{
  "best_global_step": 300,
  "best_metric": 0.18106622993946075,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 50,
  "global_step": 338,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 3.775044322013855,
      "epoch": 0.002962962962962963,
      "grad_norm": 34.59568786621094,
      "learning_rate": 0.0,
      "loss": 2.628934621810913,
      "mean_token_accuracy": 0.6272874362766743,
      "num_tokens": 16384.0,
      "step": 1
    },
    {
      "entropy": 3.8163256645202637,
      "epoch": 0.005925925925925926,
      "grad_norm": 30.7418212890625,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.4589571952819824,
      "mean_token_accuracy": 0.6306841522455215,
      "num_tokens": 32768.0,
      "step": 2
    },
    {
      "entropy": 3.4858527183532715,
      "epoch": 0.008888888888888889,
      "grad_norm": 28.812597274780273,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.1560466289520264,
      "mean_token_accuracy": 0.6737441420555115,
      "num_tokens": 49152.0,
      "step": 3
    },
    {
      "entropy": 3.8859923779964447,
      "epoch": 0.011851851851851851,
      "grad_norm": 33.05686569213867,
      "learning_rate": 6e-06,
      "loss": 2.643793821334839,
      "mean_token_accuracy": 0.619361087679863,
      "num_tokens": 65536.0,
      "step": 4
    },
    {
      "entropy": 3.7973375618457794,
      "epoch": 0.014814814814814815,
      "grad_norm": 33.31623458862305,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.9691884517669678,
      "mean_token_accuracy": 0.6821821480989456,
      "num_tokens": 81920.0,
      "step": 5
    },
    {
      "entropy": 3.6109770834445953,
      "epoch": 0.017777777777777778,
      "grad_norm": 19.2442684173584,
      "learning_rate": 1e-05,
      "loss": 1.6936599016189575,
      "mean_token_accuracy": 0.6984509825706482,
      "num_tokens": 98304.0,
      "step": 6
    },
    {
      "entropy": 3.315625935792923,
      "epoch": 0.02074074074074074,
      "grad_norm": 9.899496078491211,
      "learning_rate": 9.96996996996997e-06,
      "loss": 1.184874176979065,
      "mean_token_accuracy": 0.7923205718398094,
      "num_tokens": 114688.0,
      "step": 7
    },
    {
      "entropy": 3.6855034828186035,
      "epoch": 0.023703703703703703,
      "grad_norm": 8.208796501159668,
      "learning_rate": 9.93993993993994e-06,
      "loss": 1.1628938913345337,
      "mean_token_accuracy": 0.7737714499235153,
      "num_tokens": 131072.0,
      "step": 8
    },
    {
      "entropy": 3.525162488222122,
      "epoch": 0.02666666666666667,
      "grad_norm": 7.624190807342529,
      "learning_rate": 9.90990990990991e-06,
      "loss": 1.3821073770523071,
      "mean_token_accuracy": 0.7611671090126038,
      "num_tokens": 147456.0,
      "step": 9
    },
    {
      "entropy": 3.209045648574829,
      "epoch": 0.02962962962962963,
      "grad_norm": 5.016626358032227,
      "learning_rate": 9.879879879879881e-06,
      "loss": 0.912333607673645,
      "mean_token_accuracy": 0.8256426826119423,
      "num_tokens": 163840.0,
      "step": 10
    },
    {
      "entropy": 3.525566816329956,
      "epoch": 0.03259259259259259,
      "grad_norm": 4.671548843383789,
      "learning_rate": 9.849849849849851e-06,
      "loss": 0.886758029460907,
      "mean_token_accuracy": 0.8259894177317619,
      "num_tokens": 180224.0,
      "step": 11
    },
    {
      "entropy": 3.328701972961426,
      "epoch": 0.035555555555555556,
      "grad_norm": 5.48918342590332,
      "learning_rate": 9.81981981981982e-06,
      "loss": 1.2401018142700195,
      "mean_token_accuracy": 0.7638061940670013,
      "num_tokens": 196608.0,
      "step": 12
    },
    {
      "entropy": 3.365691661834717,
      "epoch": 0.03851851851851852,
      "grad_norm": 3.175600528717041,
      "learning_rate": 9.78978978978979e-06,
      "loss": 0.6797101497650146,
      "mean_token_accuracy": 0.8517325296998024,
      "num_tokens": 212992.0,
      "step": 13
    },
    {
      "entropy": 3.2332957088947296,
      "epoch": 0.04148148148148148,
      "grad_norm": 4.001441955566406,
      "learning_rate": 9.75975975975976e-06,
      "loss": 0.9612078070640564,
      "mean_token_accuracy": 0.8044033944606781,
      "num_tokens": 229376.0,
      "step": 14
    },
    {
      "entropy": 3.103601574897766,
      "epoch": 0.044444444444444446,
      "grad_norm": 3.249359130859375,
      "learning_rate": 9.729729729729732e-06,
      "loss": 0.6849663257598877,
      "mean_token_accuracy": 0.8227220624685287,
      "num_tokens": 245760.0,
      "step": 15
    },
    {
      "entropy": 2.968333125114441,
      "epoch": 0.047407407407407405,
      "grad_norm": 3.55027174949646,
      "learning_rate": 9.699699699699701e-06,
      "loss": 0.840543806552887,
      "mean_token_accuracy": 0.8257830664515495,
      "num_tokens": 262144.0,
      "step": 16
    },
    {
      "entropy": 3.075542002916336,
      "epoch": 0.05037037037037037,
      "grad_norm": 2.2393057346343994,
      "learning_rate": 9.669669669669671e-06,
      "loss": 0.4668881893157959,
      "mean_token_accuracy": 0.8647300750017166,
      "num_tokens": 278528.0,
      "step": 17
    },
    {
      "entropy": 2.8269824981689453,
      "epoch": 0.05333333333333334,
      "grad_norm": 3.366901159286499,
      "learning_rate": 9.63963963963964e-06,
      "loss": 0.6040822267532349,
      "mean_token_accuracy": 0.8704800009727478,
      "num_tokens": 294912.0,
      "step": 18
    },
    {
      "entropy": 3.2838538885116577,
      "epoch": 0.056296296296296296,
      "grad_norm": 2.3267767429351807,
      "learning_rate": 9.60960960960961e-06,
      "loss": 0.515355110168457,
      "mean_token_accuracy": 0.8790227174758911,
      "num_tokens": 311296.0,
      "step": 19
    },
    {
      "entropy": 2.8646548986434937,
      "epoch": 0.05925925925925926,
      "grad_norm": 2.4615976810455322,
      "learning_rate": 9.57957957957958e-06,
      "loss": 0.46175044775009155,
      "mean_token_accuracy": 0.8869305402040482,
      "num_tokens": 327680.0,
      "step": 20
    },
    {
      "entropy": 3.0974441468715668,
      "epoch": 0.06222222222222222,
      "grad_norm": 1.576017141342163,
      "learning_rate": 9.54954954954955e-06,
      "loss": 0.3317035734653473,
      "mean_token_accuracy": 0.9089246764779091,
      "num_tokens": 344064.0,
      "step": 21
    },
    {
      "entropy": 3.305128753185272,
      "epoch": 0.06518518518518518,
      "grad_norm": 2.2638797760009766,
      "learning_rate": 9.51951951951952e-06,
      "loss": 0.42593857645988464,
      "mean_token_accuracy": 0.8900623396039009,
      "num_tokens": 360448.0,
      "step": 22
    },
    {
      "entropy": 3.1091228127479553,
      "epoch": 0.06814814814814815,
      "grad_norm": 2.7307915687561035,
      "learning_rate": 9.489489489489491e-06,
      "loss": 0.5931733846664429,
      "mean_token_accuracy": 0.8545943424105644,
      "num_tokens": 376832.0,
      "step": 23
    },
    {
      "entropy": 3.254545032978058,
      "epoch": 0.07111111111111111,
      "grad_norm": 1.9047656059265137,
      "learning_rate": 9.45945945945946e-06,
      "loss": 0.40604934096336365,
      "mean_token_accuracy": 0.8869926705956459,
      "num_tokens": 393216.0,
      "step": 24
    },
    {
      "entropy": 3.5277936160564423,
      "epoch": 0.07407407407407407,
      "grad_norm": 2.0294620990753174,
      "learning_rate": 9.42942942942943e-06,
      "loss": 0.5543591976165771,
      "mean_token_accuracy": 0.8743266761302948,
      "num_tokens": 409600.0,
      "step": 25
    },
    {
      "entropy": 3.4342984557151794,
      "epoch": 0.07703703703703704,
      "grad_norm": 1.7908217906951904,
      "learning_rate": 9.3993993993994e-06,
      "loss": 0.42756205797195435,
      "mean_token_accuracy": 0.8926278129220009,
      "num_tokens": 425984.0,
      "step": 26
    },
    {
      "entropy": 3.2699481546878815,
      "epoch": 0.08,
      "grad_norm": 1.3884456157684326,
      "learning_rate": 9.36936936936937e-06,
      "loss": 0.2644193172454834,
      "mean_token_accuracy": 0.9317126050591469,
      "num_tokens": 442368.0,
      "step": 27
    },
    {
      "entropy": 3.277682840824127,
      "epoch": 0.08296296296296296,
      "grad_norm": 1.7567063570022583,
      "learning_rate": 9.339339339339341e-06,
      "loss": 0.37994882464408875,
      "mean_token_accuracy": 0.9112046957015991,
      "num_tokens": 458752.0,
      "step": 28
    },
    {
      "entropy": 3.319206029176712,
      "epoch": 0.08592592592592592,
      "grad_norm": 1.8274697065353394,
      "learning_rate": 9.30930930930931e-06,
      "loss": 0.42217355966567993,
      "mean_token_accuracy": 0.8770820647478104,
      "num_tokens": 475136.0,
      "step": 29
    },
    {
      "entropy": 3.1831164062023163,
      "epoch": 0.08888888888888889,
      "grad_norm": 1.5749222040176392,
      "learning_rate": 9.27927927927928e-06,
      "loss": 0.31777235865592957,
      "mean_token_accuracy": 0.9044679179787636,
      "num_tokens": 491520.0,
      "step": 30
    },
    {
      "entropy": 3.363261044025421,
      "epoch": 0.09185185185185185,
      "grad_norm": 1.7027117013931274,
      "learning_rate": 9.24924924924925e-06,
      "loss": 0.4325081408023834,
      "mean_token_accuracy": 0.8854201510548592,
      "num_tokens": 507904.0,
      "step": 31
    },
    {
      "entropy": 3.029395565390587,
      "epoch": 0.09481481481481481,
      "grad_norm": 1.3480552434921265,
      "learning_rate": 9.21921921921922e-06,
      "loss": 0.305914968252182,
      "mean_token_accuracy": 0.9203363656997681,
      "num_tokens": 524288.0,
      "step": 32
    },
    {
      "entropy": 3.2645480632781982,
      "epoch": 0.09777777777777778,
      "grad_norm": 1.431160569190979,
      "learning_rate": 9.189189189189191e-06,
      "loss": 0.31388601660728455,
      "mean_token_accuracy": 0.9192172512412071,
      "num_tokens": 540672.0,
      "step": 33
    },
    {
      "entropy": 3.094198167324066,
      "epoch": 0.10074074074074074,
      "grad_norm": 1.1171891689300537,
      "learning_rate": 9.15915915915916e-06,
      "loss": 0.20084995031356812,
      "mean_token_accuracy": 0.9483551234006882,
      "num_tokens": 557056.0,
      "step": 34
    },
    {
