
      "learning_rate": 3.023977402208813e-05,
      "loss": 0.9017,
      "mean_token_accuracy": 0.8168147139251232,
      "num_tokens": 29374649.0,
      "step": 3830
    },
    {
      "entropy": 0.9352445684373378,
      "epoch": 0.5585454545454546,
      "grad_norm": 0.3158262372016907,
      "learning_rate": 3.0097047494156492e-05,
      "loss": 0.9155,
      "mean_token_accuracy": 0.8125953778624535,
      "num_tokens": 29450368.0,
      "step": 3840
    },
    {
      "entropy": 0.9447495467960835,
      "epoch": 0.56,
      "grad_norm": 0.4749705195426941,
      "learning_rate": 2.995414854966983e-05,
      "loss": 0.9305,
      "mean_token_accuracy": 0.8096226342022419,
      "num_tokens": 29528900.0,
      "step": 3850
    },
    {
      "entropy": 0.9433772042393684,
      "epoch": 0.5614545454545454,
      "grad_norm": 0.30530375242233276,
      "learning_rate": 2.981108207032257e-05,
      "loss": 0.9351,
      "mean_token_accuracy": 0.8102800861001015,
      "num_tokens": 29606098.0,
      "step": 3860
    },
    {
      "entropy": 0.9182851135730743,
      "epoch": 0.5629090909090909,
      "grad_norm": 0.3289918005466461,
      "learning_rate": 2.966785294353248e-05,
      "loss": 0.9045,
      "mean_token_accuracy": 0.8157595112919808,
      "num_tokens": 29683042.0,
      "step": 3870
    },
    {
      "entropy": 0.940113590657711,
      "epoch": 0.5643636363636364,
      "grad_norm": 0.3084245026111603,
      "learning_rate": 2.9524466062273637e-05,
      "loss": 0.9618,
      "mean_token_accuracy": 0.80933144018054,
      "num_tokens": 29761341.0,
      "step": 3880
    },
    {
      "entropy": 0.9369757734239101,
      "epoch": 0.5658181818181818,
      "grad_norm": 0.3422188460826874,
      "learning_rate": 2.9380926324909326e-05,
      "loss": 0.9148,
      "mean_token_accuracy": 0.8134120874106884,
      "num_tokens": 29838964.0,
      "step": 3890
    },
    {
      "entropy": 0.920634201169014,
      "epoch": 0.5672727272727273,
      "grad_norm": 0.315481960773468,
      "learning_rate": 2.9237238635024655e-05,
      "loss": 0.9019,
      "mean_token_accuracy": 0.8159717105329036,
      "num_tokens": 29913008.0,
      "step": 3900
    },
    {
      "entropy": 0.926985715329647,
      "epoch": 0.5687272727272727,
      "grad_norm": 0.31134775280952454,
      "learning_rate": 2.9093407901259113e-05,
      "loss": 0.9133,
      "mean_token_accuracy": 0.8131841793656349,
      "num_tokens": 29989402.0,
      "step": 3910
    },
    {
      "entropy": 0.9281412333250045,
      "epoch": 0.5701818181818182,
      "grad_norm": 0.30873990058898926,
      "learning_rate": 2.8949439037138787e-05,
      "loss": 0.9139,
      "mean_token_accuracy": 0.8122556217014789,
      "num_tokens": 30066188.0,
      "step": 3920
    },
    {
      "entropy": 0.9479538016021252,
      "epoch": 0.5716363636363636,
      "grad_norm": 0.29268190264701843,
      "learning_rate": 2.880533696090858e-05,
      "loss": 0.9255,
      "mean_token_accuracy": 0.8109686590731144,
      "num_tokens": 30145885.0,
      "step": 3930
    },
    {
      "entropy": 0.9348795086145401,
      "epoch": 0.5730909090909091,
      "grad_norm": 0.2962764799594879,
      "learning_rate": 2.866110659536418e-05,
      "loss": 0.9204,
      "mean_token_accuracy": 0.8116168014705181,
      "num_tokens": 30221980.0,
      "step": 3940
    },
    {
      "entropy": 0.9421909719705581,
      "epoch": 0.5745454545454546,
      "grad_norm": 0.37171006202697754,
      "learning_rate": 2.8516752867683834e-05,
      "loss": 0.9345,
      "mean_token_accuracy": 0.809541080147028,
      "num_tokens": 30298060.0,
      "step": 3950
    },
    {
      "entropy": 0.9366579465568066,
      "epoch": 0.576,
      "grad_norm": 0.3353271484375,
      "learning_rate": 2.8372280709260112e-05,
      "loss": 0.9193,
      "mean_token_accuracy": 0.8101016081869602,
      "num_tokens": 30375772.0,
      "step": 3960
    },
    {
      "entropy": 0.9318700306117534,
      "epoch": 0.5774545454545454,
      "grad_norm": 0.30664509534835815,
      "learning_rate": 2.8227695055531372e-05,
      "loss": 0.9232,
      "mean_token_accuracy": 0.811845350265503,
      "num_tokens": 30453404.0,
      "step": 3970
    },
    {
      "entropy": 0.9198355250060558,
      "epoch": 0.5789090909090909,
      "grad_norm": 0.3414006233215332,
      "learning_rate": 2.80830008458132e-05,
      "loss": 0.9084,
      "mean_token_accuracy": 0.8146342925727368,
      "num_tokens": 30531652.0,
      "step": 3980
    },
    {
      "entropy": 0.9534987896680832,
      "epoch": 0.5803636363636364,
      "grad_norm": 0.3388223648071289,
      "learning_rate": 2.7938203023129655e-05,
      "loss": 0.9357,
      "mean_token_accuracy": 0.8100730545818806,
      "num_tokens": 30608793.0,
      "step": 3990
    },
    {
      "entropy": 0.9408224396407604,
      "epoch": 0.5818181818181818,
      "grad_norm": 0.4285614788532257,
      "learning_rate": 2.7793306534044407e-05,
      "loss": 0.9233,
      "mean_token_accuracy": 0.8130472995340824,
      "num_tokens": 30684417.0,
      "step": 4000
    },
    {
      "epoch": 0.5818181818181818,
      "eval_entropy": 0.9120176056861877,
      "eval_loss": 0.9038572311401367,
      "eval_mean_token_accuracy": 0.8106667428016663,
      "eval_num_tokens": 30684417.0,
      "eval_runtime": 77.2157,
      "eval_samples_per_second": 64.754,
      "eval_steps_per_second": 8.094,
      "step": 4000
    },
    {
      "entropy": 0.9213396511971951,
      "epoch": 0.5832727272727273,
      "grad_norm": 0.30884382128715515,
      "learning_rate": 2.764831632849174e-05,
      "loss": 0.9105,
      "mean_token_accuracy": 0.8138939343392849,
      "num_tokens": 30758705.0,
      "step": 4010
    },
    {
      "entropy": 0.9280633635818958,
      "epoch": 0.5847272727272728,
      "grad_norm": 0.3091529607772827,
      "learning_rate": 2.750323735960749e-05,
      "loss": 0.9153,
      "mean_token_accuracy": 0.8144114501774311,
      "num_tokens": 30836054.0,
      "step": 4020
    },
    {
      "entropy": 0.9196830488741398,
      "epoch": 0.5861818181818181,
      "grad_norm": 0.34157034754753113,
      "learning_rate": 2.7358074583559805e-05,
      "loss": 0.896,
      "mean_token_accuracy": 0.8160037316381932,
      "num_tokens": 30910893.0,
      "step": 4030
    },
    {
      "entropy": 0.9281592950224876,
      "epoch": 0.5876363636363636,
      "grad_norm": 0.32032203674316406,
      "learning_rate": 2.7212832959379832e-05,
      "loss": 0.913,
      "mean_token_accuracy": 0.8135712184011936,
      "num_tokens": 30985770.0,
      "step": 4040
    },
    {
      "entropy": 0.9192128323018551,
      "epoch": 0.5890909090909091,
      "grad_norm": 0.30986231565475464,
      "learning_rate": 2.7067517448792328e-05,
      "loss": 0.9043,
      "mean_token_accuracy": 0.8159645989537239,
      "num_tokens": 31062009.0,
      "step": 4050
    },
    {
      "entropy": 0.9104260437190532,
      "epoch": 0.5905454545454546,
      "grad_norm": 0.2997622489929199,
      "learning_rate": 2.6922133016046153e-05,
      "loss": 0.9021,
      "mean_token_accuracy": 0.8168829850852489,
      "num_tokens": 31136291.0,
      "step": 4060
    },
    {
      "entropy": 0.9366477951407433,
      "epoch": 0.592,
      "grad_norm": 0.3440490961074829,
      "learning_rate": 2.6776684627744652e-05,
      "loss": 0.9097,
      "mean_token_accuracy": 0.8135413631796837,
      "num_tokens": 31211763.0,
      "step": 4070
    },
    {
      "entropy": 0.9435247026383877,
      "epoch": 0.5934545454545455,
      "grad_norm": 0.3243034780025482,
      "learning_rate": 2.6631177252676037e-05,
      "loss": 0.9552,
      "mean_token_accuracy": 0.8111616026610136,
      "num_tokens": 31290765.0,
      "step": 4080
    },
    {
      "entropy": 0.9122216440737247,
      "epoch": 0.5949090909090909,
      "grad_norm": 0.3088412582874298,
      "learning_rate": 2.6485615861643615e-05,
      "loss": 0.8996,
      "mean_token_accuracy": 0.8152236327528953,
      "num_tokens": 31365328.0,
      "step": 4090
    },
    {
      "entropy": 0.9196409486234188,
      "epoch": 0.5963636363636363,
      "grad_norm": 0.30076760053634644,
      "learning_rate": 2.6340005427295967e-05,
      "loss": 0.8972,
      "mean_token_accuracy": 0.8156254440546036,
      "num_tokens": 31438526.0,
      "step": 4100
    },
    {
      "entropy": 0.9299661487340927,
      "epoch": 0.5978181818181818,
      "grad_norm": 0.3447272479534149,
      "learning_rate": 2.619435092395708e-05,
      "loss": 0.9164,
      "mean_token_accuracy": 0.8133956313133239,
      "num_tokens": 31514224.0,
      "step": 4110
    },
    {
      "entropy": 0.9186855144798756,
      "epoch": 0.5992727272727273,
      "grad_norm": 0.2987687587738037,
      "learning_rate": 2.604865732745645e-05,
      "loss": 0.9055,
      "mean_token_accuracy": 0.8139917500317096,
      "num_tokens": 31591033.0,
      "step": 4120
    },
    {
      "entropy": 0.9176096372306347,
      "epoch": 0.6007272727272728,
      "grad_norm": 0.304896742105484,
      "learning_rate": 2.590292961495903e-05,
      "loss": 0.8979,
      "mean_token_accuracy": 0.8159045994281768,
      "num_tokens": 31666674.0,
      "step": 4130
    },
    {
      "entropy": 0.9512632295489312,
      "epoch": 0.6021818181818182,
      "grad_norm": 0.3454456627368927,
      "learning_rate": 2.5757172764795272e-05,
      "loss": 0.965,
      "mean_token_accuracy": 0.8078057669103146,
      "num_tokens": 31742137.0,
      "step": 4140
    },
    {
      "entropy": 0.933166479319334,
      "epoch": 0.6036363636363636,
      "grad_norm": 0.4448087215423584,
      "learning_rate": 2.5611391756291015e-05,
      "loss": 0.915,
      "mean_token_accuracy": 0.8134665213525295,
      "num_tokens": 31819369.0,
      "step": 4150
    },
    {
      "entropy": 0.9195875227451324,
      "epoch": 0.6050909090909091,
      "grad_norm": 0.4040326774120331,
      "learning_rate": 2.5465591569597376e-05,
      "loss": 0.9083,
      "mean_token_accuracy": 0.8156167544424534,
      "num_tokens": 31894986.0,
      "step": 4160
    },
    {
      "entropy": 0.9244043946266174,
      "epoch": 0.6065454545454545,
      "grad_norm": 0.31537356972694397,
      "learning_rate": 2.5319777185520637e-05,
      "loss": 0.9157,
      "mean_token_accuracy": 0.8129312485456467,
      "num_tokens": 31970758.0,
      "step": 4170
    },
    {
      "entropy": 0.9341804251074791,
      "epoch": 0.608,
      "grad_norm": 0.30743053555488586,
      "learning_rate": 2.517395358535212e-05,
      "loss": 0.9069,
      "mean_token_accuracy": 0.8125176221132279,
      "num_tokens": 32046461.0,
      "step": 4180
    },
    {
      "entropy": 0.9229583241045475,
      "epoch": 0.6094545454545455,
      "grad_norm": 0.2892451584339142,
      "learning_rate": 2.5028125750697956e-05,
      "loss": 0.9167,
      "mean_token_accuracy": 0.8145610079169273,
      "num_tokens": 32124401.0,
      "step": 4190
    },
    {
      "entropy": 0.9227637097239494,
      "epoch": 0.610909090909091,
      "grad_norm": 0.34083735942840576,
      "learning_rate": 2.4882298663308938e-05,
      "loss": 0.9095,
      "mean_token_accuracy": 0.8163429066538811,