      "entropy": 2.9308216273784637,
      "epoch": 0.1037037037037037,
      "grad_norm": 1.0611443519592285,
      "learning_rate": 9.129129129129129e-06,
      "loss": 0.2329464554786682,
      "mean_token_accuracy": 0.943843275308609,
      "num_tokens": 573440.0,
      "step": 35
    },
    {
      "entropy": 2.862432152032852,
      "epoch": 0.10666666666666667,
      "grad_norm": 1.6051803827285767,
      "learning_rate": 9.0990990990991e-06,
      "loss": 0.4780491888523102,
      "mean_token_accuracy": 0.8832163512706757,
      "num_tokens": 589824.0,
      "step": 36
    },
    {
      "entropy": 2.8164885342121124,
      "epoch": 0.10962962962962963,
      "grad_norm": 1.1160067319869995,
      "learning_rate": 9.06906906906907e-06,
      "loss": 0.21563442051410675,
      "mean_token_accuracy": 0.9486184194684029,
      "num_tokens": 606208.0,
      "step": 37
    },
    {
      "entropy": 2.6916711181402206,
      "epoch": 0.11259259259259259,
      "grad_norm": 1.0230157375335693,
      "learning_rate": 9.03903903903904e-06,
      "loss": 0.18906882405281067,
      "mean_token_accuracy": 0.9485798850655556,
      "num_tokens": 622592.0,
      "step": 38
    },
    {
      "entropy": 2.820018947124481,
      "epoch": 0.11555555555555555,
      "grad_norm": 1.1773149967193604,
      "learning_rate": 9.00900900900901e-06,
      "loss": 0.25656864047050476,
      "mean_token_accuracy": 0.9308653548359871,
      "num_tokens": 638976.0,
      "step": 39
    },
    {
      "entropy": 2.89467953145504,
      "epoch": 0.11851851851851852,
      "grad_norm": 1.4061431884765625,
      "learning_rate": 8.97897897897898e-06,
      "loss": 0.347366601228714,
      "mean_token_accuracy": 0.910783052444458,
      "num_tokens": 655360.0,
      "step": 40
    },
    {
      "entropy": 2.7638043463230133,
      "epoch": 0.12148148148148148,
      "grad_norm": 1.164523720741272,
      "learning_rate": 8.94894894894895e-06,
      "loss": 0.32213348150253296,
      "mean_token_accuracy": 0.9277858883142471,
      "num_tokens": 671744.0,
      "step": 41
    },
    {
      "entropy": 2.706083133816719,
      "epoch": 0.12444444444444444,
      "grad_norm": 1.518288493156433,
      "learning_rate": 8.91891891891892e-06,
      "loss": 0.33784160017967224,
      "mean_token_accuracy": 0.9121690541505814,
      "num_tokens": 688128.0,
      "step": 42
    },
    {
      "entropy": 2.864104151725769,
      "epoch": 0.1274074074074074,
      "grad_norm": 1.2777132987976074,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.27047857642173767,
      "mean_token_accuracy": 0.9295907989144325,
      "num_tokens": 704512.0,
      "step": 43
    },
    {
      "entropy": 3.2951397001743317,
      "epoch": 0.13037037037037036,
      "grad_norm": 1.4618570804595947,
      "learning_rate": 8.85885885885886e-06,
      "loss": 0.35400456190109253,
      "mean_token_accuracy": 0.9075571447610855,
      "num_tokens": 720896.0,
      "step": 44
    },
    {
      "entropy": 2.64460352063179,
      "epoch": 0.13333333333333333,
      "grad_norm": 1.2006359100341797,
      "learning_rate": 8.82882882882883e-06,
      "loss": 0.3249610662460327,
      "mean_token_accuracy": 0.9128459766507149,
      "num_tokens": 737280.0,
      "step": 45
    },
    {
      "entropy": 2.5624406039714813,
      "epoch": 0.1362962962962963,
      "grad_norm": 0.9426330327987671,
      "learning_rate": 8.798798798798799e-06,
      "loss": 0.1744665950536728,
      "mean_token_accuracy": 0.946782298386097,
      "num_tokens": 753664.0,
      "step": 46
    },
    {
      "entropy": 3.0884713530540466,
      "epoch": 0.13925925925925925,
      "grad_norm": 1.4970332384109497,
      "learning_rate": 8.768768768768769e-06,
      "loss": 0.29483044147491455,
      "mean_token_accuracy": 0.9224485829472542,
      "num_tokens": 770048.0,
      "step": 47
    },
    {
      "entropy": 2.705291450023651,
      "epoch": 0.14222222222222222,
      "grad_norm": 0.8782689571380615,
      "learning_rate": 8.738738738738739e-06,
      "loss": 0.13383053243160248,
      "mean_token_accuracy": 0.9630647003650665,
      "num_tokens": 786432.0,
      "step": 48
    },
    {
      "entropy": 2.8029175400733948,
      "epoch": 0.1451851851851852,
      "grad_norm": 1.2086946964263916,
      "learning_rate": 8.70870870870871e-06,
      "loss": 0.26702427864074707,
      "mean_token_accuracy": 0.9370834156870842,
      "num_tokens": 802816.0,
      "step": 49
    },
    {
      "entropy": 2.710858181118965,
      "epoch": 0.14814814814814814,
      "grad_norm": 1.1601444482803345,
      "learning_rate": 8.67867867867868e-06,
      "loss": 0.3043576776981354,
      "mean_token_accuracy": 0.9150896891951561,
      "num_tokens": 819200.0,
      "step": 50
    },
    {
      "epoch": 0.14814814814814814,
      "eval_entropy": 2.8709708706537884,
      "eval_loss": 0.2575305700302124,
      "eval_mean_token_accuracy": 0.9331695469220479,
      "eval_num_tokens": 819200.0,
      "eval_runtime": 46.181,
      "eval_samples_per_second": 12.992,
      "eval_steps_per_second": 1.624,
      "step": 50
    },
    {
      "entropy": 2.8717176020145416,
      "epoch": 0.1511111111111111,
      "grad_norm": 1.1134225130081177,
      "learning_rate": 8.64864864864865e-06,
      "loss": 0.23674002289772034,
      "mean_token_accuracy": 0.9364274367690086,
      "num_tokens": 835584.0,
      "step": 51
    },
    {
      "entropy": 3.07846599817276,
      "epoch": 0.15407407407407409,
      "grad_norm": 1.212401032447815,
      "learning_rate": 8.618618618618619e-06,
      "loss": 0.2758175730705261,
      "mean_token_accuracy": 0.9202386289834976,
      "num_tokens": 851968.0,
      "step": 52
    },
    {
      "entropy": 2.928554356098175,
      "epoch": 0.15703703703703703,
      "grad_norm": 1.3330804109573364,
      "learning_rate": 8.588588588588589e-06,
      "loss": 0.36216461658477783,
      "mean_token_accuracy": 0.914257138967514,
      "num_tokens": 868352.0,
      "step": 53
    },
    {
      "entropy": 3.32950496673584,
      "epoch": 0.16,
      "grad_norm": 1.2147783041000366,
      "learning_rate": 8.55855855855856e-06,
      "loss": 0.28390607237815857,
      "mean_token_accuracy": 0.9206364154815674,
      "num_tokens": 884736.0,
      "step": 54
    },
    {
      "entropy": 3.175953209400177,
      "epoch": 0.16296296296296298,
      "grad_norm": 1.3068032264709473,
      "learning_rate": 8.52852852852853e-06,
      "loss": 0.23997928202152252,
      "mean_token_accuracy": 0.944263756275177,
      "num_tokens": 901120.0,
      "step": 55
    },
    {
      "entropy": 3.0237551629543304,
      "epoch": 0.16592592592592592,
      "grad_norm": 0.9807406067848206,
      "learning_rate": 8.4984984984985e-06,
      "loss": 0.22529760003089905,
      "mean_token_accuracy": 0.9395471438765526,
      "num_tokens": 917504.0,
      "step": 56
    },
    {
      "entropy": 3.324373722076416,
      "epoch": 0.1688888888888889,
      "grad_norm": 1.388300895690918,
      "learning_rate": 8.46846846846847e-06,
      "loss": 0.30894917249679565,
      "mean_token_accuracy": 0.9203754514455795,
      "num_tokens": 933888.0,
      "step": 57
    },
    {
      "entropy": 2.7935754358768463,
      "epoch": 0.17185185185185184,
      "grad_norm": 0.9797722697257996,
      "learning_rate": 8.438438438438439e-06,
      "loss": 0.24637249112129211,
      "mean_token_accuracy": 0.9311194345355034,
      "num_tokens": 950272.0,
      "step": 58
    },
    {
      "entropy": 3.1923772394657135,
      "epoch": 0.1748148148148148,
      "grad_norm": 1.0256638526916504,
      "learning_rate": 8.408408408408409e-06,
      "loss": 0.21505799889564514,
      "mean_token_accuracy": 0.9484226405620575,
      "num_tokens": 966656.0,
      "step": 59
    },
    {
      "entropy": 3.2502397298812866,
      "epoch": 0.17777777777777778,
      "grad_norm": 0.8857279419898987,
      "learning_rate": 8.378378378378378e-06,
      "loss": 0.12016204744577408,
      "mean_token_accuracy": 0.9599257409572601,
      "num_tokens": 983040.0,
      "step": 60
    },
    {
      "entropy": 2.906019926071167,
      "epoch": 0.18074074074074073,
      "grad_norm": 1.305699348449707,
      "learning_rate": 8.348348348348348e-06,
      "loss": 0.301448792219162,
      "mean_token_accuracy": 0.9159760102629662,
      "num_tokens": 999424.0,
      "step": 61
    },
    {
      "entropy": 3.066522181034088,
      "epoch": 0.1837037037037037,
      "grad_norm": 1.236630916595459,
      "learning_rate": 8.31831831831832e-06,
      "loss": 0.31222307682037354,
      "mean_token_accuracy": 0.9108500853180885,
      "num_tokens": 1015808.0,
      "step": 62
    },
    {
      "entropy": 3.1751776039600372,
      "epoch": 0.18666666666666668,
      "grad_norm": 1.095410943031311,
      "learning_rate": 8.288288288288289e-06,
      "loss": 0.2780705988407135,
      "mean_token_accuracy": 0.9205356687307358,
      "num_tokens": 1032192.0,
      "step": 63
    },
    {
      "entropy": 3.069409281015396,
      "epoch": 0.18962962962962962,
      "grad_norm": 1.1574227809906006,
      "learning_rate": 8.258258258258259e-06,
      "loss": 0.24383366107940674,
      "mean_token_accuracy": 0.940370611846447,
      "num_tokens": 1048576.0,
      "step": 64
    },
    {
      "entropy": 3.271954983472824,
      "epoch": 0.1925925925925926,
      "grad_norm": 1.4608582258224487,
      "learning_rate": 8.228228228228229e-06,
      "loss": 0.3501364588737488,
      "mean_token_accuracy": 0.896758683025837,
      "num_tokens": 1064960.0,
      "step": 65
    },
    {
      "entropy": 3.1587420105934143,
      "epoch": 0.19555555555555557,
      "grad_norm": 1.1120469570159912,
      "learning_rate": 8.198198198198198e-06,
      "loss": 0.19030891358852386,
      "mean_token_accuracy": 0.9521929174661636,
      "num_tokens": 1081344.0,
      "step": 66
    },
    {
      "entropy": 3.175812691450119,
      "epoch": 0.1985185185185185,
      "grad_norm": 1.6005663871765137,
      "learning_rate": 8.16816816816817e-06,
      "loss": 0.373776376247406,
      "mean_token_accuracy": 0.8996462225914001,
      "num_tokens": 1097728.0,
      "step": 67
    },
    {
      "entropy": 3.047615200281143,
      "epoch": 0.20148148148148148,