      "num_tokens": 32198279.0,
      "step": 4200
    },
    {
      "entropy": 0.9213042184710503,
      "epoch": 0.6123636363636363,
      "grad_norm": 0.31918326020240784,
      "learning_rate": 2.4736477304910353e-05,
      "loss": 0.9113,
      "mean_token_accuracy": 0.8158605769276619,
      "num_tokens": 32274861.0,
      "step": 4210
    },
    {
      "entropy": 0.9200929164886474,
      "epoch": 0.6138181818181818,
      "grad_norm": 0.3179524540901184,
      "learning_rate": 2.459066665703174e-05,
      "loss": 0.8979,
      "mean_token_accuracy": 0.8158435106277466,
      "num_tokens": 32351403.0,
      "step": 4220
    },
    {
      "entropy": 0.9181513540446758,
      "epoch": 0.6152727272727273,
      "grad_norm": 0.3022609353065491,
      "learning_rate": 2.4444871700836773e-05,
      "loss": 0.8922,
      "mean_token_accuracy": 0.8171645559370517,
      "num_tokens": 32426917.0,
      "step": 4230
    },
    {
      "entropy": 0.9327409908175468,
      "epoch": 0.6167272727272727,
      "grad_norm": 0.3763042390346527,
      "learning_rate": 2.4299097416953057e-05,
      "loss": 0.9288,
      "mean_token_accuracy": 0.8123154416680336,
      "num_tokens": 32505611.0,
      "step": 4240
    },
    {
      "entropy": 0.9272146694362163,
      "epoch": 0.6181818181818182,
      "grad_norm": 0.31237301230430603,
      "learning_rate": 2.415334878530198e-05,
      "loss": 0.9159,
      "mean_token_accuracy": 0.8113645888864994,
      "num_tokens": 32584298.0,
      "step": 4250
    },
    {
      "entropy": 0.9192642048001289,
      "epoch": 0.6196363636363637,
      "grad_norm": 0.40542471408843994,
      "learning_rate": 2.400763078492865e-05,
      "loss": 0.9064,
      "mean_token_accuracy": 0.8155330330133438,
      "num_tokens": 32662584.0,
      "step": 4260
    },
    {
      "entropy": 0.9455889090895653,
      "epoch": 0.6210909090909091,
      "grad_norm": 0.3466758131980896,
      "learning_rate": 2.3861948393831693e-05,
      "loss": 0.9806,
      "mean_token_accuracy": 0.8104759208858013,
      "num_tokens": 32739409.0,
      "step": 4270
    },
    {
      "entropy": 0.93989123031497,
      "epoch": 0.6225454545454545,
      "grad_norm": 0.367541640996933,
      "learning_rate": 2.3716306588793298e-05,
      "loss": 0.9274,
      "mean_token_accuracy": 0.8119640313088894,
      "num_tokens": 32816667.0,
      "step": 4280
    },
    {
      "entropy": 0.931169830262661,
      "epoch": 0.624,
      "grad_norm": 0.3801480233669281,
      "learning_rate": 2.357071034520914e-05,
      "loss": 0.9083,
      "mean_token_accuracy": 0.8134308315813541,
      "num_tokens": 32892144.0,
      "step": 4290
    },
    {
      "entropy": 0.926163075864315,
      "epoch": 0.6254545454545455,
      "grad_norm": 0.3418377637863159,
      "learning_rate": 2.342516463691844e-05,
      "loss": 0.9163,
      "mean_token_accuracy": 0.8127011097967625,
      "num_tokens": 32968698.0,
      "step": 4300
    },
    {
      "entropy": 0.958393444865942,
      "epoch": 0.6269090909090909,
      "grad_norm": 0.31486940383911133,
      "learning_rate": 2.327967443603402e-05,
      "loss": 0.9388,
      "mean_token_accuracy": 0.8094369448721409,
      "num_tokens": 33047170.0,
      "step": 4310
    },
    {
      "entropy": 0.9316714957356453,
      "epoch": 0.6283636363636363,
      "grad_norm": 0.5651655197143555,
      "learning_rate": 2.313424471277248e-05,
      "loss": 0.9221,
      "mean_token_accuracy": 0.8118502162396908,
      "num_tokens": 33126959.0,
      "step": 4320
    },
    {
      "entropy": 0.9491320624947548,
      "epoch": 0.6298181818181818,
      "grad_norm": 0.29574310779571533,
      "learning_rate": 2.2988880435284402e-05,
      "loss": 0.946,
      "mean_token_accuracy": 0.8098831824958325,
      "num_tokens": 33209043.0,
      "step": 4330
    },
    {
      "entropy": 0.9107597202062607,
      "epoch": 0.6312727272727273,
      "grad_norm": 0.3819224536418915,
      "learning_rate": 2.2843586569484592e-05,
      "loss": 0.9164,
      "mean_token_accuracy": 0.8156860545277596,
      "num_tokens": 33286119.0,
      "step": 4340
    },
    {
      "entropy": 0.9466536194086075,
      "epoch": 0.6327272727272727,
      "grad_norm": 0.36104846000671387,
      "learning_rate": 2.269836807888249e-05,
      "loss": 0.9356,
      "mean_token_accuracy": 0.8103018321096898,
      "num_tokens": 33365064.0,
      "step": 4350
    },
    {
      "entropy": 0.9229753375053406,
      "epoch": 0.6341818181818182,
      "grad_norm": 0.4024662375450134,
      "learning_rate": 2.2553229924412553e-05,
      "loss": 0.908,
      "mean_token_accuracy": 0.8140193805098533,
      "num_tokens": 33440374.0,
      "step": 4360
    },
    {
      "entropy": 0.9263255707919598,
      "epoch": 0.6356363636363637,
      "grad_norm": 0.4088100492954254,
      "learning_rate": 2.2408177064264824e-05,
      "loss": 0.9084,
      "mean_token_accuracy": 0.8131275482475757,
      "num_tokens": 33516998.0,
      "step": 4370
    },
    {
      "entropy": 0.9208009012043477,
      "epoch": 0.637090909090909,
      "grad_norm": 0.2979762554168701,
      "learning_rate": 2.2263214453715518e-05,
      "loss": 0.9037,
      "mean_token_accuracy": 0.8130398415029049,
      "num_tokens": 33597009.0,
      "step": 4380
    },
    {
      "entropy": 0.9229357041418552,
      "epoch": 0.6385454545454545,
      "grad_norm": 0.3448769748210907,
      "learning_rate": 2.2118347044957774e-05,
      "loss": 0.9173,
      "mean_token_accuracy": 0.8124949038028717,
      "num_tokens": 33674601.0,
      "step": 4390
    },
    {
      "entropy": 0.9393012635409832,
      "epoch": 0.64,
      "grad_norm": 0.33367374539375305,
      "learning_rate": 2.1973579786932434e-05,
      "loss": 0.92,
      "mean_token_accuracy": 0.8117619611322879,
      "num_tokens": 33752265.0,
      "step": 4400
    },
    {
      "entropy": 0.9376168206334115,
      "epoch": 0.6414545454545455,
      "grad_norm": 0.32508036494255066,
      "learning_rate": 2.1828917625159047e-05,
      "loss": 0.9162,
      "mean_token_accuracy": 0.8128947488963604,
      "num_tokens": 33827985.0,
      "step": 4410
    },
    {
      "entropy": 0.9196195363998413,
      "epoch": 0.6429090909090909,
      "grad_norm": 0.3176663815975189,
      "learning_rate": 2.1684365501566832e-05,
      "loss": 0.8974,
      "mean_token_accuracy": 0.8134889304637909,
      "num_tokens": 33904188.0,
      "step": 4420
    },
    {
      "entropy": 0.9178881429135799,
      "epoch": 0.6443636363636364,
      "grad_norm": 0.36815205216407776,
      "learning_rate": 2.1539928354325917e-05,
      "loss": 0.9106,
      "mean_token_accuracy": 0.8146425768733024,
      "num_tokens": 33980757.0,
      "step": 4430
    },
    {
      "entropy": 0.9099920958280563,
      "epoch": 0.6458181818181818,
      "grad_norm": 0.3318701386451721,
      "learning_rate": 2.1395611117678632e-05,
      "loss": 0.8932,
      "mean_token_accuracy": 0.8180915333330632,
      "num_tokens": 34055905.0,
      "step": 4440
    },
    {
      "entropy": 0.9483034580945968,
      "epoch": 0.6472727272727272,
      "grad_norm": 0.3110158145427704,
      "learning_rate": 2.1251418721770917e-05,
      "loss": 0.9315,
      "mean_token_accuracy": 0.8075625628232956,
      "num_tokens": 34133393.0,
      "step": 4450
    },
    {
      "entropy": 0.9322702303528786,
      "epoch": 0.6487272727272727,
      "grad_norm": 0.4107281565666199,
      "learning_rate": 2.1107356092483943e-05,
      "loss": 0.9174,
      "mean_token_accuracy": 0.8124357827007771,
      "num_tokens": 34211203.0,
      "step": 4460
    },
    {
      "entropy": 0.9236367382109165,
      "epoch": 0.6501818181818182,
      "grad_norm": 0.3252826929092407,
      "learning_rate": 2.0963428151265793e-05,
      "loss": 0.9173,
      "mean_token_accuracy": 0.8112751476466655,
      "num_tokens": 34288790.0,
      "step": 4470
    },
    {
      "entropy": 0.939553078263998,
      "epoch": 0.6516363636363637,
      "grad_norm": 0.3687044084072113,
      "learning_rate": 2.0819639814963375e-05,
      "loss": 0.9276,
      "mean_token_accuracy": 0.8118774957954884,
      "num_tokens": 34365549.0,
      "step": 4480
    },
    {
      "entropy": 0.9295577295124531,
      "epoch": 0.653090909090909,
      "grad_norm": 0.28752341866493225,
      "learning_rate": 2.0675995995654405e-05,
      "loss": 0.913,
      "mean_token_accuracy": 0.8133269488811493,
      "num_tokens": 34440923.0,
      "step": 4490
    },
    {
      "entropy": 0.9382489174604416,
      "epoch": 0.6545454545454545,
      "grad_norm": 0.3165334165096283,
      "learning_rate": 2.0532501600479647e-05,
      "loss": 0.9301,
      "mean_token_accuracy": 0.8110613726079464,
      "num_tokens": 34518649.0,
      "step": 4500
    },
    {
      "entropy": 0.9152321480214596,
      "epoch": 0.656,
      "grad_norm": 0.29478880763053894,
      "learning_rate": 2.0389161531475262e-05,
      "loss": 0.9068,
      "mean_token_accuracy": 0.8136940069496632,
      "num_tokens": 34595637.0,
      "step": 4510
    },
    {
      "entropy": 0.9248601853847503,
      "epoch": 0.6574545454545454,
      "grad_norm": 0.32419294118881226,
      "learning_rate": 2.0245980685405313e-05,
      "loss": 0.9107,
      "mean_token_accuracy": 0.8130759067833424,
      "num_tokens": 34672428.0,
      "step": 4520
    },
    {
      "entropy": 0.9113852202892303,
      "epoch": 0.6589090909090909,
      "grad_norm": 0.365728497505188,
      "learning_rate": 2.010296395359456e-05,
      "loss": 0.9063,
      "mean_token_accuracy": 0.8161646939814091,
      "num_tokens": 34748433.0,
      "step": 4530
    },
    {
      "entropy": 0.9144539162516594,
      "epoch": 0.6603636363636364,
      "grad_norm": 0.31549787521362305,
      "learning_rate": 1.9960116221761258e-05,
      "loss": 0.9131,
      "mean_token_accuracy": 0.8138697162270546,
      "num_tokens": 34825874.0,
      "step": 4540
    },
    {
      "entropy": 0.9262143172323704,
      "epoch": 0.6618181818181819,
      "grad_norm": 0.3173759877681732,
      "learning_rate": 1.9817442369850353e-05,
      "loss": 0.9081,
      "mean_token_accuracy": 0.8120600923895835,
      "num_tokens": 34904254.0,
      "step": 4550
    },
    {
      "entropy": 0.9321594141423702,
      "epoch": 0.6632727272727272,
      "grad_norm": 0.32515621185302734,
      "learning_rate": 1.96749472718667e-05,
      "loss": 0.9168,
      "mean_token_accuracy": 0.8133991360664368,
      "num_tokens": 34981283.0,
      "step": 4560
    },
    {
      "entropy": 0.9301972724497318,
      "epoch": 0.6647272727272727,
      "grad_norm": 0.3155202269554138,
      "learning_rate": 1.9532635795708604e-05,
      "loss": 0.9121,
      "mean_token_accuracy": 0.8124547980725765,
      "num_tokens": 35058109.0,
      "step": 4570
    },
    {
      "entropy": 0.9356323592364788,
      "epoch": 0.6661818181818182,
      "grad_norm": 0.33614522218704224,
      "learning_rate": 1.939051280300148e-05,
      "loss": 0.9175,
      "mean_token_accuracy": 0.812773285806179,
      "num_tokens": 35135740.0,
      "step": 4580
    },
    {
      "entropy": 0.927642934024334,
      "epoch": 0.6676363636363636,
      "grad_norm": 0.31026774644851685,