      "grad_norm": 1.1062686443328857,
      "learning_rate": 8.13813813813814e-06,
      "loss": 0.20890432596206665,
      "mean_token_accuracy": 0.9433711990714073,
      "num_tokens": 1114112.0,
      "step": 68
    },
    {
      "entropy": 2.5505146980285645,
      "epoch": 0.20444444444444446,
      "grad_norm": 0.8427297472953796,
      "learning_rate": 8.108108108108109e-06,
      "loss": 0.13734661042690277,
      "mean_token_accuracy": 0.9604312181472778,
      "num_tokens": 1130496.0,
      "step": 69
    },
    {
      "entropy": 2.911862313747406,
      "epoch": 0.2074074074074074,
      "grad_norm": 1.2946555614471436,
      "learning_rate": 8.078078078078079e-06,
      "loss": 0.22861933708190918,
      "mean_token_accuracy": 0.9296081736683846,
      "num_tokens": 1146880.0,
      "step": 70
    },
    {
      "entropy": 2.86180517077446,
      "epoch": 0.21037037037037037,
      "grad_norm": 1.1980105638504028,
      "learning_rate": 8.048048048048048e-06,
      "loss": 0.2675926685333252,
      "mean_token_accuracy": 0.9246023669838905,
      "num_tokens": 1163264.0,
      "step": 71
    },
    {
      "entropy": 3.2405199706554413,
      "epoch": 0.21333333333333335,
      "grad_norm": 1.065236210823059,
      "learning_rate": 8.018018018018018e-06,
      "loss": 0.23876021802425385,
      "mean_token_accuracy": 0.9408073797821999,
      "num_tokens": 1179648.0,
      "step": 72
    },
    {
      "entropy": 2.9095458686351776,
      "epoch": 0.2162962962962963,
      "grad_norm": 0.8881495594978333,
      "learning_rate": 7.987987987987988e-06,
      "loss": 0.18038365244865417,
      "mean_token_accuracy": 0.9497027918696404,
      "num_tokens": 1196032.0,
      "step": 73
    },
    {
      "entropy": 3.1068210899829865,
      "epoch": 0.21925925925925926,
      "grad_norm": 1.3357926607131958,
      "learning_rate": 7.95795795795796e-06,
      "loss": 0.35012519359588623,
      "mean_token_accuracy": 0.9102579206228256,
      "num_tokens": 1212416.0,
      "step": 74
    },
    {
      "entropy": 3.077962279319763,
      "epoch": 0.2222222222222222,
      "grad_norm": 1.329525351524353,
      "learning_rate": 7.927927927927929e-06,
      "loss": 0.27650079131126404,
      "mean_token_accuracy": 0.9293218851089478,
      "num_tokens": 1228800.0,
      "step": 75
    },
    {
      "entropy": 2.9894566237926483,
      "epoch": 0.22518518518518518,
      "grad_norm": 1.0652294158935547,
      "learning_rate": 7.897897897897899e-06,
      "loss": 0.23853328824043274,
      "mean_token_accuracy": 0.9362653195858002,
      "num_tokens": 1245184.0,
      "step": 76
    },
    {
      "entropy": 2.8854158222675323,
      "epoch": 0.22814814814814816,
      "grad_norm": 1.154850959777832,
      "learning_rate": 7.867867867867868e-06,
      "loss": 0.2553675174713135,
      "mean_token_accuracy": 0.9320229142904282,
      "num_tokens": 1261568.0,
      "step": 77
    },
    {
      "entropy": 3.015660434961319,
      "epoch": 0.2311111111111111,
      "grad_norm": 1.5270060300827026,
      "learning_rate": 7.837837837837838e-06,
      "loss": 0.3484768569469452,
      "mean_token_accuracy": 0.9260349422693253,
      "num_tokens": 1277952.0,
      "step": 78
    },
    {
      "entropy": 2.9225947558879852,
      "epoch": 0.23407407407407407,
      "grad_norm": 1.3085774183273315,
      "learning_rate": 7.807807807807808e-06,
      "loss": 0.2533874809741974,
      "mean_token_accuracy": 0.9236741736531258,
      "num_tokens": 1294336.0,
      "step": 79
    },
    {
      "entropy": 3.013516530394554,
      "epoch": 0.23703703703703705,
      "grad_norm": 1.0632835626602173,
      "learning_rate": 7.77777777777778e-06,
      "loss": 0.2117396742105484,
      "mean_token_accuracy": 0.9344713315367699,
      "num_tokens": 1310720.0,
      "step": 80
    },
    {
      "entropy": 2.938462883234024,
      "epoch": 0.24,
      "grad_norm": 0.9691763520240784,
      "learning_rate": 7.747747747747749e-06,
      "loss": 0.18630896508693695,
      "mean_token_accuracy": 0.9417071342468262,
      "num_tokens": 1327104.0,
      "step": 81
    },
    {
      "entropy": 3.2490529119968414,
      "epoch": 0.24296296296296296,
      "grad_norm": 1.1068042516708374,
      "learning_rate": 7.717717717717719e-06,
      "loss": 0.1994417905807495,
      "mean_token_accuracy": 0.9386816993355751,
      "num_tokens": 1343488.0,
      "step": 82
    },
    {
      "entropy": 2.7314798682928085,
      "epoch": 0.24592592592592594,
      "grad_norm": 1.0012438297271729,
      "learning_rate": 7.687687687687688e-06,
      "loss": 0.19472742080688477,
      "mean_token_accuracy": 0.9480613172054291,
      "num_tokens": 1359872.0,
      "step": 83
    },
    {
      "entropy": 2.899082988500595,
      "epoch": 0.24888888888888888,
      "grad_norm": 0.9162399172782898,
      "learning_rate": 7.657657657657658e-06,
      "loss": 0.15804064273834229,
      "mean_token_accuracy": 0.9564466997981071,
      "num_tokens": 1376256.0,
      "step": 84
    },
    {
      "entropy": 3.07162669301033,
      "epoch": 0.2518518518518518,
      "grad_norm": 1.0021650791168213,
      "learning_rate": 7.6276276276276285e-06,
      "loss": 0.19935385882854462,
      "mean_token_accuracy": 0.9392828047275543,
      "num_tokens": 1392640.0,
      "step": 85
    },
    {
      "entropy": 3.085723578929901,
      "epoch": 0.2548148148148148,
      "grad_norm": 1.0430418252944946,
      "learning_rate": 7.597597597597598e-06,
      "loss": 0.16674907505512238,
      "mean_token_accuracy": 0.9378847703337669,
      "num_tokens": 1409024.0,
      "step": 86
    },
    {
      "entropy": 2.8088001906871796,
      "epoch": 0.2577777777777778,
      "grad_norm": 1.2396692037582397,
      "learning_rate": 7.567567567567569e-06,
      "loss": 0.2708350121974945,
      "mean_token_accuracy": 0.9323693364858627,
      "num_tokens": 1425408.0,
      "step": 87
    },
    {
      "entropy": 2.567664086818695,
      "epoch": 0.2607407407407407,
      "grad_norm": 0.9123518466949463,
      "learning_rate": 7.5375375375375385e-06,
      "loss": 0.17232848703861237,
      "mean_token_accuracy": 0.9493565931916237,
      "num_tokens": 1441792.0,
      "step": 88
    },
    {
      "entropy": 2.7368578910827637,
      "epoch": 0.2637037037037037,
      "grad_norm": 1.1355301141738892,
      "learning_rate": 7.507507507507507e-06,
      "loss": 0.203575000166893,
      "mean_token_accuracy": 0.9447965919971466,
      "num_tokens": 1458176.0,
      "step": 89
    },
    {
      "entropy": 2.9226705133914948,
      "epoch": 0.26666666666666666,
      "grad_norm": 0.8666026592254639,
      "learning_rate": 7.477477477477479e-06,
      "loss": 0.15507856011390686,
      "mean_token_accuracy": 0.957670234143734,
      "num_tokens": 1474560.0,
      "step": 90
    },
    {
      "entropy": 3.011612117290497,
      "epoch": 0.2696296296296296,
      "grad_norm": 0.9835048317909241,
      "learning_rate": 7.447447447447448e-06,
      "loss": 0.19631719589233398,
      "mean_token_accuracy": 0.9474716559052467,
      "num_tokens": 1490944.0,
      "step": 91
    },
    {
      "entropy": 3.0336680710315704,
      "epoch": 0.2725925925925926,
      "grad_norm": 1.0945093631744385,
      "learning_rate": 7.417417417417418e-06,
      "loss": 0.18349133431911469,
      "mean_token_accuracy": 0.9488312005996704,
      "num_tokens": 1507328.0,
      "step": 92
    },
    {
      "entropy": 2.9798973202705383,
      "epoch": 0.27555555555555555,
      "grad_norm": 1.2349214553833008,
      "learning_rate": 7.387387387387388e-06,
      "loss": 0.2918708324432373,
      "mean_token_accuracy": 0.9265884160995483,
      "num_tokens": 1523712.0,
      "step": 93
    },
    {
      "entropy": 3.068564772605896,
      "epoch": 0.2785185185185185,
      "grad_norm": 1.2158499956130981,
      "learning_rate": 7.3573573573573575e-06,
      "loss": 0.2323075234889984,
      "mean_token_accuracy": 0.9372951462864876,
      "num_tokens": 1540096.0,
      "step": 94
    },
    {
      "entropy": 2.868547946214676,
      "epoch": 0.2814814814814815,
      "grad_norm": 0.8773489594459534,
      "learning_rate": 7.327327327327328e-06,
      "loss": 0.1515350490808487,
      "mean_token_accuracy": 0.94919104129076,
      "num_tokens": 1556480.0,
      "step": 95
    },
    {
      "entropy": 2.9125973880290985,
      "epoch": 0.28444444444444444,
      "grad_norm": 1.1339553594589233,
      "learning_rate": 7.297297297297298e-06,
      "loss": 0.18595059216022491,
      "mean_token_accuracy": 0.9450299069285393,
      "num_tokens": 1572864.0,
      "step": 96
    },
    {
      "entropy": 2.830490857362747,
      "epoch": 0.2874074074074074,
      "grad_norm": 0.8534502983093262,
      "learning_rate": 7.267267267267268e-06,
      "loss": 0.1191851794719696,
      "mean_token_accuracy": 0.9684600383043289,
      "num_tokens": 1589248.0,
      "step": 97
    },
    {
      "entropy": 2.8104536086320877,
      "epoch": 0.2903703703703704,
      "grad_norm": 1.0791709423065186,
      "learning_rate": 7.237237237237238e-06,
      "loss": 0.20497238636016846,
      "mean_token_accuracy": 0.9426870495080948,
      "num_tokens": 1605632.0,
      "step": 98
    },
    {
      "entropy": 2.718022584915161,
      "epoch": 0.29333333333333333,
      "grad_norm": 0.8226582407951355,
      "learning_rate": 7.207207207207208e-06,
      "loss": 0.14803673326969147,
      "mean_token_accuracy": 0.9597157314419746,
      "num_tokens": 1622016.0,
      "step": 99
    },
    {
      "entropy": 2.726251184940338,
      "epoch": 0.2962962962962963,
      "grad_norm": 0.8652470111846924,
      "learning_rate": 7.177177177177178e-06,
      "loss": 0.1922772228717804,
      "mean_token_accuracy": 0.9483191147446632,
      "num_tokens": 1638400.0,
      "step": 100
    },
    {
      "epoch": 0.2962962962962963,
      "eval_entropy": 2.9197415765126546,
      "eval_loss": 0.20984426140785217,
      "eval_mean_token_accuracy": 0.9438043769200642,
      "eval_num_tokens": 1638400.0,
      "eval_runtime": 46.0698,
      "eval_samples_per_second": 13.024,
      "eval_steps_per_second": 1.628,
      "step": 100
    },
    {
      "entropy": 2.682308256626129,