      "learning_rate": 1.9248583148931846e-05,
      "loss": 0.9079,
      "mean_token_accuracy": 0.8145812973380089,
      "num_tokens": 35211772.0,
      "step": 4590
    },
    {
      "entropy": 0.9314297124743461,
      "epoch": 0.6690909090909091,
      "grad_norm": 0.32177138328552246,
      "learning_rate": 1.910685168208135e-05,
      "loss": 0.9151,
      "mean_token_accuracy": 0.8127957999706268,
      "num_tokens": 35291233.0,
      "step": 4600
    },
    {
      "entropy": 0.9267405033111572,
      "epoch": 0.6705454545454546,
      "grad_norm": 0.3147774636745453,
      "learning_rate": 1.896532324426125e-05,
      "loss": 0.9262,
      "mean_token_accuracy": 0.8127358593046665,
      "num_tokens": 35367300.0,
      "step": 4610
    },
    {
      "entropy": 0.9300662837922573,
      "epoch": 0.672,
      "grad_norm": 0.30565381050109863,
      "learning_rate": 1.8824002670346923e-05,
      "loss": 0.9164,
      "mean_token_accuracy": 0.8130706764757634,
      "num_tokens": 35445684.0,
      "step": 4620
    },
    {
      "entropy": 0.9152273058891296,
      "epoch": 0.6734545454545454,
      "grad_norm": 0.30101245641708374,
      "learning_rate": 1.868289478811272e-05,
      "loss": 0.9025,
      "mean_token_accuracy": 0.8156548984348774,
      "num_tokens": 35522237.0,
      "step": 4630
    },
    {
      "entropy": 0.9321424387395382,
      "epoch": 0.6749090909090909,
      "grad_norm": 0.3155101537704468,
      "learning_rate": 1.854200441806708e-05,
      "loss": 0.9206,
      "mean_token_accuracy": 0.81043586358428,
      "num_tokens": 35601986.0,
      "step": 4640
    },
    {
      "entropy": 0.9121891938149929,
      "epoch": 0.6763636363636364,
      "grad_norm": 0.3027799129486084,
      "learning_rate": 1.840133637328778e-05,
      "loss": 0.889,
      "mean_token_accuracy": 0.816604009270668,
      "num_tokens": 35678707.0,
      "step": 4650
    },
    {
      "entropy": 0.9230089358985424,
      "epoch": 0.6778181818181818,
      "grad_norm": 0.3036416172981262,
      "learning_rate": 1.8260895459257594e-05,
      "loss": 0.9062,
      "mean_token_accuracy": 0.8145593546330929,
      "num_tokens": 35752636.0,
      "step": 4660
    },
    {
      "entropy": 0.9228417217731476,
      "epoch": 0.6792727272727273,
      "grad_norm": 0.3162209391593933,
      "learning_rate": 1.812068647370005e-05,
      "loss": 0.9125,
      "mean_token_accuracy": 0.814147113263607,
      "num_tokens": 35828700.0,
      "step": 4670
    },
    {
      "entropy": 0.9207356870174408,
      "epoch": 0.6807272727272727,
      "grad_norm": 0.30621933937072754,
      "learning_rate": 1.7980714206415594e-05,
      "loss": 0.9039,
      "mean_token_accuracy": 0.8144191935658455,
      "num_tokens": 35905856.0,
      "step": 4680
    },
    {
      "entropy": 0.9312183283269405,
      "epoch": 0.6821818181818182,
      "grad_norm": 0.3484281599521637,
      "learning_rate": 1.7840983439117913e-05,
      "loss": 0.9061,
      "mean_token_accuracy": 0.8130019463598728,
      "num_tokens": 35984612.0,
      "step": 4690
    },
    {
      "entropy": 0.9221517466008663,
      "epoch": 0.6836363636363636,
      "grad_norm": 0.31173935532569885,
      "learning_rate": 1.7701498945270617e-05,
      "loss": 0.9086,
      "mean_token_accuracy": 0.8135214596986771,
      "num_tokens": 36062136.0,
      "step": 4700
    },
    {
      "entropy": 0.9280834481120109,
      "epoch": 0.6850909090909091,
      "grad_norm": 0.30636629462242126,
      "learning_rate": 1.7562265489924152e-05,
      "loss": 0.9117,
      "mean_token_accuracy": 0.8145467966794968,
      "num_tokens": 36138870.0,
      "step": 4710
    },
    {
      "entropy": 0.916886804997921,
      "epoch": 0.6865454545454546,
      "grad_norm": 0.2903667986392975,
      "learning_rate": 1.7423287829553007e-05,
      "loss": 0.9037,
      "mean_token_accuracy": 0.8137389704585075,
      "num_tokens": 36215545.0,
      "step": 4720
    },
    {
      "entropy": 0.9208665400743484,
      "epoch": 0.688,
      "grad_norm": 0.330870121717453,
      "learning_rate": 1.728457071189327e-05,
      "loss": 0.9113,
      "mean_token_accuracy": 0.8159119389951229,
      "num_tokens": 36289626.0,
      "step": 4730
    },
    {
      "entropy": 0.9310987628996372,
      "epoch": 0.6894545454545454,
      "grad_norm": 0.31531065702438354,
      "learning_rate": 1.7146118875780366e-05,
      "loss": 0.9106,
      "mean_token_accuracy": 0.8122837901115417,
      "num_tokens": 36368248.0,
      "step": 4740
    },
    {
      "entropy": 0.9374513991177083,
      "epoch": 0.6909090909090909,
      "grad_norm": 0.2993552088737488,
      "learning_rate": 1.7007937050987248e-05,
      "loss": 0.9147,
      "mean_token_accuracy": 0.8125071600079536,
      "num_tokens": 36444504.0,
      "step": 4750
    },
    {
      "entropy": 0.9231074757874012,
      "epoch": 0.6923636363636364,
      "grad_norm": 0.3486793637275696,
      "learning_rate": 1.6870029958062736e-05,
      "loss": 0.9172,
      "mean_token_accuracy": 0.8129670150578022,
      "num_tokens": 36520449.0,
      "step": 4760
    },
    {
      "entropy": 0.9267033264040947,
      "epoch": 0.6938181818181818,
      "grad_norm": 0.2960473597049713,
      "learning_rate": 1.6732402308170334e-05,
      "loss": 0.9095,
      "mean_token_accuracy": 0.8121006153523922,
      "num_tokens": 36596976.0,
      "step": 4770
    },
    {
      "entropy": 0.930424588918686,
      "epoch": 0.6952727272727273,
      "grad_norm": 0.3184972405433655,
      "learning_rate": 1.6595058802927218e-05,
      "loss": 0.9215,
      "mean_token_accuracy": 0.8138814494013786,
      "num_tokens": 36675808.0,
      "step": 4780
    },
    {
      "entropy": 0.9042963065207005,
      "epoch": 0.6967272727272728,
      "grad_norm": 0.3160466253757477,
      "learning_rate": 1.645800413424369e-05,
      "loss": 0.8805,
      "mean_token_accuracy": 0.8197195284068585,
      "num_tokens": 36749873.0,
      "step": 4790
    },
    {
      "entropy": 0.9051633194088936,
      "epoch": 0.6981818181818182,
      "grad_norm": 0.3440122902393341,
      "learning_rate": 1.6321242984162818e-05,
      "loss": 0.9007,
      "mean_token_accuracy": 0.8161387205123901,
      "num_tokens": 36827728.0,
      "step": 4800
    },
    {
      "entropy": 0.9146598286926746,
      "epoch": 0.6996363636363636,
      "grad_norm": 0.3145407736301422,
      "learning_rate": 1.6184780024700565e-05,
      "loss": 0.9016,
      "mean_token_accuracy": 0.814163988828659,
      "num_tokens": 36906330.0,
      "step": 4810
    },
    {
      "entropy": 0.9468378983438015,
      "epoch": 0.7010909090909091,
      "grad_norm": 0.2996998429298401,
      "learning_rate": 1.6048619917686113e-05,
      "loss": 0.9359,
      "mean_token_accuracy": 0.8094095788896084,
      "num_tokens": 36985047.0,
      "step": 4820
    },
    {
      "entropy": 0.9358403541147708,
      "epoch": 0.7025454545454546,
      "grad_norm": 0.3024042546749115,
      "learning_rate": 1.591276731460265e-05,
      "loss": 0.919,
      "mean_token_accuracy": 0.8127869322896004,
      "num_tokens": 37065350.0,
      "step": 4830
    },
    {
      "entropy": 0.9225737951695919,
      "epoch": 0.704,
      "grad_norm": 0.3492467403411865,
      "learning_rate": 1.5777226856428474e-05,
      "loss": 0.911,
      "mean_token_accuracy": 0.814387359470129,
      "num_tokens": 37143676.0,
      "step": 4840
    },
    {
      "entropy": 0.9143257580697537,
      "epoch": 0.7054545454545454,
      "grad_norm": 0.31178757548332214,
      "learning_rate": 1.5642003173478397e-05,
      "loss": 0.9005,
      "mean_token_accuracy": 0.8143624864518643,
      "num_tokens": 37220127.0,
      "step": 4850
    },
    {
      "entropy": 0.939749064296484,
      "epoch": 0.7069090909090909,
      "grad_norm": 0.32759299874305725,
      "learning_rate": 1.550710088524564e-05,
      "loss": 0.9352,
      "mean_token_accuracy": 0.8100519716739655,
      "num_tokens": 37299479.0,
      "step": 4860
    },
    {
      "entropy": 0.918349402397871,
      "epoch": 0.7083636363636364,
      "grad_norm": 0.32404211163520813,
      "learning_rate": 1.5372524600243933e-05,
      "loss": 0.9063,
      "mean_token_accuracy": 0.8145528778433799,
      "num_tokens": 37377354.0,
      "step": 4870
    },
    {
      "entropy": 0.9249159656465054,
      "epoch": 0.7098181818181818,
      "grad_norm": 0.306998074054718,
      "learning_rate": 1.5238278915850179e-05,
      "loss": 0.9132,
      "mean_token_accuracy": 0.8122608505189419,
      "num_tokens": 37453558.0,
      "step": 4880
    },
    {
      "entropy": 0.9320059776306152,
      "epoch": 0.7112727272727273,
      "grad_norm": 0.35612940788269043,
      "learning_rate": 1.5104368418147308e-05,
      "loss": 0.9192,
      "mean_token_accuracy": 0.8133574321866035,
      "num_tokens": 37529895.0,
      "step": 4890
    },
    {
      "entropy": 0.9335394233465195,
      "epoch": 0.7127272727272728,
      "grad_norm": 0.35255157947540283,
      "learning_rate": 1.497079768176767e-05,
      "loss": 0.9292,
      "mean_token_accuracy": 0.81116853505373,
      "num_tokens": 37608124.0,
      "step": 4900
    },
    {
      "entropy": 0.9473002113401889,
      "epoch": 0.7141818181818181,
      "grad_norm": 0.37226906418800354,
      "learning_rate": 1.4837571269736716e-05,
      "loss": 0.9384,
      "mean_token_accuracy": 0.8106353215873241,
      "num_tokens": 37685482.0,
      "step": 4910
    },
    {
      "entropy": 0.9388401240110398,
      "epoch": 0.7156363636363636,
      "grad_norm": 0.3105103671550751,
      "learning_rate": 1.4704693733317152e-05,
      "loss": 0.9147,
      "mean_token_accuracy": 0.8120019629597663,
      "num_tokens": 37763552.0,
      "step": 4920
    },
    {
      "entropy": 0.9257414855062962,
      "epoch": 0.7170909090909091,
      "grad_norm": 0.3366066515445709,
      "learning_rate": 1.4572169611853437e-05,
      "loss": 0.9009,
      "mean_token_accuracy": 0.8152667842805386,
      "num_tokens": 37840526.0,
      "step": 4930
    },
    {
      "entropy": 0.9142756037414074,
      "epoch": 0.7185454545454546,
      "grad_norm": 0.29677334427833557,
      "learning_rate": 1.4440003432616713e-05,
      "loss": 0.8997,
      "mean_token_accuracy": 0.81498968526721,
      "num_tokens": 37914807.0,
      "step": 4940
    },
    {
      "entropy": 0.9175783574581147,
      "epoch": 0.72,
      "grad_norm": 0.32036203145980835,
      "learning_rate": 1.4308199710650153e-05,
      "loss": 0.9059,
      "mean_token_accuracy": 0.8153199687600136,
      "num_tokens": 37990579.0,
      "step": 4950
    },
    {
      "entropy": 0.9218224361538887,
      "epoch": 0.7214545454545455,
      "grad_norm": 0.3067290782928467,
      "learning_rate": 1.4176762948614697e-05,
      "loss": 0.8976,
      "mean_token_accuracy": 0.8162059687077999,
      "num_tokens": 38065673.0,
      "step": 4960
    },
    {
      "entropy": 0.9500917561352253,
      "epoch": 0.722909090909091,
      "grad_norm": 0.3111895024776459,
      "learning_rate": 1.4045697636635283e-05,
      "loss": 0.9362,
      "mean_token_accuracy": 0.8087316453456879,
      "num_tokens": 38143782.0,
      "step": 4970
    },
    {
      "entropy": 0.9383980259299278,
      "epoch": 0.7243636363636363,
      "grad_norm": 0.3183054029941559,
      "learning_rate": 1.391500825214738e-05,
      "loss": 0.9191,
      "mean_token_accuracy": 0.8115313574671745,
      "num_tokens": 38223280.0,
      "step": 4980
    },
    {
      "entropy": 0.9261168956756591,
      "epoch": 0.7258181818181818,
      "grad_norm": 0.30299633741378784,
      "learning_rate": 1.378469925974412e-05,
      "loss": 0.9299,
      "mean_token_accuracy": 0.8118037767708302,
      "num_tokens": 38302455.0,
      "step": 4990
    },
    {
      "entropy": 0.910749813914299,
      "epoch": 0.7272727272727273,
      "grad_norm": 0.33091458678245544,
      "learning_rate": 1.3654775111023681e-05,
      "loss": 0.9022,
      "mean_token_accuracy": 0.8166116662323475,
      "num_tokens": 38380992.0,
      "step": 5000
    },
    {
      "epoch": 0.7272727272727273,
      "eval_entropy": 0.9231304630279541,
      "eval_loss": 0.8997952342033386,
      "eval_mean_token_accuracy": 0.811289277935028,
      "eval_num_tokens": 38380992.0,
      "eval_runtime": 77.91,
      "eval_samples_per_second": 64.177,
      "eval_steps_per_second": 8.022,
      "step": 5000
    },
    {
      "entropy": 0.9267726093530655,
      "epoch": 0.7287272727272728,
      "grad_norm": 0.31813424825668335,
      "learning_rate": 1.3525240244437315e-05,
      "loss": 0.9266,
      "mean_token_accuracy": 0.8130132332444191,
      "num_tokens": 38457879.0,
      "step": 5010
    },
    {
      "entropy": 0.9128262452781201,
      "epoch": 0.7301818181818182,
      "grad_norm": 0.38921618461608887,
      "learning_rate": 1.3396099085137629e-05,
      "loss": 0.8976,
      "mean_token_accuracy": 0.8168335042893886,
      "num_tokens": 38532295.0,
      "step": 5020
    },
    {
      "entropy": 0.9510353311896325,
      "epoch": 0.7316363636363636,
      "grad_norm": 0.3156704604625702,
      "learning_rate": 1.3267356044827464e-05,
      "loss": 0.9732,
      "mean_token_accuracy": 0.8092717811465263,
      "num_tokens": 38610836.0,
      "step": 5030
    },
    {
      "entropy": 0.9076557137072087,
      "epoch": 0.7330909090909091,
      "grad_norm": 0.3028537631034851,
      "learning_rate": 1.3139015521609203e-05,
      "loss": 0.8993,
      "mean_token_accuracy": 0.8164118349552154,
      "num_tokens": 38687664.0,
      "step": 5040
    },
    {
      "entropy": 0.9417129889130592,
      "epoch": 0.7345454545454545,
      "grad_norm": 0.30686843395233154,
      "learning_rate": 1.3011081899834443e-05,
      "loss": 0.9302,
      "mean_token_accuracy": 0.8112225525081158,
      "num_tokens": 38766894.0,
      "step": 5050
    },
    {
      "entropy": 0.9178734488785267,
      "epoch": 0.736,
      "grad_norm": 0.3113689124584198,
      "learning_rate": 1.2883559549954321e-05,
      "loss": 0.9046,
      "mean_token_accuracy": 0.8157093286514282,
      "num_tokens": 38844280.0,
      "step": 5060
    },
    {
      "entropy": 0.922054061293602,
      "epoch": 0.7374545454545455,
      "grad_norm": 0.3138102889060974,
      "learning_rate": 1.275645282837013e-05,
      "loss": 0.9071,
      "mean_token_accuracy": 0.814234309643507,
      "num_tokens": 38921783.0,
      "step": 5070
    },
    {
      "entropy": 0.9010880179703236,
      "epoch": 0.738909090909091,
      "grad_norm": 0.2947777211666107,
      "learning_rate": 1.2629766077284524e-05,
      "loss": 0.8823,
      "mean_token_accuracy": 0.8186369761824608,
      "num_tokens": 38998093.0,
      "step": 5080
    },
    {
      "entropy": 0.9084647111594677,
      "epoch": 0.7403636363636363,
      "grad_norm": 0.4752785861492157,
      "learning_rate": 1.250350362455318e-05,
      "loss": 0.8925,
      "mean_token_accuracy": 0.8168115198612214,
      "num_tokens": 39075179.0,
      "step": 5090
    },
    {
      "entropy": 0.9198925085365772,
      "epoch": 0.7418181818181818,
      "grad_norm": 0.309291809797287,
      "learning_rate": 1.2377669783536986e-05,
      "loss": 0.9024,
      "mean_token_accuracy": 0.814837496727705,
      "num_tokens": 39152927.0,
      "step": 5100
    },
    {
      "entropy": 0.9089493103325367,
      "epoch": 0.7432727272727273,
      "grad_norm": 0.3210631310939789,
      "learning_rate": 1.225226885295461e-05,
      "loss": 0.9023,
      "mean_token_accuracy": 0.8153677061200142,
      "num_tokens": 39228412.0,
      "step": 5110
    },
    {
      "entropy": 0.9285098657011985,
      "epoch": 0.7447272727272727,
      "grad_norm": 0.37961098551750183,
      "learning_rate": 1.2127305116735708e-05,
      "loss": 0.9171,
      "mean_token_accuracy": 0.8129767127335071,
      "num_tokens": 39305265.0,
      "step": 5120
    },
    {
      "entropy": 0.92706044241786,
      "epoch": 0.7461818181818182,
      "grad_norm": 0.2997186779975891,
      "learning_rate": 1.2002782843874574e-05,
      "loss": 0.9073,
      "mean_token_accuracy": 0.81295927464962,
      "num_tokens": 39382407.0,
      "step": 5130
    },
    {
      "entropy": 0.9154023215174675,
      "epoch": 0.7476363636363637,
      "grad_norm": 0.31855660676956177,
      "learning_rate": 1.1878706288284272e-05,
      "loss": 0.8967,
      "mean_token_accuracy": 0.817341311275959,
      "num_tokens": 39458813.0,
      "step": 5140
    },
    {
      "entropy": 0.9350193403661251,
      "epoch": 0.7490909090909091,
      "grad_norm": 0.31996530294418335,
      "learning_rate": 1.1755079688651353e-05,
      "loss": 0.926,
      "mean_token_accuracy": 0.8099291563034058,
      "num_tokens": 39537349.0,
      "step": 5150
    },
    {
      "entropy": 0.9326572187244893,
      "epoch": 0.7505454545454545,
      "grad_norm": 0.30809709429740906,
      "learning_rate": 1.1631907268291008e-05,
      "loss": 0.9197,
      "mean_token_accuracy": 0.8114828899502754,
      "num_tokens": 39615186.0,
      "step": 5160
    },
    {
      "entropy": 0.9393107175827027,
      "epoch": 0.752,
      "grad_norm": 0.3391665518283844,
      "learning_rate": 1.1509193235002843e-05,
      "loss": 0.927,
      "mean_token_accuracy": 0.8121741153299809,
      "num_tokens": 39694700.0,
      "step": 5170
    },
    {
      "entropy": 0.948701660335064,
      "epoch": 0.7534545454545455,
      "grad_norm": 0.3133850395679474,
      "learning_rate": 1.1386941780927079e-05,
      "loss": 0.9397,
      "mean_token_accuracy": 0.8088832281529903,
      "num_tokens": 39771680.0,
      "step": 5180
    },
    {
      "entropy": 0.9011263348162174,
      "epoch": 0.7549090909090909,
      "grad_norm": 0.3303625285625458,
      "learning_rate": 1.1265157082401404e-05,
      "loss": 0.8786,
      "mean_token_accuracy": 0.8216502912342548,
      "num_tokens": 39843328.0,
      "step": 5190
    },
    {
      "entropy": 0.9218214929103852,
      "epoch": 0.7563636363636363,
      "grad_norm": 0.3165566921234131,
      "learning_rate": 1.114384329981823e-05,
      "loss": 0.9143,
      "mean_token_accuracy": 0.8160425268113614,
      "num_tokens": 39920612.0,
      "step": 5200
    },
    {
      "entropy": 0.9363315224647522,
      "epoch": 0.7578181818181818,
      "grad_norm": 0.3346816897392273,
      "learning_rate": 1.1023004577482625e-05,
      "loss": 0.9155,
      "mean_token_accuracy": 0.8121711872518063,
      "num_tokens": 39995562.0,
      "step": 5210
    },
    {
      "entropy": 0.9151855200529099,
      "epoch": 0.7592727272727273,
      "grad_norm": 0.3320513069629669,
      "learning_rate": 1.0902645043470701e-05,
      "loss": 0.9058,
      "mean_token_accuracy": 0.8149781577289105,
      "num_tokens": 40071662.0,
      "step": 5220
    },
    {
      "entropy": 0.9302772752940655,
      "epoch": 0.7607272727272727,
      "grad_norm": 0.31996816396713257,
      "learning_rate": 1.0782768809488602e-05,
      "loss": 0.9047,
      "mean_token_accuracy": 0.8155157797038555,
      "num_tokens": 40148712.0,
      "step": 5230
    },
    {
      "entropy": 0.9082226067781448,
      "epoch": 0.7621818181818182,
      "grad_norm": 0.3083097040653229,
      "learning_rate": 1.0663379970732061e-05,
      "loss": 0.9028,
      "mean_token_accuracy": 0.8169299229979515,
      "num_tokens": 40223661.0,
      "step": 5240
    },
    {
      "entropy": 0.9109638199210167,
      "epoch": 0.7636363636363637,
      "grad_norm": 0.3519935607910156,
      "learning_rate": 1.0544482605746455e-05,
      "loss": 0.8945,
      "mean_token_accuracy": 0.818182872980833,
      "num_tokens": 40299756.0,
      "step": 5250
    },
    {
      "entropy": 0.9138317100703717,
      "epoch": 0.765090909090909,
      "grad_norm": 0.322620153427124,
      "learning_rate": 1.0426080776287528e-05,
      "loss": 0.8896,
      "mean_token_accuracy": 0.816927520930767,
      "num_tokens": 40373317.0,
      "step": 5260
    },
    {
      "entropy": 0.9338168516755104,
      "epoch": 0.7665454545454545,
      "grad_norm": 0.3295387029647827,
      "learning_rate": 1.0308178527182572e-05,
      "loss": 0.9173,
      "mean_token_accuracy": 0.812485358864069,
      "num_tokens": 40452850.0,
      "step": 5270
    },
    {
      "entropy": 0.9220724381506443,
      "epoch": 0.768,
      "grad_norm": 0.30471867322921753,
      "learning_rate": 1.019077988619232e-05,
      "loss": 0.9134,
      "mean_token_accuracy": 0.8132412627339363,
      "num_tokens": 40528672.0,
      "step": 5280
    },
    {
      "entropy": 0.9006679959595203,
      "epoch": 0.7694545454545455,
      "grad_norm": 0.3501948118209839,
      "learning_rate": 1.0073888863873281e-05,
      "loss": 0.884,
      "mean_token_accuracy": 0.8187536239624024,
      "num_tokens": 40603405.0,
      "step": 5290
    },
    {
      "entropy": 0.9375061370432377,
      "epoch": 0.7709090909090909,
      "grad_norm": 0.4156853258609772,
      "learning_rate": 9.957509453440788e-06,
      "loss": 0.9187,
      "mean_token_accuracy": 0.8133987367153168,
      "num_tokens": 40680296.0,
      "step": 5300
    },
    {
      "entropy": 0.9344054982066154,
      "epoch": 0.7723636363636364,
      "grad_norm": 0.3439910113811493,
      "learning_rate": 9.84164563063254e-06,
      "loss": 0.9292,
      "mean_token_accuracy": 0.8120204225182533,
      "num_tokens": 40759619.0,
      "step": 5310
    },
    {
      "entropy": 0.949238508939743,
      "epoch": 0.7738181818181818,
      "grad_norm": 0.3470531105995178,
      "learning_rate": 9.7263013535728e-06,
      "loss": 0.9667,
      "mean_token_accuracy": 0.8093096021562814,
      "num_tokens": 40835342.0,
      "step": 5320
    },
    {
      "entropy": 0.9372798420488835,
      "epoch": 0.7752727272727272,
      "grad_norm": 0.37672653794288635,
      "learning_rate": 9.611480562637207e-06,
      "loss": 0.9112,
      "mean_token_accuracy": 0.8122121192514896,
      "num_tokens": 40913392.0,
      "step": 5330
    },
    {
      "entropy": 0.9161439433693885,
      "epoch": 0.7767272727272727,
      "grad_norm": 0.3224441707134247,
      "learning_rate": 9.4971871803181e-06,
      "loss": 0.8972,
      "mean_token_accuracy": 0.8151120066642761,
      "num_tokens": 40988136.0,
      "step": 5340
    },
    {
      "entropy": 0.9237129762768745,
      "epoch": 0.7781818181818182,
      "grad_norm": 0.3345123529434204,