      "epoch": 0.2992592592592593,
      "grad_norm": 0.8887585401535034,
      "learning_rate": 7.147147147147148e-06,
      "loss": 0.14127542078495026,
      "mean_token_accuracy": 0.9570844620466232,
      "num_tokens": 1654784.0,
      "step": 101
    },
    {
      "entropy": 2.7055087238550186,
      "epoch": 0.3022222222222222,
      "grad_norm": 1.0312213897705078,
      "learning_rate": 7.117117117117117e-06,
      "loss": 0.18035940825939178,
      "mean_token_accuracy": 0.9473381191492081,
      "num_tokens": 1671168.0,
      "step": 102
    },
    {
      "entropy": 2.8517268896102905,
      "epoch": 0.30518518518518517,
      "grad_norm": 1.208492398262024,
      "learning_rate": 7.087087087087087e-06,
      "loss": 0.2457740157842636,
      "mean_token_accuracy": 0.9201536476612091,
      "num_tokens": 1687552.0,
      "step": 103
    },
    {
      "entropy": 3.0285078287124634,
      "epoch": 0.30814814814814817,
      "grad_norm": 1.3902666568756104,
      "learning_rate": 7.057057057057057e-06,
      "loss": 0.33393436670303345,
      "mean_token_accuracy": 0.9189444556832314,
      "num_tokens": 1703936.0,
      "step": 104
    },
    {
      "entropy": 2.8324770629405975,
      "epoch": 0.3111111111111111,
      "grad_norm": 0.9838982224464417,
      "learning_rate": 7.027027027027028e-06,
      "loss": 0.12197890132665634,
      "mean_token_accuracy": 0.9580222964286804,
      "num_tokens": 1720320.0,
      "step": 105
    },
    {
      "entropy": 2.78508061170578,
      "epoch": 0.31407407407407406,
      "grad_norm": 1.1180880069732666,
      "learning_rate": 6.996996996996997e-06,
      "loss": 0.22881369292736053,
      "mean_token_accuracy": 0.9426014497876167,
      "num_tokens": 1736704.0,
      "step": 106
    },
    {
      "entropy": 3.0421781837940216,
      "epoch": 0.31703703703703706,
      "grad_norm": 0.9388776421546936,
      "learning_rate": 6.966966966966967e-06,
      "loss": 0.1317838430404663,
      "mean_token_accuracy": 0.9635236337780952,
      "num_tokens": 1753088.0,
      "step": 107
    },
    {
      "entropy": 2.7839683145284653,
      "epoch": 0.32,
      "grad_norm": 0.9274376630783081,
      "learning_rate": 6.936936936936938e-06,
      "loss": 0.20459799468517303,
      "mean_token_accuracy": 0.9468235448002815,
      "num_tokens": 1769472.0,
      "step": 108
    },
    {
      "entropy": 3.221632182598114,
      "epoch": 0.32296296296296295,
      "grad_norm": 1.4066088199615479,
      "learning_rate": 6.906906906906907e-06,
      "loss": 0.2937573790550232,
      "mean_token_accuracy": 0.9293190464377403,
      "num_tokens": 1785856.0,
      "step": 109
    },
    {
      "entropy": 3.1568872332572937,
      "epoch": 0.32592592592592595,
      "grad_norm": 1.0390819311141968,
      "learning_rate": 6.876876876876878e-06,
      "loss": 0.2432892620563507,
      "mean_token_accuracy": 0.9426203146576881,
      "num_tokens": 1802240.0,
      "step": 110
    },
    {
      "entropy": 3.1188774406909943,
      "epoch": 0.3288888888888889,
      "grad_norm": 1.1714736223220825,
      "learning_rate": 6.846846846846848e-06,
      "loss": 0.21787814795970917,
      "mean_token_accuracy": 0.9426657408475876,
      "num_tokens": 1818624.0,
      "step": 111
    },
    {
      "entropy": 3.054466634988785,
      "epoch": 0.33185185185185184,
      "grad_norm": 1.0604864358901978,
      "learning_rate": 6.816816816816817e-06,
      "loss": 0.17631398141384125,
      "mean_token_accuracy": 0.9548331052064896,
      "num_tokens": 1835008.0,
      "step": 112
    },
    {
      "entropy": 2.9591097831726074,
      "epoch": 0.3348148148148148,
      "grad_norm": 1.0762364864349365,
      "learning_rate": 6.786786786786788e-06,
      "loss": 0.20261330902576447,
      "mean_token_accuracy": 0.9430303275585175,
      "num_tokens": 1851392.0,
      "step": 113
    },
    {
      "entropy": 2.8580664098262787,
      "epoch": 0.3377777777777778,
      "grad_norm": 1.0885573625564575,
      "learning_rate": 6.7567567567567575e-06,
      "loss": 0.2401440441608429,
      "mean_token_accuracy": 0.9320252984762192,
      "num_tokens": 1867776.0,
      "step": 114
    },
    {
      "entropy": 3.0035135447978973,
      "epoch": 0.34074074074074073,
      "grad_norm": 1.2362713813781738,
      "learning_rate": 6.726726726726728e-06,
      "loss": 0.25023746490478516,
      "mean_token_accuracy": 0.9219900891184807,
      "num_tokens": 1884160.0,
      "step": 115
    },
    {
      "entropy": 2.792519822716713,
      "epoch": 0.3437037037037037,
      "grad_norm": 0.8280065655708313,
      "learning_rate": 6.696696696696697e-06,
      "loss": 0.15134726464748383,
      "mean_token_accuracy": 0.9595373570919037,
      "num_tokens": 1900544.0,
      "step": 116
    },
    {
      "entropy": 2.9771476686000824,
      "epoch": 0.3466666666666667,
      "grad_norm": 1.0389312505722046,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.1933242678642273,
      "mean_token_accuracy": 0.9470393136143684,
      "num_tokens": 1916928.0,
      "step": 117
    },
    {
      "entropy": 2.6197174042463303,
      "epoch": 0.3496296296296296,
      "grad_norm": 0.9474180340766907,
      "learning_rate": 6.636636636636637e-06,
      "loss": 0.2134747952222824,
      "mean_token_accuracy": 0.9342445358633995,
      "num_tokens": 1933312.0,
      "step": 118
    },
    {
      "entropy": 3.18684458732605,
      "epoch": 0.35259259259259257,
      "grad_norm": 1.1820570230484009,
      "learning_rate": 6.606606606606607e-06,
      "loss": 0.2462218701839447,
      "mean_token_accuracy": 0.9311666861176491,
      "num_tokens": 1949696.0,
      "step": 119
    },
    {
      "entropy": 2.737267106771469,
      "epoch": 0.35555555555555557,
      "grad_norm": 0.9350874423980713,
      "learning_rate": 6.5765765765765775e-06,
      "loss": 0.1752600371837616,
      "mean_token_accuracy": 0.941002868115902,
      "num_tokens": 1966080.0,
      "step": 120
    },
    {
      "entropy": 3.251073867082596,
      "epoch": 0.3585185185185185,
      "grad_norm": 1.0711486339569092,
      "learning_rate": 6.546546546546547e-06,
      "loss": 0.200066477060318,
      "mean_token_accuracy": 0.9415752366185188,
      "num_tokens": 1982464.0,
      "step": 121
    },
    {
      "entropy": 2.8459617495536804,
      "epoch": 0.36148148148148146,
      "grad_norm": 0.9384996891021729,
      "learning_rate": 6.516516516516517e-06,
      "loss": 0.13886728882789612,
      "mean_token_accuracy": 0.9644250869750977,
      "num_tokens": 1998848.0,
      "step": 122
    },
    {
      "entropy": 2.9354823231697083,
      "epoch": 0.36444444444444446,
      "grad_norm": 0.9482925534248352,
      "learning_rate": 6.486486486486487e-06,
      "loss": 0.18203622102737427,
      "mean_token_accuracy": 0.944021001458168,
      "num_tokens": 2015232.0,
      "step": 123
    },
    {
      "entropy": 2.7248374968767166,
      "epoch": 0.3674074074074074,
      "grad_norm": 0.7138831615447998,
      "learning_rate": 6.456456456456457e-06,
      "loss": 0.12275013327598572,
      "mean_token_accuracy": 0.9598664492368698,
      "num_tokens": 2031616.0,
      "step": 124
    },
    {
      "entropy": 2.5504330098629,
      "epoch": 0.37037037037037035,
      "grad_norm": 0.7979923486709595,
      "learning_rate": 6.426426426426427e-06,
      "loss": 0.1486181914806366,
      "mean_token_accuracy": 0.9559812396764755,
      "num_tokens": 2048000.0,
      "step": 125
    },
    {
      "entropy": 2.960262417793274,
      "epoch": 0.37333333333333335,
      "grad_norm": 0.9023007750511169,
      "learning_rate": 6.396396396396397e-06,
      "loss": 0.14857268333435059,
      "mean_token_accuracy": 0.9502998143434525,
      "num_tokens": 2064384.0,
      "step": 126
    },
    {
      "entropy": 3.066712349653244,
      "epoch": 0.3762962962962963,
      "grad_norm": 1.2226768732070923,
      "learning_rate": 6.366366366366366e-06,
      "loss": 0.22188934683799744,
      "mean_token_accuracy": 0.9333562925457954,
      "num_tokens": 2080768.0,
      "step": 127
    },
    {
      "entropy": 2.8543275594711304,
      "epoch": 0.37925925925925924,
      "grad_norm": 1.0072193145751953,
      "learning_rate": 6.336336336336338e-06,
      "loss": 0.16556726396083832,
      "mean_token_accuracy": 0.9592267498373985,
      "num_tokens": 2097152.0,
      "step": 128
    },
    {
      "entropy": 2.787966251373291,
      "epoch": 0.38222222222222224,
      "grad_norm": 0.9507020711898804,
      "learning_rate": 6.3063063063063065e-06,
      "loss": 0.2076268494129181,
      "mean_token_accuracy": 0.9394065663218498,
      "num_tokens": 2113536.0,
      "step": 129
    },
    {
      "entropy": 3.078700691461563,
      "epoch": 0.3851851851851852,
      "grad_norm": 1.1816531419754028,
      "learning_rate": 6.276276276276276e-06,
      "loss": 0.2084028422832489,
      "mean_token_accuracy": 0.9474279880523682,
      "num_tokens": 2129920.0,
      "step": 130
    },
    {
      "entropy": 2.738174542784691,
      "epoch": 0.38814814814814813,
      "grad_norm": 1.0200042724609375,
      "learning_rate": 6.246246246246247e-06,
      "loss": 0.2508554458618164,
      "mean_token_accuracy": 0.9188810512423515,
      "num_tokens": 2146304.0,
      "step": 131
    },
    {
      "entropy": 2.9517208635807037,
      "epoch": 0.39111111111111113,
      "grad_norm": 1.1556061506271362,
      "learning_rate": 6.2162162162162164e-06,
      "loss": 0.203628271818161,
      "mean_token_accuracy": 0.9529605433344841,
      "num_tokens": 2162688.0,
      "step": 132
    },
    {
      "entropy": 2.6178929209709167,
      "epoch": 0.3940740740740741,
      "grad_norm": 0.8564301133155823,
      "learning_rate": 6.186186186186187e-06,
      "loss": 0.15471211075782776,
      "mean_token_accuracy": 0.9534137547016144,
      "num_tokens": 2179072.0,
      "step": 133
    },
    {
      "entropy": 3.0443992614746094,
      "epoch": 0.397037037037037,
      "grad_norm": 1.066476583480835,
      "learning_rate": 6.156156156156157e-06,
      "loss": 0.18798089027404785,
      "mean_token_accuracy": 0.9450087919831276,
      "num_tokens": 2195456.0,
      "step": 134
    },
    {
      "entropy": 2.85013285279274,
      "epoch": 0.4,
      "grad_norm": 1.095873236656189,
      "learning_rate": 6.126126126126126e-06,
      "loss": 0.21489809453487396,
      "mean_token_accuracy": 0.9296712279319763,
      "num_tokens": 2211840.0,
      "step": 135
    },
    {
      "entropy": 2.9270103573799133,
      "epoch": 0.40296296296296297,
      "grad_norm": 0.9951763153076172,
      "learning_rate": 6.096096096096097e-06,
      "loss": 0.18810951709747314,
      "mean_token_accuracy": 0.9445398077368736,
      "num_tokens": 2228224.0,
      "step": 136
    },
    {
      "entropy": 2.7944080233573914,
      "epoch": 0.4059259259259259,
      "grad_norm": 1.0390986204147339,
      "learning_rate": 6.066066066066067e-06,
      "loss": 0.18749360740184784,
      "mean_token_accuracy": 0.952758900821209,
      "num_tokens": 2244608.0,
      "step": 137
    },
    {
      "entropy": 3.1080731749534607,
      "epoch": 0.4088888888888889,
      "grad_norm": 1.1147621870040894,
      "learning_rate": 6.036036036036037e-06,
      "loss": 0.1984114646911621,
      "mean_token_accuracy": 0.9459464251995087,
      "num_tokens": 2260992.0,
      "step": 138
    },
    {
      "entropy": 2.874028980731964,
      "epoch": 0.41185185185185186,
      "grad_norm": 1.1710313558578491,
      "learning_rate": 6.006006006006007e-06,
      "loss": 0.2441575825214386,
      "mean_token_accuracy": 0.9331995025277138,
      "num_tokens": 2277376.0,
      "step": 139
    },
    {
      "entropy": 2.7302171885967255,
      "epoch": 0.4148148148148148,
      "grad_norm": 0.9056249260902405,
      "learning_rate": 5.975975975975976e-06,
      "loss": 0.15125545859336853,
      "mean_token_accuracy": 0.959392860531807,
      "num_tokens": 2293760.0,
      "step": 140
    },
    {
      "entropy": 2.8780124485492706,
      "epoch": 0.4177777777777778,
      "grad_norm": 0.9433802366256714,
      "learning_rate": 5.945945945945947e-06,
      "loss": 0.14991971850395203,
      "mean_token_accuracy": 0.9578480198979378,
      "num_tokens": 2310144.0,
      "step": 141
    },
    {
      "entropy": 2.8998009264469147,
      "epoch": 0.42074074074074075,
      "grad_norm": 1.197164535522461,
      "learning_rate": 5.915915915915916e-06,
      "loss": 0.24937231838703156,
      "mean_token_accuracy": 0.9226873740553856,
      "num_tokens": 2326528.0,
      "step": 142
    },
    {
      "entropy": 2.951385796070099,
      "epoch": 0.4237037037037037,
      "grad_norm": 0.9769899249076843,
      "learning_rate": 5.885885885885886e-06,
      "loss": 0.14887379109859467,
      "mean_token_accuracy": 0.9541207551956177,
      "num_tokens": 2342912.0,
      "step": 143
    },
    {
      "entropy": 2.766965001821518,
      "epoch": 0.4266666666666667,
      "grad_norm": 1.288712739944458,
      "learning_rate": 5.855855855855856e-06,
      "loss": 0.2876666784286499,
      "mean_token_accuracy": 0.9222293272614479,
      "num_tokens": 2359296.0,
      "step": 144
    },
    {
      "entropy": 2.9683837443590164,
      "epoch": 0.42962962962962964,
      "grad_norm": 1.2986119985580444,
      "learning_rate": 5.825825825825826e-06,
      "loss": 0.2985563278198242,
      "mean_token_accuracy": 0.9242896735668182,
      "num_tokens": 2375680.0,
      "step": 145
    },
    {
      "entropy": 2.5206188559532166,
      "epoch": 0.4325925925925926,
      "grad_norm": 0.7270355224609375,
      "learning_rate": 5.7957957957957965e-06,
      "loss": 0.09969068318605423,
      "mean_token_accuracy": 0.9722582995891571,
      "num_tokens": 2392064.0,
      "step": 146
    },
    {
      "entropy": 2.7732113003730774,
      "epoch": 0.43555555555555553,
      "grad_norm": 1.0483299493789673,
      "learning_rate": 5.765765765765766e-06,
      "loss": 0.2085704654455185,
      "mean_token_accuracy": 0.9264028370380402,
      "num_tokens": 2408448.0,
      "step": 147
    },
    {
      "entropy": 2.611909866333008,
      "epoch": 0.43851851851851853,
      "grad_norm": 0.9123360514640808,
      "learning_rate": 5.735735735735736e-06,
      "loss": 0.18682576715946198,
      "mean_token_accuracy": 0.94195157289505,
      "num_tokens": 2424832.0,
      "step": 148
    },
    {
      "entropy": 2.851868659257889,
      "epoch": 0.4414814814814815,
      "grad_norm": 0.8913617730140686,
      "learning_rate": 5.7057057057057065e-06,
      "loss": 0.15075425803661346,
      "mean_token_accuracy": 0.9590912833809853,
      "num_tokens": 2441216.0,
      "step": 149
    },
    {
      "entropy": 2.83494171500206,
      "epoch": 0.4444444444444444,
      "grad_norm": 0.9222665429115295,
      "learning_rate": 5.675675675675676e-06,
      "loss": 0.13384118676185608,
      "mean_token_accuracy": 0.958981528878212,
      "num_tokens": 2457600.0,
      "step": 150
    },
    {
      "epoch": 0.4444444444444444,
      "eval_entropy": 2.8018603912989297,
      "eval_loss": 0.19661802053451538,
      "eval_mean_token_accuracy": 0.9459546693166097,
      "eval_num_tokens": 2457600.0,
      "eval_runtime": 46.0405,
      "eval_samples_per_second": 13.032,
      "eval_steps_per_second": 1.629,
      "step": 150
    },
    {
      "entropy": 2.8351137042045593,
      "epoch": 0.4474074074074074,
      "grad_norm": 0.99371737241745,
      "learning_rate": 5.645645645645647e-06,
      "loss": 0.16289640963077545,
      "mean_token_accuracy": 0.9574097767472267,
      "num_tokens": 2473984.0,
      "step": 151
    },
    {
      "entropy": 2.9848466515541077,
      "epoch": 0.45037037037037037,
      "grad_norm": 1.0490005016326904,
      "learning_rate": 5.615615615615616e-06,
      "loss": 0.1935531347990036,
      "mean_token_accuracy": 0.9483331888914108,
      "num_tokens": 2490368.0,
      "step": 152
    },
    {
      "entropy": 2.6824145317077637,
      "epoch": 0.4533333333333333,
      "grad_norm": 0.9250248074531555,
      "learning_rate": 5.585585585585585e-06,
      "loss": 0.15339694917201996,
      "mean_token_accuracy": 0.9620754644274712,
      "num_tokens": 2506752.0,
      "step": 153
    },
    {
      "entropy": 2.6512254178524017,
      "epoch": 0.4562962962962963,
      "grad_norm": 0.7534433007240295,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.10185273736715317,
      "mean_token_accuracy": 0.9696594327688217,
      "num_tokens": 2523136.0,
      "step": 154
    },
    {
      "entropy": 2.948331445455551,
      "epoch": 0.45925925925925926,
      "grad_norm": 1.1155580282211304,
      "learning_rate": 5.5255255255255255e-06,
      "loss": 0.19951286911964417,
      "mean_token_accuracy": 0.9360602721571922,
      "num_tokens": 2539520.0,
      "step": 155
    },
    {
      "entropy": 3.118526369333267,
      "epoch": 0.4622222222222222,
      "grad_norm": 1.1425875425338745,
      "learning_rate": 5.495495495495496e-06,
      "loss": 0.1999339461326599,
      "mean_token_accuracy": 0.9489343166351318,
      "num_tokens": 2555904.0,
      "step": 156
    },
    {
      "entropy": 2.866454780101776,
      "epoch": 0.4651851851851852,
      "grad_norm": 1.0070029497146606,
      "learning_rate": 5.465465465465466e-06,
      "loss": 0.1702185422182083,
      "mean_token_accuracy": 0.9476803690195084,
      "num_tokens": 2572288.0,
      "step": 157
    },
    {
      "entropy": 2.8745312988758087,
      "epoch": 0.46814814814814815,
      "grad_norm": 1.107635498046875,
      "learning_rate": 5.4354354354354355e-06,
      "loss": 0.21794094145298004,
      "mean_token_accuracy": 0.9443532079458237,
      "num_tokens": 2588672.0,
      "step": 158
    },
    {
      "entropy": 2.971309572458267,
      "epoch": 0.4711111111111111,
      "grad_norm": 0.7945137023925781,
      "learning_rate": 5.405405405405406e-06,
      "loss": 0.12246584892272949,
      "mean_token_accuracy": 0.9603362455964088,
      "num_tokens": 2605056.0,
      "step": 159
    },
    {
      "entropy": 3.0638364255428314,
      "epoch": 0.4740740740740741,
      "grad_norm": 1.3628849983215332,
      "learning_rate": 5.375375375375376e-06,
      "loss": 0.303911417722702,
      "mean_token_accuracy": 0.9074493572115898,
      "num_tokens": 2621440.0,
      "step": 160
    },
    {
      "entropy": 2.7165317684412003,
      "epoch": 0.47703703703703704,
      "grad_norm": 0.9763112664222717,
      "learning_rate": 5.345345345345346e-06,
      "loss": 0.20524877309799194,
      "mean_token_accuracy": 0.9534239247441292,
      "num_tokens": 2637824.0,
      "step": 161
    },
    {
      "entropy": 2.697300225496292,
      "epoch": 0.48,
      "grad_norm": 0.9978896379470825,
      "learning_rate": 5.315315315315316e-06,
      "loss": 0.15991677343845367,
      "mean_token_accuracy": 0.9553363919258118,
      "num_tokens": 2654208.0,
      "step": 162
    },
    {
      "entropy": 3.078951895236969,
      "epoch": 0.482962962962963,
      "grad_norm": 1.002069115638733,
      "learning_rate": 5.285285285285286e-06,
      "loss": 0.1946626603603363,
      "mean_token_accuracy": 0.9440996497869492,
      "num_tokens": 2670592.0,
      "step": 163
    },
    {
      "entropy": 3.1597962975502014,
      "epoch": 0.48592592592592593,
      "grad_norm": 1.2001583576202393,
      "learning_rate": 5.255255255255256e-06,
      "loss": 0.2090940624475479,
      "mean_token_accuracy": 0.9325530007481575,
      "num_tokens": 2686976.0,
      "step": 164
    },
    {
      "entropy": 3.025523602962494,
      "epoch": 0.4888888888888889,
      "grad_norm": 0.8787401914596558,
      "learning_rate": 5.225225225225226e-06,
      "loss": 0.12661801278591156,
      "mean_token_accuracy": 0.9651984125375748,
      "num_tokens": 2703360.0,
      "step": 165
    },
    {
      "entropy": 3.077287882566452,
      "epoch": 0.4918518518518519,
      "grad_norm": 1.0118409395217896,
      "learning_rate": 5.195195195195195e-06,
      "loss": 0.1550823152065277,
      "mean_token_accuracy": 0.9478819221258163,
      "num_tokens": 2719744.0,
      "step": 166
    },
    {
      "entropy": 2.828215479850769,
      "epoch": 0.4948148148148148,
      "grad_norm": 0.9409404397010803,
      "learning_rate": 5.165165165165165e-06,
      "loss": 0.15371187031269073,
      "mean_token_accuracy": 0.9487152621150017,
      "num_tokens": 2736128.0,
      "step": 167
    },
    {
      "entropy": 2.451965168118477,