      "learning_rate": 9.383425111090586e-06,
      "loss": 0.923,
      "mean_token_accuracy": 0.8145867928862571,
      "num_tokens": 41062502.0,
      "step": 5350
    },
    {
      "entropy": 0.9012173816561699,
      "epoch": 0.7796363636363637,
      "grad_norm": 0.32990553975105286,
      "learning_rate": 9.270198241279093e-06,
      "loss": 0.8865,
      "mean_token_accuracy": 0.8177733510732651,
      "num_tokens": 41136820.0,
      "step": 5360
    },
    {
      "entropy": 0.9296848550438881,
      "epoch": 0.7810909090909091,
      "grad_norm": 0.32448163628578186,
      "learning_rate": 9.157510438924686e-06,
      "loss": 0.9188,
      "mean_token_accuracy": 0.8128324948251248,
      "num_tokens": 41212913.0,
      "step": 5370
    },
    {
      "entropy": 0.9221142314374446,
      "epoch": 0.7825454545454545,
      "grad_norm": 0.33526450395584106,
      "learning_rate": 9.045365553652833e-06,
      "loss": 0.901,
      "mean_token_accuracy": 0.8145244039595128,
      "num_tokens": 41289724.0,
      "step": 5380
    },
    {
      "entropy": 0.9362643413245678,
      "epoch": 0.784,
      "grad_norm": 0.3785208761692047,
      "learning_rate": 8.933767416541978e-06,
      "loss": 0.9243,
      "mean_token_accuracy": 0.8105296589434147,
      "num_tokens": 41370085.0,
      "step": 5390
    },
    {
      "entropy": 0.9407142043113709,
      "epoch": 0.7854545454545454,
      "grad_norm": 0.34329891204833984,
      "learning_rate": 8.822719839992597e-06,
      "loss": 0.9366,
      "mean_token_accuracy": 0.8094825424253941,
      "num_tokens": 41447850.0,
      "step": 5400
    },
    {
      "entropy": 0.924130903929472,
      "epoch": 0.7869090909090909,
      "grad_norm": 0.32263872027397156,
      "learning_rate": 8.712226617597028e-06,
      "loss": 0.9176,
      "mean_token_accuracy": 0.8136730268597603,
      "num_tokens": 41524404.0,
      "step": 5410
    },
    {
      "entropy": 0.9223329506814479,
      "epoch": 0.7883636363636364,
      "grad_norm": 0.3371281325817108,
      "learning_rate": 8.6022915240098e-06,
      "loss": 0.9104,
      "mean_token_accuracy": 0.8147990137338639,
      "num_tokens": 41597942.0,
      "step": 5420
    },
    {
      "entropy": 0.924953754991293,
      "epoch": 0.7898181818181819,
      "grad_norm": 0.3209916651248932,
      "learning_rate": 8.49291831481873e-06,
      "loss": 0.9151,
      "mean_token_accuracy": 0.8145290650427341,
      "num_tokens": 41675038.0,
      "step": 5430
    },
    {
      "entropy": 0.907654071599245,
      "epoch": 0.7912727272727272,
      "grad_norm": 0.3175525963306427,
      "learning_rate": 8.384110726416642e-06,
      "loss": 0.8971,
      "mean_token_accuracy": 0.8170280657708645,
      "num_tokens": 41750100.0,
      "step": 5440
    },
    {
      "entropy": 0.9338749371469021,
      "epoch": 0.7927272727272727,
      "grad_norm": 0.309807151556015,
      "learning_rate": 8.275872475873662e-06,
      "loss": 0.9112,
      "mean_token_accuracy": 0.8137181624770164,
      "num_tokens": 41828144.0,
      "step": 5450
    },
    {
      "entropy": 0.9331824757158756,
      "epoch": 0.7941818181818182,
      "grad_norm": 0.6010003089904785,
      "learning_rate": 8.16820726081031e-06,
      "loss": 0.9261,
      "mean_token_accuracy": 0.813402933627367,
      "num_tokens": 41905473.0,
      "step": 5460
    },
    {
      "entropy": 0.9054892651736737,
      "epoch": 0.7956363636363636,
      "grad_norm": 0.3187941014766693,
      "learning_rate": 8.06111875927111e-06,
      "loss": 0.8861,
      "mean_token_accuracy": 0.8182264745235444,
      "num_tokens": 41981650.0,
      "step": 5470
    },
    {
      "entropy": 0.9242243312299252,
      "epoch": 0.7970909090909091,
      "grad_norm": 0.3306691348552704,
      "learning_rate": 7.954610629598996e-06,
      "loss": 0.9252,
      "mean_token_accuracy": 0.8116098716855049,
      "num_tokens": 42059323.0,
      "step": 5480
    },
    {
      "entropy": 0.9288264907896518,
      "epoch": 0.7985454545454546,
      "grad_norm": 0.3101803660392761,
      "learning_rate": 7.848686510310307e-06,
      "loss": 0.9264,
      "mean_token_accuracy": 0.8119940876960754,
      "num_tokens": 42135684.0,
      "step": 5490
    },
    {
      "entropy": 0.9204266957938672,
      "epoch": 0.8,
      "grad_norm": 0.32018566131591797,
      "learning_rate": 7.74335001997052e-06,
      "loss": 0.9086,
      "mean_token_accuracy": 0.8140264257788659,
      "num_tokens": 42212157.0,
      "step": 5500
    },
    {
      "entropy": 0.9341423124074936,
      "epoch": 0.8014545454545454,
      "grad_norm": 0.30274590849876404,
      "learning_rate": 7.638604757070585e-06,
      "loss": 0.9178,
      "mean_token_accuracy": 0.8140484564006328,
      "num_tokens": 42289342.0,
      "step": 5510
    },
    {
      "entropy": 0.9244139701128006,
      "epoch": 0.8029090909090909,
      "grad_norm": 0.2957451343536377,
      "learning_rate": 7.5344542999040255e-06,
      "loss": 0.9046,
      "mean_token_accuracy": 0.8142490774393082,
      "num_tokens": 42366646.0,
      "step": 5520
    },
    {
      "entropy": 0.9261615209281444,
      "epoch": 0.8043636363636364,
      "grad_norm": 0.3075684905052185,
      "learning_rate": 7.430902206444711e-06,
      "loss": 0.9171,
      "mean_token_accuracy": 0.8156520687043667,
      "num_tokens": 42442815.0,
      "step": 5530
    },
    {
      "entropy": 0.9371460095047951,
      "epoch": 0.8058181818181818,
      "grad_norm": 0.3488467037677765,
      "learning_rate": 7.327952014225259e-06,
      "loss": 0.9259,
      "mean_token_accuracy": 0.8112338967621326,
      "num_tokens": 42522857.0,
      "step": 5540
    },
    {
      "entropy": 0.9135711878538132,
      "epoch": 0.8072727272727273,
      "grad_norm": 0.3251098692417145,
      "learning_rate": 7.225607240216253e-06,
      "loss": 0.9053,
      "mean_token_accuracy": 0.8151080518960953,
      "num_tokens": 42598873.0,
      "step": 5550
    },
    {
      "entropy": 0.9226764656603337,
      "epoch": 0.8087272727272727,
      "grad_norm": 0.3301449120044708,
      "learning_rate": 7.123871380706026e-06,
      "loss": 0.9059,
      "mean_token_accuracy": 0.8157884009182453,
      "num_tokens": 42676099.0,
      "step": 5560
    },
    {
      "entropy": 0.9240290597081184,
      "epoch": 0.8101818181818182,
      "grad_norm": 0.33067309856414795,
      "learning_rate": 7.022747911181295e-06,
      "loss": 0.9127,
      "mean_token_accuracy": 0.8157339677214622,
      "num_tokens": 42752810.0,
      "step": 5570
    },
    {
      "entropy": 0.9300024129450322,
      "epoch": 0.8116363636363636,
      "grad_norm": 0.3276396095752716,
      "learning_rate": 6.9222402862083525e-06,
      "loss": 0.9297,
      "mean_token_accuracy": 0.8102625668048858,
      "num_tokens": 42830186.0,
      "step": 5580
    },
    {
      "entropy": 0.9295799404382705,
      "epoch": 0.8130909090909091,
      "grad_norm": 0.3359651267528534,
      "learning_rate": 6.822351939315129e-06,
      "loss": 0.9165,
      "mean_token_accuracy": 0.813194601982832,
      "num_tokens": 42905771.0,
      "step": 5590
    },
    {
      "entropy": 0.9260146826505661,
      "epoch": 0.8145454545454546,
      "grad_norm": 0.3210604190826416,
      "learning_rate": 6.723086282873835e-06,
      "loss": 0.9019,
      "mean_token_accuracy": 0.8183946497738361,
      "num_tokens": 42981816.0,
      "step": 5600
    },
    {
      "entropy": 0.9256181947886943,
      "epoch": 0.816,
      "grad_norm": 0.3306752145290375,
      "learning_rate": 6.624446707984416e-06,
      "loss": 0.9133,
      "mean_token_accuracy": 0.8128662265837192,
      "num_tokens": 43059336.0,
      "step": 5610
    },
    {
      "entropy": 0.9192026577889919,
      "epoch": 0.8174545454545454,
      "grad_norm": 0.35311999917030334,
      "learning_rate": 6.5264365843587265e-06,
      "loss": 0.9047,
      "mean_token_accuracy": 0.8151163265109063,
      "num_tokens": 43136391.0,
      "step": 5620
    },
    {
      "entropy": 0.9239639192819595,
      "epoch": 0.8189090909090909,
      "grad_norm": 0.3081344962120056,
      "learning_rate": 6.4290592602053665e-06,
      "loss": 0.9135,
      "mean_token_accuracy": 0.8140504479408264,
      "num_tokens": 43214113.0,
      "step": 5630
    },
    {
      "entropy": 0.9212824292480946,
      "epoch": 0.8203636363636364,
      "grad_norm": 0.3152661919593811,
      "learning_rate": 6.332318062115341e-06,
      "loss": 0.9101,
      "mean_token_accuracy": 0.8138195432722568,
      "num_tokens": 43291551.0,
      "step": 5640
    },
    {
      "entropy": 0.9230553589761257,
      "epoch": 0.8218181818181818,
      "grad_norm": 0.31583520770072937,
      "learning_rate": 6.236216294948384e-06,
      "loss": 0.8995,
      "mean_token_accuracy": 0.8151654459536075,
      "num_tokens": 43369496.0,
      "step": 5650
    },
    {
      "entropy": 0.901103076338768,
      "epoch": 0.8232727272727273,
      "grad_norm": 0.33287662267684937,
      "learning_rate": 6.14075724172011e-06,
      "loss": 0.8962,
      "mean_token_accuracy": 0.8168199427425862,
      "num_tokens": 43443535.0,
      "step": 5660
    },
    {
      "entropy": 0.9405964225530624,
      "epoch": 0.8247272727272728,
      "grad_norm": 0.40133121609687805,
      "learning_rate": 6.045944163489785e-06,
      "loss": 0.9241,
      "mean_token_accuracy": 0.8112705871462822,
      "num_tokens": 43522587.0,
      "step": 5670
    },
    {
      "entropy": 0.9063078582286834,
      "epoch": 0.8261818181818181,
      "grad_norm": 0.34281712770462036,
      "learning_rate": 5.951780299249006e-06,
      "loss": 0.8985,
      "mean_token_accuracy": 0.8165899373590946,
      "num_tokens": 43597778.0,
      "step": 5680
    },
    {
      "entropy": 0.9149762600660324,
      "epoch": 0.8276363636363636,
      "grad_norm": 0.3659612536430359,
      "learning_rate": 5.85826886581097e-06,
      "loss": 0.9258,
      "mean_token_accuracy": 0.8148726306855678,
      "num_tokens": 43673314.0,
      "step": 5690
    },
    {
      "entropy": 0.9137645930051803,
      "epoch": 0.8290909090909091,
      "grad_norm": 0.36775919795036316,
      "learning_rate": 5.765413057700659e-06,
      "loss": 0.9152,
      "mean_token_accuracy": 0.8137751542031765,
      "num_tokens": 43752330.0,
      "step": 5700
    },
    {
      "entropy": 0.9262564107775688,
      "epoch": 0.8305454545454546,
      "grad_norm": 0.3278714418411255,
      "learning_rate": 5.673216047045637e-06,
      "loss": 0.9042,
      "mean_token_accuracy": 0.8133993491530418,
      "num_tokens": 43827542.0,
      "step": 5710
    },
    {
      "entropy": 0.9020561717450619,
      "epoch": 0.832,
      "grad_norm": 0.3041229844093323,
      "learning_rate": 5.581680983467746e-06,
      "loss": 0.8723,
      "mean_token_accuracy": 0.8194869726896286,
      "num_tokens": 43903331.0,
      "step": 5720
    },
    {
      "entropy": 0.9359213314950466,
      "epoch": 0.8334545454545454,
      "grad_norm": 0.33220741152763367,
      "learning_rate": 5.4908109939754605e-06,
      "loss": 0.9153,
      "mean_token_accuracy": 0.8136885397136211,
      "num_tokens": 43980944.0,
      "step": 5730
    },
    {