      "epoch": 0.49777777777777776,
      "grad_norm": 0.6742821931838989,
      "learning_rate": 5.135135135135135e-06,
      "loss": 0.08776291459798813,
      "mean_token_accuracy": 0.9738249182701111,
      "num_tokens": 2752512.0,
      "step": 168
    },
    {
      "entropy": 3.0690193474292755,
      "epoch": 0.5007407407407407,
      "grad_norm": 0.8734592199325562,
      "learning_rate": 5.105105105105106e-06,
      "loss": 0.11133302748203278,
      "mean_token_accuracy": 0.9670532643795013,
      "num_tokens": 2768896.0,
      "step": 169
    },
    {
      "entropy": 2.7598538994789124,
      "epoch": 0.5037037037037037,
      "grad_norm": 1.0697872638702393,
      "learning_rate": 5.075075075075075e-06,
      "loss": 0.20876695215702057,
      "mean_token_accuracy": 0.9453472048044205,
      "num_tokens": 2785280.0,
      "step": 170
    },
    {
      "entropy": 3.2945614755153656,
      "epoch": 0.5066666666666667,
      "grad_norm": 1.2122254371643066,
      "learning_rate": 5.045045045045045e-06,
      "loss": 0.21309486031532288,
      "mean_token_accuracy": 0.9333294779062271,
      "num_tokens": 2801664.0,
      "step": 171
    },
    {
      "entropy": 3.102342188358307,
      "epoch": 0.5096296296296297,
      "grad_norm": 1.2963942289352417,
      "learning_rate": 5.0150150150150156e-06,
      "loss": 0.22489085793495178,
      "mean_token_accuracy": 0.9435132145881653,
      "num_tokens": 2818048.0,
      "step": 172
    },
    {
      "entropy": 3.4107375741004944,
      "epoch": 0.5125925925925926,
      "grad_norm": 1.265325903892517,
      "learning_rate": 4.984984984984985e-06,
      "loss": 0.30620458722114563,
      "mean_token_accuracy": 0.9148801937699318,
      "num_tokens": 2834432.0,
      "step": 173
    },
    {
      "entropy": 2.9490450620651245,
      "epoch": 0.5155555555555555,
      "grad_norm": 1.1190979480743408,
      "learning_rate": 4.954954954954955e-06,
      "loss": 0.16873480379581451,
      "mean_token_accuracy": 0.9528811648488045,
      "num_tokens": 2850816.0,
      "step": 174
    },
    {
      "entropy": 2.8443517684936523,
      "epoch": 0.5185185185185185,
      "grad_norm": 0.9175981879234314,
      "learning_rate": 4.9249249249249255e-06,
      "loss": 0.17586427927017212,
      "mean_token_accuracy": 0.9409847855567932,
      "num_tokens": 2867200.0,
      "step": 175
    },
    {
      "entropy": 2.656998038291931,
      "epoch": 0.5214814814814814,
      "grad_norm": 0.9134055972099304,
      "learning_rate": 4.894894894894895e-06,
      "loss": 0.1511061191558838,
      "mean_token_accuracy": 0.9635813310742378,
      "num_tokens": 2883584.0,
      "step": 176
    },
    {
      "entropy": 2.8948962688446045,
      "epoch": 0.5244444444444445,
      "grad_norm": 0.9751296043395996,
      "learning_rate": 4.864864864864866e-06,
      "loss": 0.13517218828201294,
      "mean_token_accuracy": 0.9621190130710602,
      "num_tokens": 2899968.0,
      "step": 177
    },
    {
      "entropy": 2.9210911095142365,
      "epoch": 0.5274074074074074,
      "grad_norm": 0.9840187430381775,
      "learning_rate": 4.8348348348348355e-06,
      "loss": 0.18799883127212524,
      "mean_token_accuracy": 0.9483400359749794,
      "num_tokens": 2916352.0,
      "step": 178
    },
    {
      "entropy": 2.669636607170105,
      "epoch": 0.5303703703703704,
      "grad_norm": 0.9172318577766418,
      "learning_rate": 4.804804804804805e-06,
      "loss": 0.1985693871974945,
      "mean_token_accuracy": 0.944894015789032,
      "num_tokens": 2932736.0,
      "step": 179
    },
    {
      "entropy": 3.374725103378296,
      "epoch": 0.5333333333333333,
      "grad_norm": 1.4646100997924805,
      "learning_rate": 4.774774774774775e-06,
      "loss": 0.2822015881538391,
      "mean_token_accuracy": 0.9252517595887184,
      "num_tokens": 2949120.0,
      "step": 180
    },
    {
      "entropy": 2.7206821590662003,
      "epoch": 0.5362962962962963,
      "grad_norm": 0.7133533954620361,
      "learning_rate": 4.7447447447447454e-06,
      "loss": 0.10333085060119629,
      "mean_token_accuracy": 0.9712485820055008,
      "num_tokens": 2965504.0,
      "step": 181
    },
    {
      "entropy": 2.6812227815389633,
      "epoch": 0.5392592592592592,
      "grad_norm": 0.969178318977356,
      "learning_rate": 4.714714714714715e-06,
      "loss": 0.1855068802833557,
      "mean_token_accuracy": 0.9543507844209671,
      "num_tokens": 2981888.0,
      "step": 182
    },
    {
      "entropy": 3.1652239859104156,
      "epoch": 0.5422222222222223,
      "grad_norm": 1.1464864015579224,
      "learning_rate": 4.684684684684685e-06,
      "loss": 0.10449449717998505,
      "mean_token_accuracy": 0.9613668546080589,
      "num_tokens": 2998272.0,
      "step": 183
    },
    {
      "entropy": 3.170502334833145,
      "epoch": 0.5451851851851852,
      "grad_norm": 1.2865345478057861,
      "learning_rate": 4.654654654654655e-06,
      "loss": 0.300717830657959,
      "mean_token_accuracy": 0.9175125136971474,
      "num_tokens": 3014656.0,
      "step": 184
    },
    {
      "entropy": 2.9928721487522125,
      "epoch": 0.5481481481481482,
      "grad_norm": 0.9857022762298584,
      "learning_rate": 4.624624624624625e-06,
      "loss": 0.1558627486228943,
      "mean_token_accuracy": 0.9588551372289658,
      "num_tokens": 3031040.0,
      "step": 185
    },
    {
      "entropy": 2.838076591491699,
      "epoch": 0.5511111111111111,
      "grad_norm": 0.6746611595153809,
      "learning_rate": 4.594594594594596e-06,
      "loss": 0.07870074361562729,
      "mean_token_accuracy": 0.9770568087697029,
      "num_tokens": 3047424.0,
      "step": 186
    },
    {
      "entropy": 2.83999440073967,
      "epoch": 0.554074074074074,
      "grad_norm": 0.8938256502151489,
      "learning_rate": 4.5645645645645645e-06,
      "loss": 0.14536425471305847,
      "mean_token_accuracy": 0.9508580788969994,
      "num_tokens": 3063808.0,
      "step": 187
    },
    {
      "entropy": 2.8258763253688812,
      "epoch": 0.557037037037037,
      "grad_norm": 1.0617492198944092,
      "learning_rate": 4.534534534534535e-06,
      "loss": 0.1897096037864685,
      "mean_token_accuracy": 0.9497568681836128,
      "num_tokens": 3080192.0,
      "step": 188
    },
    {
      "entropy": 3.0183293521404266,
      "epoch": 0.56,
      "grad_norm": 1.0675630569458008,
      "learning_rate": 4.504504504504505e-06,
      "loss": 0.21500572562217712,
      "mean_token_accuracy": 0.9347843676805496,
      "num_tokens": 3096576.0,
      "step": 189
    },
    {
      "entropy": 2.857665240764618,
      "epoch": 0.562962962962963,
      "grad_norm": 0.9390116930007935,
      "learning_rate": 4.474474474474475e-06,
      "loss": 0.15814648568630219,
      "mean_token_accuracy": 0.9588152691721916,
      "num_tokens": 3112960.0,
      "step": 190
    },
    {
      "entropy": 2.9760123789310455,
      "epoch": 0.5659259259259259,
      "grad_norm": 0.6752185821533203,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.08213961869478226,
      "mean_token_accuracy": 0.9749421551823616,
      "num_tokens": 3129344.0,
      "step": 191
    },
    {
      "entropy": 3.3175204694271088,
      "epoch": 0.5688888888888889,
      "grad_norm": 1.3816381692886353,
      "learning_rate": 4.414414414414415e-06,
      "loss": 0.20562152564525604,
      "mean_token_accuracy": 0.9325385615229607,
      "num_tokens": 3145728.0,
      "step": 192
    },
    {
      "entropy": 3.2857537269592285,
      "epoch": 0.5718518518518518,
      "grad_norm": 1.4988305568695068,
      "learning_rate": 4.384384384384384e-06,
      "loss": 0.29960426688194275,
      "mean_token_accuracy": 0.911498174071312,
      "num_tokens": 3162112.0,
      "step": 193
    },
    {
      "entropy": 2.84012833237648,
      "epoch": 0.5748148148148148,
      "grad_norm": 1.0266245603561401,
      "learning_rate": 4.354354354354355e-06,
      "loss": 0.1969761699438095,
      "mean_token_accuracy": 0.9453105926513672,
      "num_tokens": 3178496.0,
      "step": 194
    },
    {
      "entropy": 2.562734991312027,
      "epoch": 0.5777777777777777,
      "grad_norm": 0.8547648787498474,
      "learning_rate": 4.324324324324325e-06,
      "loss": 0.13799740374088287,
      "mean_token_accuracy": 0.9638612940907478,
      "num_tokens": 3194880.0,
      "step": 195
    },
    {
      "entropy": 3.051138758659363,
      "epoch": 0.5807407407407408,
      "grad_norm": 1.0234289169311523,
      "learning_rate": 4.294294294294294e-06,
      "loss": 0.19907894730567932,
      "mean_token_accuracy": 0.9392252638936043,
      "num_tokens": 3211264.0,
      "step": 196
    },
    {
      "entropy": 2.978795439004898,
      "epoch": 0.5837037037037037,
      "grad_norm": 1.147609829902649,
      "learning_rate": 4.264264264264265e-06,
      "loss": 0.1958506554365158,
      "mean_token_accuracy": 0.9412998706102371,
      "num_tokens": 3227648.0,
      "step": 197
    },
    {
      "entropy": 2.717726945877075,
      "epoch": 0.5866666666666667,
      "grad_norm": 1.1399898529052734,
      "learning_rate": 4.234234234234235e-06,
      "loss": 0.17526914179325104,
      "mean_token_accuracy": 0.9513465017080307,
      "num_tokens": 3244032.0,
      "step": 198
    },
    {
      "entropy": 2.735936999320984,
      "epoch": 0.5896296296296296,
      "grad_norm": 0.9563067555427551,
      "learning_rate": 4.204204204204204e-06,
      "loss": 0.16168488562107086,
      "mean_token_accuracy": 0.9503465443849564,
      "num_tokens": 3260416.0,
      "step": 199
    },
    {
      "entropy": 2.786933973431587,
      "epoch": 0.5925925925925926,
      "grad_norm": 1.2610732316970825,
      "learning_rate": 4.174174174174174e-06,
      "loss": 0.20555342733860016,
      "mean_token_accuracy": 0.9323367550969124,
      "num_tokens": 3276800.0,
      "step": 200
    },
    {
      "epoch": 0.5925925925925926,
      "eval_entropy": 2.843206590016683,
      "eval_loss": 0.18811877071857452,
      "eval_mean_token_accuracy": 0.9479140742619833,
      "eval_num_tokens": 3276800.0,
      "eval_runtime": 46.0589,
      "eval_samples_per_second": 13.027,
      "eval_steps_per_second": 1.628,
      "step": 200
    },
    {