      "entropy": 0.9343759462237358,
      "epoch": 0.8349090909090909,
      "grad_norm": 0.31812480092048645,
      "learning_rate": 5.400609182857086e-06,
      "loss": 0.9269,
      "mean_token_accuracy": 0.8123717658221722,
      "num_tokens": 44059217.0,
      "step": 5740
    },
    {
      "entropy": 0.9189955502748489,
      "epoch": 0.8363636363636363,
      "grad_norm": 0.30972015857696533,
      "learning_rate": 5.31107863157473e-06,
      "loss": 0.9087,
      "mean_token_accuracy": 0.8125043779611587,
      "num_tokens": 44137764.0,
      "step": 5750
    },
    {
      "entropy": 0.9185071147978305,
      "epoch": 0.8378181818181818,
      "grad_norm": 0.30130535364151,
      "learning_rate": 5.2222223986589915e-06,
      "loss": 0.8974,
      "mean_token_accuracy": 0.815995067358017,
      "num_tokens": 44215864.0,
      "step": 5760
    },
    {
      "entropy": 0.9070168405771255,
      "epoch": 0.8392727272727273,
      "grad_norm": 0.33926618099212646,
      "learning_rate": 5.134043519604497e-06,
      "loss": 0.8987,
      "mean_token_accuracy": 0.8166046671569347,
      "num_tokens": 44291835.0,
      "step": 5770
    },
    {
      "entropy": 0.9173299342393875,
      "epoch": 0.8407272727272728,
      "grad_norm": 0.3847213089466095,
      "learning_rate": 5.046545006766206e-06,
      "loss": 0.9077,
      "mean_token_accuracy": 0.8146868884563446,
      "num_tokens": 44368202.0,
      "step": 5780
    },
    {
      "entropy": 0.9102527864277363,
      "epoch": 0.8421818181818181,
      "grad_norm": 0.36913734674453735,
      "learning_rate": 4.9597298492565064e-06,
      "loss": 0.9026,
      "mean_token_accuracy": 0.8161390721797943,
      "num_tokens": 44444116.0,
      "step": 5790
    },
    {
      "entropy": 0.9258656144142151,
      "epoch": 0.8436363636363636,
      "grad_norm": 0.31458625197410583,
      "learning_rate": 4.8736010128430825e-06,
      "loss": 0.9195,
      "mean_token_accuracy": 0.8141199484467506,
      "num_tokens": 44521118.0,
      "step": 5800
    },
    {
      "entropy": 0.933140229433775,
      "epoch": 0.8450909090909091,
      "grad_norm": 0.32843664288520813,
      "learning_rate": 4.788161439847618e-06,
      "loss": 0.9172,
      "mean_token_accuracy": 0.8115513384342193,
      "num_tokens": 44600130.0,
      "step": 5810
    },
    {
      "entropy": 0.9227162197232246,
      "epoch": 0.8465454545454546,
      "grad_norm": 0.3127104640007019,
      "learning_rate": 4.703414049045271e-06,
      "loss": 0.9023,
      "mean_token_accuracy": 0.8162086643278599,
      "num_tokens": 44677422.0,
      "step": 5820
    },
    {
      "entropy": 0.9307225309312344,
      "epoch": 0.848,
      "grad_norm": 0.3314415514469147,
      "learning_rate": 4.619361735564954e-06,
      "loss": 0.9228,
      "mean_token_accuracy": 0.8109805166721344,
      "num_tokens": 44756418.0,
      "step": 5830
    },
    {
      "entropy": 0.9122442126274108,
      "epoch": 0.8494545454545455,
      "grad_norm": 0.33934876322746277,
      "learning_rate": 4.536007370790463e-06,
      "loss": 0.8962,
      "mean_token_accuracy": 0.815778373926878,
      "num_tokens": 44833645.0,
      "step": 5840
    },
    {
      "entropy": 0.9105288550257683,
      "epoch": 0.850909090909091,
      "grad_norm": 0.31422221660614014,
      "learning_rate": 4.453353802262341e-06,
      "loss": 0.9018,
      "mean_token_accuracy": 0.817198246717453,
      "num_tokens": 44910482.0,
      "step": 5850
    },
    {
      "entropy": 0.8946401551365852,
      "epoch": 0.8523636363636363,
      "grad_norm": 0.3195045292377472,
      "learning_rate": 4.371403853580646e-06,
      "loss": 0.8744,
      "mean_token_accuracy": 0.820909558236599,
      "num_tokens": 44983955.0,
      "step": 5860
    },
    {
      "entropy": 0.8876437328755855,
      "epoch": 0.8538181818181818,
      "grad_norm": 0.39258408546447754,
      "learning_rate": 4.290160324308453e-06,
      "loss": 0.8768,
      "mean_token_accuracy": 0.8214518509805202,
      "num_tokens": 45056876.0,
      "step": 5870
    },
    {
      "entropy": 0.9040459990501404,
      "epoch": 0.8552727272727273,
      "grad_norm": 0.31693005561828613,
      "learning_rate": 4.2096259898762484e-06,
      "loss": 0.8934,
      "mean_token_accuracy": 0.816833870112896,
      "num_tokens": 45132382.0,
      "step": 5880
    },
    {
      "entropy": 0.9187272481620312,
      "epoch": 0.8567272727272728,
      "grad_norm": 0.3108614385128021,
      "learning_rate": 4.129803601487085e-06,
      "loss": 0.9093,
      "mean_token_accuracy": 0.8157303839921951,
      "num_tokens": 45207136.0,
      "step": 5890
    },
    {
      "entropy": 0.9228496827185154,
      "epoch": 0.8581818181818182,
      "grad_norm": 0.31431129574775696,
      "learning_rate": 4.05069588602262e-06,
      "loss": 0.9211,
      "mean_token_accuracy": 0.8127873703837395,
      "num_tokens": 45284545.0,
      "step": 5900
    },
    {
      "entropy": 0.9108579710125924,
      "epoch": 0.8596363636363636,
      "grad_norm": 0.35750770568847656,
      "learning_rate": 3.972305545949937e-06,
      "loss": 0.9018,
      "mean_token_accuracy": 0.8170554347336292,
      "num_tokens": 45360575.0,
      "step": 5910
    },
    {
      "entropy": 0.9355661682784557,
      "epoch": 0.8610909090909091,
      "grad_norm": 0.36121928691864014,
      "learning_rate": 3.894635259229265e-06,
      "loss": 0.9268,
      "mean_token_accuracy": 0.8105575621128083,
      "num_tokens": 45439579.0,
      "step": 5920
    },
    {
      "entropy": 0.961587718129158,
      "epoch": 0.8625454545454545,
      "grad_norm": 0.31216803193092346,
      "learning_rate": 3.817687679222442e-06,
      "loss": 0.9521,
      "mean_token_accuracy": 0.807161021232605,
      "num_tokens": 45522968.0,
      "step": 5930
    },
    {
      "entropy": 0.9307756863534451,
      "epoch": 0.864,
      "grad_norm": 0.34104517102241516,
      "learning_rate": 3.741465434602306e-06,
      "loss": 0.9165,
      "mean_token_accuracy": 0.814553914219141,
      "num_tokens": 45598073.0,
      "step": 5940
    },
    {
      "entropy": 0.9234868161380291,
      "epoch": 0.8654545454545455,
      "grad_norm": 0.34115663170814514,
      "learning_rate": 3.6659711292628956e-06,
      "loss": 0.9063,
      "mean_token_accuracy": 0.8158062770962715,
      "num_tokens": 45675032.0,
      "step": 5950
    },
    {
      "entropy": 0.9477932959794998,
      "epoch": 0.866909090909091,
      "grad_norm": 0.3255559504032135,
      "learning_rate": 3.5912073422304626e-06,
      "loss": 0.9264,
      "mean_token_accuracy": 0.8109208598732949,
      "num_tokens": 45753692.0,
      "step": 5960
    },
    {
      "entropy": 0.9064127787947655,
      "epoch": 0.8683636363636363,
      "grad_norm": 0.3285483419895172,
      "learning_rate": 3.5171766275754187e-06,
      "loss": 0.8953,
      "mean_token_accuracy": 0.8159914433956146,
      "num_tokens": 45831343.0,
      "step": 5970
    },
    {
      "entropy": 0.9151718072593212,
      "epoch": 0.8698181818181818,
      "grad_norm": 0.31288042664527893,
      "learning_rate": 3.4438815143250314e-06,
      "loss": 0.8918,
      "mean_token_accuracy": 0.8162871718406677,
      "num_tokens": 45907964.0,
      "step": 5980
    },
    {
      "entropy": 0.9252648554742336,
      "epoch": 0.8712727272727273,
      "grad_norm": 0.37754347920417786,
      "learning_rate": 3.3713245063770745e-06,
      "loss": 0.9441,
      "mean_token_accuracy": 0.8124652244150639,
      "num_tokens": 45984483.0,
      "step": 5990
    },
    {
      "entropy": 0.897750299423933,
      "epoch": 0.8727272727272727,
      "grad_norm": 0.33970126509666443,
      "learning_rate": 3.2995080824142425e-06,
      "loss": 0.8851,
      "mean_token_accuracy": 0.819002702832222,
      "num_tokens": 46061634.0,
      "step": 6000
    },
    {
      "epoch": 0.8727272727272727,
      "eval_entropy": 0.9209443353652954,
      "eval_loss": 0.8987749218940735,
      "eval_mean_token_accuracy": 0.8116296404838562,
      "eval_num_tokens": 46061634.0,
      "eval_runtime": 77.8817,
      "eval_samples_per_second": 64.2,
      "eval_steps_per_second": 8.025,
      "step": 6000
    },
    {
      "entropy": 0.9322944730520248,
      "epoch": 0.8741818181818182,
      "grad_norm": 0.3368423283100128,
      "learning_rate": 3.228434695819528e-06,
      "loss": 0.9194,
      "mean_token_accuracy": 0.8112942822277546,
      "num_tokens": 46139299.0,
      "step": 6010
    },
    {
      "entropy": 0.9085685111582279,
      "epoch": 0.8756363636363637,
      "grad_norm": 0.3252302408218384,
      "learning_rate": 3.15810677459236e-06,
      "loss": 0.8893,
      "mean_token_accuracy": 0.8170065075159073,
      "num_tokens": 46215455.0,
      "step": 6020
    },
    {
      "entropy": 0.9145600602030755,
      "epoch": 0.8770909090909091,
      "grad_norm": 0.3287157416343689,
      "learning_rate": 3.0885267212656892e-06,
      "loss": 0.9011,
      "mean_token_accuracy": 0.8160723358392715,
      "num_tokens": 46291495.0,
      "step": 6030
    },
    {
      "entropy": 0.9285273037850856,
      "epoch": 0.8785454545454545,
      "grad_norm": 0.31994038820266724,
      "learning_rate": 3.019696912823903e-06,
      "loss": 0.9203,
      "mean_token_accuracy": 0.8141358248889446,
      "num_tokens": 46370221.0,
      "step": 6040
    },
    {
      "entropy": 0.9345873408019543,
      "epoch": 0.88,
      "grad_norm": 0.37476661801338196,
      "learning_rate": 2.9516197006216194e-06,
      "loss": 0.9389,
      "mean_token_accuracy": 0.8111633822321892,
      "num_tokens": 46449245.0,
      "step": 6050
    },
    {
      "entropy": 0.9355661444365978,
      "epoch": 0.8814545454545455,
      "grad_norm": 0.3263081908226013,
      "learning_rate": 2.8842974103033773e-06,
      "loss": 0.9157,
      "mean_token_accuracy": 0.8130726262927055,
      "num_tokens": 46524841.0,
      "step": 6060
    },
    {
      "entropy": 0.9101524129509926,
      "epoch": 0.8829090909090909,
      "grad_norm": 0.3230786919593811,
      "learning_rate": 2.817732341724158e-06,
      "loss": 0.8942,
      "mean_token_accuracy": 0.8166044473648071,
      "num_tokens": 46598955.0,
      "step": 6070
    },
    {
      "entropy": 0.9097373388707638,
      "epoch": 0.8843636363636364,
      "grad_norm": 0.3022027611732483,
      "learning_rate": 2.7519267688708604e-06,
      "loss": 0.898,
      "mean_token_accuracy": 0.8170449204742909,
      "num_tokens": 46674399.0,
      "step": 6080
    },
    {
      "entropy": 0.9250694133341313,
      "epoch": 0.8858181818181818,
      "grad_norm": 0.34704554080963135,
      "learning_rate": 2.6868829397845656e-06,
      "loss": 0.9124,
      "mean_token_accuracy": 0.8141155697405338,
      "num_tokens": 46752826.0,
      "step": 6090
    },
    {
      "entropy": 0.9141334019601345,
      "epoch": 0.8872727272727273,
      "grad_norm": 0.338782399892807,
      "learning_rate": 2.6226030764837945e-06,
      "loss": 0.9261,
      "mean_token_accuracy": 0.8140919879078865,
      "num_tokens": 46828411.0,
      "step": 6100
    },
    {
      "entropy": 0.9200235098600388,
      "epoch": 0.8887272727272727,
      "grad_norm": 0.33723023533821106,