      "entropy": 2.8376754820346832,
      "epoch": 0.5955555555555555,
      "grad_norm": 1.1410433053970337,
      "learning_rate": 4.1441441441441446e-06,
      "loss": 0.2528444230556488,
      "mean_token_accuracy": 0.9285106882452965,
      "num_tokens": 3293184.0,
      "step": 201
    },
    {
      "entropy": 2.8771266043186188,
      "epoch": 0.5985185185185186,
      "grad_norm": 0.9486339688301086,
      "learning_rate": 4.114114114114114e-06,
      "loss": 0.17324446141719818,
      "mean_token_accuracy": 0.9460131004452705,
      "num_tokens": 3309568.0,
      "step": 202
    },
    {
      "entropy": 2.6368781328201294,
      "epoch": 0.6014814814814815,
      "grad_norm": 1.042582392692566,
      "learning_rate": 4.084084084084085e-06,
      "loss": 0.19229085743427277,
      "mean_token_accuracy": 0.9515736997127533,
      "num_tokens": 3325952.0,
      "step": 203
    },
    {
      "entropy": 3.000798851251602,
      "epoch": 0.6044444444444445,
      "grad_norm": 1.2937592267990112,
      "learning_rate": 4.0540540540540545e-06,
      "loss": 0.30858540534973145,
      "mean_token_accuracy": 0.9139964729547501,
      "num_tokens": 3342336.0,
      "step": 204
    },
    {
      "entropy": 2.483848661184311,
      "epoch": 0.6074074074074074,
      "grad_norm": 0.8479737639427185,
      "learning_rate": 4.024024024024024e-06,
      "loss": 0.1725257933139801,
      "mean_token_accuracy": 0.9434438943862915,
      "num_tokens": 3358720.0,
      "step": 205
    },
    {
      "entropy": 2.5683358758687973,
      "epoch": 0.6103703703703703,
      "grad_norm": 0.9113346338272095,
      "learning_rate": 3.993993993993994e-06,
      "loss": 0.1125510185956955,
      "mean_token_accuracy": 0.9675316885113716,
      "num_tokens": 3375104.0,
      "step": 206
    },
    {
      "entropy": 2.908651977777481,
      "epoch": 0.6133333333333333,
      "grad_norm": 1.0864291191101074,
      "learning_rate": 3.9639639639639645e-06,
      "loss": 0.23676928877830505,
      "mean_token_accuracy": 0.9158507436513901,
      "num_tokens": 3391488.0,
      "step": 207
    },
    {
      "entropy": 3.02653369307518,
      "epoch": 0.6162962962962963,
      "grad_norm": 1.1199774742126465,
      "learning_rate": 3.933933933933934e-06,
      "loss": 0.25189208984375,
      "mean_token_accuracy": 0.9315646514296532,
      "num_tokens": 3407872.0,
      "step": 208
    },
    {
      "entropy": 2.872484266757965,
      "epoch": 0.6192592592592593,
      "grad_norm": 1.0150285959243774,
      "learning_rate": 3.903903903903904e-06,
      "loss": 0.21484199166297913,
      "mean_token_accuracy": 0.9420194998383522,
      "num_tokens": 3424256.0,
      "step": 209
    },
    {
      "entropy": 3.188303828239441,
      "epoch": 0.6222222222222222,
      "grad_norm": 0.966525137424469,
      "learning_rate": 3.8738738738738744e-06,
      "loss": 0.1408887803554535,
      "mean_token_accuracy": 0.9526339918375015,
      "num_tokens": 3440640.0,
      "step": 210
    },
    {
      "entropy": 2.522002376616001,
      "epoch": 0.6251851851851852,
      "grad_norm": 1.1654863357543945,
      "learning_rate": 3.843843843843844e-06,
      "loss": 0.15462645888328552,
      "mean_token_accuracy": 0.9606514051556587,
      "num_tokens": 3457024.0,
      "step": 211
    },
    {
      "entropy": 3.116719514131546,
      "epoch": 0.6281481481481481,
      "grad_norm": 1.0795726776123047,
      "learning_rate": 3.8138138138138143e-06,
      "loss": 0.17554046213626862,
      "mean_token_accuracy": 0.9533887356519699,
      "num_tokens": 3473408.0,
      "step": 212
    },
    {
      "entropy": 3.040469765663147,
      "epoch": 0.6311111111111111,
      "grad_norm": 1.3026610612869263,
      "learning_rate": 3.7837837837837844e-06,
      "loss": 0.2474716305732727,
      "mean_token_accuracy": 0.9333793073892593,
      "num_tokens": 3489792.0,
      "step": 213
    },
    {
      "entropy": 3.274827539920807,
      "epoch": 0.6340740740740741,
      "grad_norm": 1.3957505226135254,
      "learning_rate": 3.7537537537537537e-06,
      "loss": 0.24205292761325836,
      "mean_token_accuracy": 0.9245704934000969,
      "num_tokens": 3506176.0,
      "step": 214
    },
    {
      "entropy": 3.178326040506363,
      "epoch": 0.6370370370370371,
      "grad_norm": 1.216489315032959,
      "learning_rate": 3.723723723723724e-06,
      "loss": 0.28391730785369873,
      "mean_token_accuracy": 0.9203849658370018,
      "num_tokens": 3522560.0,
      "step": 215
    },
    {
      "entropy": 3.2481323778629303,
      "epoch": 0.64,
      "grad_norm": 1.231285572052002,
      "learning_rate": 3.693693693693694e-06,
      "loss": 0.26016050577163696,
      "mean_token_accuracy": 0.9341835081577301,
      "num_tokens": 3538944.0,
      "step": 216
    },
    {
      "entropy": 3.3229857087135315,
      "epoch": 0.642962962962963,
      "grad_norm": 1.1100609302520752,
      "learning_rate": 3.663663663663664e-06,
      "loss": 0.21764174103736877,
      "mean_token_accuracy": 0.9426891654729843,
      "num_tokens": 3555328.0,
      "step": 217
    },
    {
      "entropy": 3.274485871195793,
      "epoch": 0.6459259259259259,
      "grad_norm": 1.3515102863311768,
      "learning_rate": 3.633633633633634e-06,
      "loss": 0.25322726368904114,
      "mean_token_accuracy": 0.9083448052406311,
      "num_tokens": 3571712.0,
      "step": 218
    },
    {
      "entropy": 3.1637975871562958,
      "epoch": 0.6488888888888888,
      "grad_norm": 1.1547216176986694,
      "learning_rate": 3.603603603603604e-06,
      "loss": 0.1991884857416153,
      "mean_token_accuracy": 0.9434786066412926,
      "num_tokens": 3588096.0,
      "step": 219
    },
    {
      "entropy": 3.130490243434906,
      "epoch": 0.6518518518518519,
      "grad_norm": 0.9679241180419922,
      "learning_rate": 3.573573573573574e-06,
      "loss": 0.16193579137325287,
      "mean_token_accuracy": 0.9527297466993332,
      "num_tokens": 3604480.0,
      "step": 220
    },
    {
      "entropy": 3.1223119497299194,
      "epoch": 0.6548148148148148,
      "grad_norm": 1.0283665657043457,
      "learning_rate": 3.5435435435435437e-06,
      "loss": 0.16569122672080994,
      "mean_token_accuracy": 0.9500171765685081,
      "num_tokens": 3620864.0,
      "step": 221
    },
    {
      "entropy": 3.634727567434311,
      "epoch": 0.6577777777777778,
      "grad_norm": 1.3865503072738647,
      "learning_rate": 3.513513513513514e-06,
      "loss": 0.25727787613868713,
      "mean_token_accuracy": 0.9221251383423805,
      "num_tokens": 3637248.0,
      "step": 222
    },
    {
      "entropy": 3.1611114740371704,
      "epoch": 0.6607407407407407,
      "grad_norm": 0.9414445161819458,
      "learning_rate": 3.4834834834834835e-06,
      "loss": 0.1781175583600998,
      "mean_token_accuracy": 0.9557042866945267,
      "num_tokens": 3653632.0,
      "step": 223
    },
    {
      "entropy": 3.085616856813431,
      "epoch": 0.6637037037037037,
      "grad_norm": 1.2785733938217163,
      "learning_rate": 3.4534534534534537e-06,
      "loss": 0.21069051325321198,
      "mean_token_accuracy": 0.936344563961029,
      "num_tokens": 3670016.0,
      "step": 224
    },
    {
      "entropy": 3.4318977296352386,
      "epoch": 0.6666666666666666,
      "grad_norm": 1.1817911863327026,
      "learning_rate": 3.423423423423424e-06,
      "loss": 0.18880143761634827,
      "mean_token_accuracy": 0.9423677548766136,
      "num_tokens": 3686400.0,
      "step": 225
    },
    {
      "entropy": 3.318112313747406,
      "epoch": 0.6696296296296296,
      "grad_norm": 1.0572702884674072,
      "learning_rate": 3.393393393393394e-06,
      "loss": 0.190169095993042,
      "mean_token_accuracy": 0.9346237555146217,
      "num_tokens": 3702784.0,
      "step": 226
    },
    {
      "entropy": 2.874858856201172,
      "epoch": 0.6725925925925926,
      "grad_norm": 0.8685035109519958,
      "learning_rate": 3.363363363363364e-06,
      "loss": 0.15599817037582397,
      "mean_token_accuracy": 0.9568697363138199,
      "num_tokens": 3719168.0,
      "step": 227
    },
    {
      "entropy": 3.2247777581214905,
      "epoch": 0.6755555555555556,
      "grad_norm": 1.010103464126587,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.13462236523628235,
      "mean_token_accuracy": 0.958881676197052,
      "num_tokens": 3735552.0,
      "step": 228
    },
    {
      "entropy": 3.294255703687668,
      "epoch": 0.6785185185185185,
      "grad_norm": 1.3059518337249756,
      "learning_rate": 3.3033033033033035e-06,
      "loss": 0.2621678113937378,
      "mean_token_accuracy": 0.9403132051229477,
      "num_tokens": 3751936.0,
      "step": 229
    },
    {
      "entropy": 3.4544302225112915,
      "epoch": 0.6814814814814815,
      "grad_norm": 1.263285517692566,
      "learning_rate": 3.2732732732732736e-06,
      "loss": 0.23261971771717072,
      "mean_token_accuracy": 0.9319297224283218,
      "num_tokens": 3768320.0,
      "step": 230
    },
    {
      "entropy": 2.852003574371338,
      "epoch": 0.6844444444444444,
      "grad_norm": 0.8441635966300964,
      "learning_rate": 3.2432432432432437e-06,
      "loss": 0.12731072306632996,
      "mean_token_accuracy": 0.9589616805315018,
      "num_tokens": 3784704.0,
      "step": 231
    },
    {
      "entropy": 3.351558566093445,
      "epoch": 0.6874074074074074,
      "grad_norm": 1.1281638145446777,
      "learning_rate": 3.2132132132132134e-06,
      "loss": 0.2168779969215393,
      "mean_token_accuracy": 0.9330433830618858,
      "num_tokens": 3801088.0,
      "step": 232
    },
    {
      "entropy": 2.961367964744568,
      "epoch": 0.6903703703703704,
      "grad_norm": 1.3269091844558716,
      "learning_rate": 3.183183183183183e-06,
      "loss": 0.21297143399715424,
      "mean_token_accuracy": 0.9387368410825729,
      "num_tokens": 3817472.0,
      "step": 233
    },
    {
      "entropy": 2.8256981670856476,
      "epoch": 0.6933333333333334,
      "grad_norm": 0.6941968202590942,
      "learning_rate": 3.1531531531531532e-06,
      "loss": 0.08092458546161652,
      "mean_token_accuracy": 0.9731299355626106,
      "num_tokens": 3833856.0,
      "step": 234
    },
    {