      "learning_rate": 2.559089374888545e-06,
      "loss": 0.9096,
      "mean_token_accuracy": 0.8156822443008422,
      "num_tokens": 46904972.0,
      "step": 6110
    },
    {
      "entropy": 0.9146118760108948,
      "epoch": 0.8901818181818182,
      "grad_norm": 0.31707334518432617,
      "learning_rate": 2.496344004745324e-06,
      "loss": 0.8998,
      "mean_token_accuracy": 0.816555169224739,
      "num_tokens": 46981629.0,
      "step": 6120
    },
    {
      "entropy": 0.9358232043683529,
      "epoch": 0.8916363636363637,
      "grad_norm": 0.31498560309410095,
      "learning_rate": 2.434369109552993e-06,
      "loss": 0.919,
      "mean_token_accuracy": 0.8128682546317577,
      "num_tokens": 47058559.0,
      "step": 6130
    },
    {
      "entropy": 0.9056357733905316,
      "epoch": 0.893090909090909,
      "grad_norm": 0.353083997964859,
      "learning_rate": 2.373166806489546e-06,
      "loss": 0.89,
      "mean_token_accuracy": 0.8196616470813751,
      "num_tokens": 47133139.0,
      "step": 6140
    },
    {
      "entropy": 0.9217743061482906,
      "epoch": 0.8945454545454545,
      "grad_norm": 0.3720227777957916,
      "learning_rate": 2.312739186339809e-06,
      "loss": 0.9191,
      "mean_token_accuracy": 0.8154807358980178,
      "num_tokens": 47208884.0,
      "step": 6150
    },
    {
      "entropy": 0.9359294772148132,
      "epoch": 0.896,
      "grad_norm": 0.3200484812259674,
      "learning_rate": 2.2530883134239742e-06,
      "loss": 0.9249,
      "mean_token_accuracy": 0.8138979524374008,
      "num_tokens": 47284079.0,
      "step": 6160
    },
    {
      "entropy": 0.9157271198928356,
      "epoch": 0.8974545454545455,
      "grad_norm": 0.30511274933815,
      "learning_rate": 2.1942162255271132e-06,
      "loss": 0.9046,
      "mean_token_accuracy": 0.8142931953072547,
      "num_tokens": 47359251.0,
      "step": 6170
    },
    {
      "entropy": 0.9222023263573647,
      "epoch": 0.8989090909090909,
      "grad_norm": 0.3218860924243927,
      "learning_rate": 2.1361249338295327e-06,
      "loss": 0.9076,
      "mean_token_accuracy": 0.8148991748690605,
      "num_tokens": 47435112.0,
      "step": 6180
    },
    {
      "entropy": 0.9072126895189285,
      "epoch": 0.9003636363636364,
      "grad_norm": 0.3221643567085266,
      "learning_rate": 2.078816422838109e-06,
      "loss": 0.8986,
      "mean_token_accuracy": 0.8161638289690017,
      "num_tokens": 47510845.0,
      "step": 6190
    },
    {
      "entropy": 0.9256663709878922,
      "epoch": 0.9018181818181819,
      "grad_norm": 0.311989426612854,
      "learning_rate": 2.022292650318451e-06,
      "loss": 0.9365,
      "mean_token_accuracy": 0.8134777262806893,
      "num_tokens": 47588720.0,
      "step": 6200
    },
    {
      "entropy": 0.9273410558700561,
      "epoch": 0.9032727272727272,
      "grad_norm": 0.31235963106155396,
      "learning_rate": 1.966555547228044e-06,
      "loss": 0.9101,
      "mean_token_accuracy": 0.8141037151217461,
      "num_tokens": 47666849.0,
      "step": 6210
    },
    {
      "entropy": 0.920647069811821,
      "epoch": 0.9047272727272727,
      "grad_norm": 0.3313010632991791,
      "learning_rate": 1.9116070176502807e-06,
      "loss": 0.9013,
      "mean_token_accuracy": 0.8138377077877521,
      "num_tokens": 47743639.0,
      "step": 6220
    },
    {
      "entropy": 0.9163322940468788,
      "epoch": 0.9061818181818182,
      "grad_norm": 0.31483882665634155,
      "learning_rate": 1.8574489387294034e-06,
      "loss": 0.907,
      "mean_token_accuracy": 0.8146905891597271,
      "num_tokens": 47821434.0,
      "step": 6230
    },
    {
      "entropy": 0.9224201649427414,
      "epoch": 0.9076363636363637,
      "grad_norm": 0.30767881870269775,
      "learning_rate": 1.8040831606063967e-06,
      "loss": 0.9118,
      "mean_token_accuracy": 0.8155392311513424,
      "num_tokens": 47899203.0,
      "step": 6240
    },
    {
      "entropy": 0.8979198761284352,
      "epoch": 0.9090909090909091,
      "grad_norm": 0.310153603553772,
      "learning_rate": 1.7515115063557652e-06,
      "loss": 0.8839,
      "mean_token_accuracy": 0.819090685993433,
      "num_tokens": 47972742.0,
      "step": 6250
    },
    {
      "entropy": 0.9100939020514488,
      "epoch": 0.9105454545454545,
      "grad_norm": 0.34032052755355835,
      "learning_rate": 1.6997357719232662e-06,
      "loss": 0.8999,
      "mean_token_accuracy": 0.8160591490566731,
      "num_tokens": 48049075.0,
      "step": 6260
    },
    {
      "entropy": 0.9030675068497658,
      "epoch": 0.912,
      "grad_norm": 0.32921066880226135,
      "learning_rate": 1.6487577260645411e-06,
      "loss": 0.8838,
      "mean_token_accuracy": 0.8187085382640362,
      "num_tokens": 48126883.0,
      "step": 6270
    },
    {
      "entropy": 0.9277480982244015,
      "epoch": 0.9134545454545454,
      "grad_norm": 0.3190605938434601,
      "learning_rate": 1.5985791102847185e-06,
      "loss": 0.9225,
      "mean_token_accuracy": 0.8141351498663425,
      "num_tokens": 48203971.0,
      "step": 6280
    },
    {
      "entropy": 0.9353076785802841,
      "epoch": 0.9149090909090909,
      "grad_norm": 0.303288996219635,
      "learning_rate": 1.5492016387788944e-06,
      "loss": 0.9444,
      "mean_token_accuracy": 0.8119538478553295,
      "num_tokens": 48282198.0,
      "step": 6290
    },
    {
      "entropy": 0.9261584036052227,
      "epoch": 0.9163636363636364,
      "grad_norm": 0.32700851559638977,
      "learning_rate": 1.5006269983735794e-06,
      "loss": 0.9153,
      "mean_token_accuracy": 0.812885682284832,
      "num_tokens": 48360686.0,
      "step": 6300
    },
    {
      "entropy": 0.9366669520735741,
      "epoch": 0.9178181818181819,
      "grad_norm": 0.33134371042251587,
      "learning_rate": 1.4528568484690765e-06,
      "loss": 0.9374,
      "mean_token_accuracy": 0.8110633857548237,
      "num_tokens": 48438891.0,
      "step": 6310
    },
    {
      "entropy": 0.9129975132644177,
      "epoch": 0.9192727272727272,
      "grad_norm": 0.3131563067436218,
      "learning_rate": 1.4058928209828066e-06,
      "loss": 0.9163,
      "mean_token_accuracy": 0.8145956374704838,
      "num_tokens": 48516417.0,
      "step": 6320
    },
    {
      "entropy": 0.9178128764033318,
      "epoch": 0.9207272727272727,
      "grad_norm": 0.3282724916934967,
      "learning_rate": 1.3597365202935283e-06,
      "loss": 0.8925,
      "mean_token_accuracy": 0.8164288327097893,
      "num_tokens": 48591332.0,
      "step": 6330
    },
    {
      "entropy": 0.9109235092997551,
      "epoch": 0.9221818181818182,
      "grad_norm": 0.30869513750076294,
      "learning_rate": 1.3143895231865565e-06,
      "loss": 0.8895,
      "mean_token_accuracy": 0.8177370443940163,
      "num_tokens": 48665701.0,
      "step": 6340
    },
    {
      "entropy": 0.9241031289100647,
      "epoch": 0.9236363636363636,
      "grad_norm": 0.36920541524887085,
      "learning_rate": 1.2698533787998822e-06,
      "loss": 0.9163,
      "mean_token_accuracy": 0.8146472811698914,
      "num_tokens": 48742926.0,
      "step": 6350
    },
    {
      "entropy": 0.9276870623230934,
      "epoch": 0.9250909090909091,
      "grad_norm": 0.3207862377166748,
      "learning_rate": 1.2261296085712583e-06,
      "loss": 0.9151,
      "mean_token_accuracy": 0.8119709461927413,
      "num_tokens": 48822553.0,
      "step": 6360
    },
    {
      "entropy": 0.925787340849638,
      "epoch": 0.9265454545454546,
      "grad_norm": 0.31634289026260376,
      "learning_rate": 1.1832197061862213e-06,
      "loss": 0.9126,
      "mean_token_accuracy": 0.8139682069420815,
      "num_tokens": 48901887.0,
      "step": 6370
    },
    {
      "entropy": 0.9347988933324813,
      "epoch": 0.928,
      "grad_norm": 0.3287469446659088,
      "learning_rate": 1.1411251375270554e-06,
      "loss": 0.9197,
      "mean_token_accuracy": 0.8132240571081638,
      "num_tokens": 48979966.0,
      "step": 6380
    },
    {
      "entropy": 0.9178888291120529,
      "epoch": 0.9294545454545454,
      "grad_norm": 0.33060967922210693,
      "learning_rate": 1.099847340622741e-06,
      "loss": 0.9089,
      "mean_token_accuracy": 0.8142723865807057,
      "num_tokens": 49057484.0,
      "step": 6390
    },
    {
      "entropy": 0.9371354281902313,
      "epoch": 0.9309090909090909,
      "grad_norm": 0.31378334760665894,
      "learning_rate": 1.0593877255997946e-06,
      "loss": 0.9311,
      "mean_token_accuracy": 0.812279848754406,
      "num_tokens": 49137700.0,
      "step": 6400
    },
    {
      "entropy": 0.917350922524929,
      "epoch": 0.9323636363636364,
      "grad_norm": 0.30465006828308105,
      "learning_rate": 1.0197476746341303e-06,
      "loss": 0.9027,
      "mean_token_accuracy": 0.8145864970982075,
      "num_tokens": 49214269.0,
      "step": 6410
    },
    {
      "entropy": 0.9150080628693104,
      "epoch": 0.9338181818181818,
      "grad_norm": 0.3422051966190338,
      "learning_rate": 9.8092854190381e-07,
      "loss": 0.9056,
      "mean_token_accuracy": 0.813194676488638,
      "num_tokens": 49290198.0,
      "step": 6420
    },
    {
      "entropy": 0.9078165777027607,
      "epoch": 0.9352727272727273,
      "grad_norm": 0.3076263964176178,
      "learning_rate": 9.429316535428057e-07,
      "loss": 0.8903,
      "mean_token_accuracy": 0.8160121709108352,
      "num_tokens": 49368361.0,
      "step": 6430
    },
    {
      "entropy": 0.9192823156714439,
      "epoch": 0.9367272727272727,
      "grad_norm": 0.32441800832748413,
      "learning_rate": 9.057583075956887e-07,
      "loss": 0.901,
      "mean_token_accuracy": 0.8145803168416024,
      "num_tokens": 49445497.0,
      "step": 6440
    },
    {
      "entropy": 0.9385962396860122,
      "epoch": 0.9381818181818182,
      "grad_norm": 0.34090206027030945,
      "learning_rate": 8.694097739732798e-07,
      "loss": 0.9291,
      "mean_token_accuracy": 0.8106141477823258,
      "num_tokens": 49523114.0,
      "step": 6450
    },
    {
      "entropy": 0.9091130934655667,
      "epoch": 0.9396363636363636,
      "grad_norm": 0.3067343831062317,
      "learning_rate": 8.338872944092852e-07,
      "loss": 0.8978,
      "mean_token_accuracy": 0.8167271681129933,
      "num_tokens": 49599066.0,
      "step": 6460
    },
    {
      "entropy": 0.923352313041687,
      "epoch": 0.9410909090909091,
      "grad_norm": 0.34588825702667236,
      "learning_rate": 7.991920824178455e-07,
      "loss": 0.908,
      "mean_token_accuracy": 0.812999901920557,
      "num_tokens": 49676403.0,
      "step": 6470
    },
    {
      "entropy": 0.9061731100082397,
      "epoch": 0.9425454545454546,
      "grad_norm": 0.3114166855812073,
      "learning_rate": 7.653253232521162e-07,
      "loss": 0.8886,
      "mean_token_accuracy": 0.8181905202567578,
      "num_tokens": 49748740.0,
      "step": 6480
    },
    {
      "entropy": 0.9442753471434117,
      "epoch": 0.944,
      "grad_norm": 0.3027520477771759,
      "learning_rate": 7.322881738637479e-07,
      "loss": 0.9392,
      "mean_token_accuracy": 0.8092067658901214,
      "num_tokens": 49828347.0,
      "step": 6490
    },
    {
      "entropy": 0.909515468031168,
      "epoch": 0.9454545454545454,
      "grad_norm": 0.31182175874710083,
      "learning_rate": 7.000817628633776e-07,
      "loss": 0.9053,
      "mean_token_accuracy": 0.8153221309185028,
      "num_tokens": 49905096.0,
      "step": 6500
    },
    {
      "entropy": 0.9085680104792118,
      "epoch": 0.9469090909090909,
      "grad_norm": 0.33979618549346924,
      "learning_rate": 6.687071904820677e-07,
      "loss": 0.8953,
      "mean_token_accuracy": 0.8190076671540737,
      "num_tokens": 49981386.0,
      "step": 6510
    },
    {
      "entropy": 0.9074861466884613,
      "epoch": 0.9483636363636364,
      "grad_norm": 0.3361256420612335,
      "learning_rate": 6.381655285337279e-07,
      "loss": 0.8951,
      "mean_token_accuracy": 0.816261338442564,
      "num_tokens": 50057245.0,
      "step": 6520
    },
    {
      "entropy": 0.9325119577348232,
      "epoch": 0.9498181818181818,
      "grad_norm": 0.36365851759910583,
      "learning_rate": 6.084578203784874e-07,
      "loss": 0.9214,
      "mean_token_accuracy": 0.8115631319582463,
      "num_tokens": 50133281.0,
      "step": 6530
    },
    {
      "entropy": 0.9262050181627274,
      "epoch": 0.9512727272727273,
      "grad_norm": 0.31893861293792725,
      "learning_rate": 5.795850808870589e-07,
      "loss": 0.9183,
      "mean_token_accuracy": 0.8127511434257031,
      "num_tokens": 50209580.0,
      "step": 6540
    },
    {
      "entropy": 0.9360543221235276,
      "epoch": 0.9527272727272728,
      "grad_norm": 0.3180423974990845,
      "learning_rate": 5.515482964060745e-07,
      "loss": 0.937,
      "mean_token_accuracy": 0.8129704773426056,
      "num_tokens": 50285252.0,
      "step": 6550
    },
    {
      "entropy": 0.9206071726977825,
      "epoch": 0.9541818181818181,
      "grad_norm": 0.3428179621696472,
      "learning_rate": 5.243484247243779e-07,
      "loss": 0.9134,
      "mean_token_accuracy": 0.8148298509418964,
      "num_tokens": 50362033.0,
      "step": 6560
    },
    {
      "entropy": 0.9003576777875424,
      "epoch": 0.9556363636363636,
      "grad_norm": 0.31086480617523193,
      "learning_rate": 4.979863950403129e-07,
      "loss": 0.8885,
      "mean_token_accuracy": 0.8173546336591244,
      "num_tokens": 50436962.0,
      "step": 6570
    },
    {
      "entropy": 0.9236943081021309,
      "epoch": 0.9570909090909091,
      "grad_norm": 0.31479397416114807,
      "learning_rate": 4.7246310792997683e-07,
      "loss": 0.9093,
      "mean_token_accuracy": 0.8160760276019573,
      "num_tokens": 50513406.0,
      "step": 6580
    },
    {
      "entropy": 0.934125005453825,
      "epoch": 0.9585454545454546,
      "grad_norm": 0.3748307526111603,
      "learning_rate": 4.4777943531646196e-07,
      "loss": 0.9291,
      "mean_token_accuracy": 0.8114527925848961,
      "num_tokens": 50591962.0,
      "step": 6590
    },
    {
      "entropy": 0.9182382963597775,
      "epoch": 0.96,
      "grad_norm": 0.3080226480960846,
      "learning_rate": 4.2393622044005333e-07,
      "loss": 0.9065,
      "mean_token_accuracy": 0.8157414674758912,
      "num_tokens": 50668724.0,
      "step": 6600
    },
    {
      "entropy": 0.9102472327649593,
      "epoch": 0.9614545454545455,
      "grad_norm": 0.3206920325756073,
      "learning_rate": 4.0093427782944265e-07,
      "loss": 0.9043,
      "mean_token_accuracy": 0.8165704101324082,
      "num_tokens": 50744450.0,
      "step": 6610
    },
    {
      "entropy": 0.927041707187891,
      "epoch": 0.9629090909090909,
      "grad_norm": 0.3061367869377136,
      "learning_rate": 3.787743932738898e-07,
      "loss": 0.906,
      "mean_token_accuracy": 0.8148591019213199,
      "num_tokens": 50821886.0,
      "step": 6620
    },
    {
      "entropy": 0.9167927749454975,
      "epoch": 0.9643636363636363,
      "grad_norm": 0.3160494565963745,
      "learning_rate": 3.5745732379637053e-07,
      "loss": 0.8901,
      "mean_token_accuracy": 0.8175670191645622,
      "num_tokens": 50897361.0,
      "step": 6630
    },
    {
      "entropy": 0.91978540122509,
      "epoch": 0.9658181818181818,
      "grad_norm": 0.31838276982307434,
      "learning_rate": 3.369837976277431e-07,
      "loss": 0.9123,
      "mean_token_accuracy": 0.8131505504250527,
      "num_tokens": 50976682.0,
      "step": 6640
    },
    {
      "entropy": 0.9417281582951545,
      "epoch": 0.9672727272727273,
      "grad_norm": 0.3298901915550232,
      "learning_rate": 3.173545141818379e-07,
      "loss": 0.9281,
      "mean_token_accuracy": 0.8116386346518993,
      "num_tokens": 51053043.0,
      "step": 6650
    },
    {
      "entropy": 0.9071712777018547,
      "epoch": 0.9687272727272728,
      "grad_norm": 0.31101104617118835,
      "learning_rate": 2.9857014403158787e-07,
      "loss": 0.9025,
      "mean_token_accuracy": 0.816408134251833,
      "num_tokens": 51129073.0,
      "step": 6660
    },
    {
      "entropy": 0.9059986241161824,
      "epoch": 0.9701818181818181,
      "grad_norm": 0.30451127886772156,
      "learning_rate": 2.8063132888611017e-07,
      "loss": 0.8866,
      "mean_token_accuracy": 0.8188144057989121,
      "num_tokens": 51203044.0,
      "step": 6670
    },
    {
      "entropy": 0.9010291390120984,
      "epoch": 0.9716363636363636,
      "grad_norm": 0.31770607829093933,
      "learning_rate": 2.6353868156878635e-07,
      "loss": 0.8896,
      "mean_token_accuracy": 0.8195660427212715,
      "num_tokens": 51277927.0,
      "step": 6680
    },
    {
      "entropy": 0.9211806155741215,
      "epoch": 0.9730909090909091,
      "grad_norm": 0.30818912386894226,
      "learning_rate": 2.472927859963272e-07,
      "loss": 0.9006,
      "mean_token_accuracy": 0.8144206136465073,
      "num_tokens": 51355111.0,
      "step": 6690
    },
    {
      "entropy": 0.9047483175992965,
      "epoch": 0.9745454545454545,
      "grad_norm": 0.3617713153362274,
      "learning_rate": 2.3189419715881735e-07,
      "loss": 0.8907,
      "mean_token_accuracy": 0.8167600952088833,
      "num_tokens": 51430620.0,
      "step": 6700
    },
    {
      "entropy": 0.8989372082054615,
      "epoch": 0.976,
      "grad_norm": 0.31459614634513855,
      "learning_rate": 2.173434411007767e-07,
      "loss": 0.8799,
      "mean_token_accuracy": 0.8186790086328983,
      "num_tokens": 51504902.0,
      "step": 6710
    },
    {
      "entropy": 0.9168683797121048,
      "epoch": 0.9774545454545455,
      "grad_norm": 0.33197900652885437,
      "learning_rate": 2.036410149031669e-07,
      "loss": 0.8963,
      "mean_token_accuracy": 0.8159121349453926,
      "num_tokens": 51579884.0,
      "step": 6720
    },
    {
      "entropy": 0.9148142851889134,
      "epoch": 0.978909090909091,
      "grad_norm": 0.3526170551776886,
      "learning_rate": 1.907873866664188e-07,
      "loss": 0.8994,
      "mean_token_accuracy": 0.8167805947363377,
      "num_tokens": 51655972.0,
      "step": 6730
    },
    {
      "entropy": 0.8993205860257149,
      "epoch": 0.9803636363636363,
      "grad_norm": 0.3287599980831146,
      "learning_rate": 1.7878299549444739e-07,
      "loss": 0.8829,
      "mean_token_accuracy": 0.8196272850036621,
      "num_tokens": 51733439.0,
      "step": 6740
    },
    {
      "entropy": 0.9057880103588104,
      "epoch": 0.9818181818181818,
      "grad_norm": 0.30253535509109497,
      "learning_rate": 1.6762825147964345e-07,
      "loss": 0.8909,
      "mean_token_accuracy": 0.818729893118143,
      "num_tokens": 51807602.0,
      "step": 6750
    },
    {
      "entropy": 0.9042447462677956,
      "epoch": 0.9832727272727273,
      "grad_norm": 0.32170674204826355,
      "learning_rate": 1.5732353568886325e-07,
      "loss": 0.8988,
      "mean_token_accuracy": 0.8172514781355857,
      "num_tokens": 51882868.0,
      "step": 6760
    },
    {
      "entropy": 0.9321618229150772,
      "epoch": 0.9847272727272727,
      "grad_norm": 0.33762219548225403,
      "learning_rate": 1.4786920015042024e-07,
      "loss": 0.9191,
      "mean_token_accuracy": 0.812790434062481,
      "num_tokens": 51960042.0,
      "step": 6770
    },
    {
      "entropy": 0.9293069154024124,
      "epoch": 0.9861818181818182,
      "grad_norm": 0.318977952003479,
      "learning_rate": 1.3926556784204404e-07,
      "loss": 0.921,
      "mean_token_accuracy": 0.8137647397816181,
      "num_tokens": 52035665.0,
      "step": 6780
    },
    {
      "entropy": 0.9178681053221226,
      "epoch": 0.9876363636363636,
      "grad_norm": 0.339473694562912,
      "learning_rate": 1.3151293267986418e-07,
      "loss": 0.9104,
      "mean_token_accuracy": 0.8139139614999295,
      "num_tokens": 52111485.0,
      "step": 6790
    },
    {
      "entropy": 0.9090309455990792,
      "epoch": 0.9890909090909091,
      "grad_norm": 0.31843292713165283,
      "learning_rate": 1.2461155950835192e-07,
      "loss": 0.9014,
      "mean_token_accuracy": 0.8167950376868248,
      "num_tokens": 52188521.0,
      "step": 6800
    },
    {
      "entropy": 0.9286883376538754,
      "epoch": 0.9905454545454545,
      "grad_norm": 0.3196488320827484,
      "learning_rate": 1.1856168409129034e-07,
      "loss": 0.922,
      "mean_token_accuracy": 0.8129499651491642,
      "num_tokens": 52264651.0,
      "step": 6810
    },
    {
      "entropy": 0.9294882476329803,
      "epoch": 0.992,
      "grad_norm": 0.3718339502811432,
      "learning_rate": 1.1336351310370239e-07,
      "loss": 0.9114,
      "mean_token_accuracy": 0.8124035835266114,
      "num_tokens": 52341010.0,
      "step": 6820
    },
    {
      "entropy": 0.9305678024888039,
      "epoch": 0.9934545454545455,
      "grad_norm": 0.34219232201576233,
      "learning_rate": 1.0901722412480678e-07,
      "loss": 0.9258,
      "mean_token_accuracy": 0.8104947313666344,
      "num_tokens": 52419005.0,
      "step": 6830
    },
    {
      "entropy": 0.9108941599726676,
      "epoch": 0.9949090909090909,
      "grad_norm": 0.32488080859184265,
      "learning_rate": 1.0552296563194055e-07,
      "loss": 0.899,
      "mean_token_accuracy": 0.8151013270020485,
      "num_tokens": 52497112.0,
      "step": 6840
    },
    {
      "entropy": 0.9331740163266659,
      "epoch": 0.9963636363636363,
      "grad_norm": 0.31630396842956543,
      "learning_rate": 1.0288085699548733e-07,
      "loss": 0.9185,
      "mean_token_accuracy": 0.8122143365442753,
      "num_tokens": 52575897.0,
      "step": 6850
    },
    {
      "entropy": 0.9373132646083832,
      "epoch": 0.9978181818181818,
      "grad_norm": 0.3261500895023346,
      "learning_rate": 1.01090988474808e-07,
      "loss": 0.929,
      "mean_token_accuracy": 0.812446454167366,
      "num_tokens": 52653425.0,
      "step": 6860
    },
    {
      "entropy": 0.9317662589251995,
      "epoch": 0.9992727272727273,
      "grad_norm": 0.32642170786857605,
      "learning_rate": 1.0015342121514953e-07,
      "loss": 0.9138,
      "mean_token_accuracy": 0.8126379199326038,
      "num_tokens": 52733879.0,
      "step": 6870
    }
  ],
  "logging_steps": 10,
  "max_steps": 6875,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.344753466386416e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}