      "entropy": 3.1228906214237213,
      "epoch": 0.6962962962962963,
      "grad_norm": 1.0961962938308716,
      "learning_rate": 3.1231231231231234e-06,
      "loss": 0.18822498619556427,
      "mean_token_accuracy": 0.9481627866625786,
      "num_tokens": 3850240.0,
      "step": 235
    },
    {
      "entropy": 2.942607879638672,
      "epoch": 0.6992592592592592,
      "grad_norm": 1.1089555025100708,
      "learning_rate": 3.0930930930930935e-06,
      "loss": 0.18346789479255676,
      "mean_token_accuracy": 0.9427254498004913,
      "num_tokens": 3866624.0,
      "step": 236
    },
    {
      "entropy": 2.7327932566404343,
      "epoch": 0.7022222222222222,
      "grad_norm": 0.9384790658950806,
      "learning_rate": 3.063063063063063e-06,
      "loss": 0.15320709347724915,
      "mean_token_accuracy": 0.9529693126678467,
      "num_tokens": 3883008.0,
      "step": 237
    },
    {
      "entropy": 3.1973710656166077,
      "epoch": 0.7051851851851851,
      "grad_norm": 1.2108032703399658,
      "learning_rate": 3.0330330330330333e-06,
      "loss": 0.2533422112464905,
      "mean_token_accuracy": 0.9376625046133995,
      "num_tokens": 3899392.0,
      "step": 238
    },
    {
      "entropy": 2.988917350769043,
      "epoch": 0.7081481481481482,
      "grad_norm": 1.2388724088668823,
      "learning_rate": 3.0030030030030034e-06,
      "loss": 0.27100202441215515,
      "mean_token_accuracy": 0.9289768561720848,
      "num_tokens": 3915776.0,
      "step": 239
    },
    {
      "entropy": 2.9555850327014923,
      "epoch": 0.7111111111111111,
      "grad_norm": 1.5143336057662964,
      "learning_rate": 2.9729729729729736e-06,
      "loss": 0.3464142382144928,
      "mean_token_accuracy": 0.9164462983608246,
      "num_tokens": 3932160.0,
      "step": 240
    },
    {
      "entropy": 2.8657090961933136,
      "epoch": 0.7140740740740741,
      "grad_norm": 0.8298128247261047,
      "learning_rate": 2.942942942942943e-06,
      "loss": 0.12520447373390198,
      "mean_token_accuracy": 0.9673245996236801,
      "num_tokens": 3948544.0,
      "step": 241
    },
    {
      "entropy": 2.8615731298923492,
      "epoch": 0.717037037037037,
      "grad_norm": 1.2797398567199707,
      "learning_rate": 2.912912912912913e-06,
      "loss": 0.26079538464546204,
      "mean_token_accuracy": 0.9286767542362213,
      "num_tokens": 3964928.0,
      "step": 242
    },
    {
      "entropy": 3.159029483795166,
      "epoch": 0.72,
      "grad_norm": 1.1734626293182373,
      "learning_rate": 2.882882882882883e-06,
      "loss": 0.17298385500907898,
      "mean_token_accuracy": 0.9454841911792755,
      "num_tokens": 3981312.0,
      "step": 243
    },
    {
      "entropy": 2.9439398646354675,
      "epoch": 0.7229629629629629,
      "grad_norm": 1.187084674835205,
      "learning_rate": 2.8528528528528532e-06,
      "loss": 0.20587158203125,
      "mean_token_accuracy": 0.9424489960074425,
      "num_tokens": 3997696.0,
      "step": 244
    },
    {
      "entropy": 2.640188694000244,
      "epoch": 0.725925925925926,
      "grad_norm": 0.8445670008659363,
      "learning_rate": 2.8228228228228234e-06,
      "loss": 0.1405259370803833,
      "mean_token_accuracy": 0.9596375972032547,
      "num_tokens": 4014080.0,
      "step": 245
    },
    {
      "entropy": 3.065976858139038,
      "epoch": 0.7288888888888889,
      "grad_norm": 1.2772263288497925,
      "learning_rate": 2.7927927927927926e-06,
      "loss": 0.3095298111438751,
      "mean_token_accuracy": 0.9271637871861458,
      "num_tokens": 4030464.0,
      "step": 246
    },
    {
      "entropy": 2.9902926087379456,
      "epoch": 0.7318518518518519,
      "grad_norm": 1.0105092525482178,
      "learning_rate": 2.7627627627627628e-06,
      "loss": 0.1472294181585312,
      "mean_token_accuracy": 0.957129955291748,
      "num_tokens": 4046848.0,
      "step": 247
    },
    {
      "entropy": 2.8248405754566193,
      "epoch": 0.7348148148148148,
      "grad_norm": 0.9832578301429749,
      "learning_rate": 2.732732732732733e-06,
      "loss": 0.13547852635383606,
      "mean_token_accuracy": 0.9589240774512291,
      "num_tokens": 4063232.0,
      "step": 248
    },
    {
      "entropy": 3.128548353910446,
      "epoch": 0.7377777777777778,
      "grad_norm": 1.0722336769104004,
      "learning_rate": 2.702702702702703e-06,
      "loss": 0.1782652884721756,
      "mean_token_accuracy": 0.9541125819087029,
      "num_tokens": 4079616.0,
      "step": 249
    },
    {
      "entropy": 3.122624695301056,
      "epoch": 0.7407407407407407,
      "grad_norm": 1.0744754076004028,
      "learning_rate": 2.672672672672673e-06,
      "loss": 0.17159108817577362,
      "mean_token_accuracy": 0.9499552398920059,
      "num_tokens": 4096000.0,
      "step": 250
    },
    {
      "epoch": 0.7407407407407407,
      "eval_entropy": 2.852132587432861,
      "eval_loss": 0.18360987305641174,
      "eval_mean_token_accuracy": 0.9489432787895202,
      "eval_num_tokens": 4096000.0,
      "eval_runtime": 46.0566,
      "eval_samples_per_second": 13.027,
      "eval_steps_per_second": 1.628,
      "step": 250
    },
    {
      "entropy": 3.0413025617599487,
      "epoch": 0.7437037037037038,
      "grad_norm": 1.0630085468292236,
      "learning_rate": 2.642642642642643e-06,
      "loss": 0.1608152836561203,
      "mean_token_accuracy": 0.9527239575982094,
      "num_tokens": 4112384.0,
      "step": 251
    },
    {
      "entropy": 3.0769342482089996,
      "epoch": 0.7466666666666667,
      "grad_norm": 1.155750036239624,
      "learning_rate": 2.612612612612613e-06,
      "loss": 0.2295750081539154,
      "mean_token_accuracy": 0.9408787041902542,
      "num_tokens": 4128768.0,
      "step": 252
    },
    {
      "entropy": 2.544542968273163,
      "epoch": 0.7496296296296296,
      "grad_norm": 0.8062906861305237,
      "learning_rate": 2.5825825825825827e-06,
      "loss": 0.09673324227333069,
      "mean_token_accuracy": 0.9708146378397942,
      "num_tokens": 4145152.0,
      "step": 253
    },
    {
      "entropy": 2.376124307513237,
      "epoch": 0.7525925925925926,
      "grad_norm": 0.8885618448257446,
      "learning_rate": 2.552552552552553e-06,
      "loss": 0.1497853845357895,
      "mean_token_accuracy": 0.9607590362429619,
      "num_tokens": 4161536.0,
      "step": 254
    },
    {
      "entropy": 2.9372702836990356,
      "epoch": 0.7555555555555555,
      "grad_norm": 0.9668951630592346,
      "learning_rate": 2.5225225225225225e-06,
      "loss": 0.16193467378616333,
      "mean_token_accuracy": 0.956824317574501,
      "num_tokens": 4177920.0,
      "step": 255
    },
    {
      "entropy": 3.096995621919632,
      "epoch": 0.7585185185185185,
      "grad_norm": 1.040653109550476,
      "learning_rate": 2.4924924924924926e-06,
      "loss": 0.15727289021015167,
      "mean_token_accuracy": 0.9528902173042297,
      "num_tokens": 4194304.0,
      "step": 256
    },
    {
      "entropy": 2.7445938736200333,
      "epoch": 0.7614814814814815,
      "grad_norm": 0.9870290756225586,
      "learning_rate": 2.4624624624624628e-06,
      "loss": 0.16061754524707794,
      "mean_token_accuracy": 0.944463774561882,
      "num_tokens": 4210688.0,
      "step": 257
    },
    {
      "entropy": 3.0143087804317474,
      "epoch": 0.7644444444444445,
      "grad_norm": 1.063123345375061,
      "learning_rate": 2.432432432432433e-06,
      "loss": 0.23332087695598602,
      "mean_token_accuracy": 0.9388647153973579,
      "num_tokens": 4227072.0,
      "step": 258
    },
    {
      "entropy": 2.296273797750473,
      "epoch": 0.7674074074074074,
      "grad_norm": 0.5959308743476868,
      "learning_rate": 2.4024024024024026e-06,
      "loss": 0.06353914737701416,
      "mean_token_accuracy": 0.975685179233551,
      "num_tokens": 4243456.0,
      "step": 259
    },
    {
      "entropy": 2.560692310333252,
      "epoch": 0.7703703703703704,
      "grad_norm": 0.9101587533950806,
      "learning_rate": 2.3723723723723727e-06,
      "loss": 0.1989908069372177,
      "mean_token_accuracy": 0.9402595013380051,
      "num_tokens": 4259840.0,
      "step": 260
    },
    {
      "entropy": 2.547641634941101,
      "epoch": 0.7733333333333333,
      "grad_norm": 0.9849615693092346,
      "learning_rate": 2.3423423423423424e-06,
      "loss": 0.19354493916034698,
      "mean_token_accuracy": 0.943332850933075,
      "num_tokens": 4276224.0,
      "step": 261
    },
    {
      "entropy": 2.405847728252411,
      "epoch": 0.7762962962962963,
      "grad_norm": 0.7766302824020386,
      "learning_rate": 2.3123123123123125e-06,
      "loss": 0.1316036880016327,
      "mean_token_accuracy": 0.9651554375886917,
      "num_tokens": 4292608.0,
      "step": 262
    },
    {
      "entropy": 2.7922613620758057,
      "epoch": 0.7792592592592592,
      "grad_norm": 1.1916041374206543,
      "learning_rate": 2.2822822822822822e-06,
      "loss": 0.2569754123687744,
      "mean_token_accuracy": 0.9235717058181763,
      "num_tokens": 4308992.0,
      "step": 263
    },
    {
      "entropy": 2.726495712995529,
      "epoch": 0.7822222222222223,
      "grad_norm": 1.3649945259094238,
      "learning_rate": 2.2522522522522524e-06,
      "loss": 0.23288977146148682,
      "mean_token_accuracy": 0.9412301778793335,
      "num_tokens": 4325376.0,
      "step": 264
    },
    {
      "entropy": 3.0609046816825867,
      "epoch": 0.7851851851851852,
      "grad_norm": 0.9855526089668274,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.18245509266853333,
      "mean_token_accuracy": 0.9424236342310905,
      "num_tokens": 4341760.0,
      "step": 265
    },
    {
      "entropy": 2.693583309650421,
      "epoch": 0.7881481481481482,
      "grad_norm": 1.293851613998413,
      "learning_rate": 2.192192192192192e-06,
      "loss": 0.27080366015434265,
      "mean_token_accuracy": 0.9291214868426323,
      "num_tokens": 4358144.0,
      "step": 266
    },
    {
      "entropy": 2.966628283262253,
      "epoch": 0.7911111111111111,
      "grad_norm": 1.4858123064041138,
      "learning_rate": 2.1621621621621623e-06,
      "loss": 0.2997784912586212,
      "mean_token_accuracy": 0.9088621661067009,
      "num_tokens": 4374528.0,
      "step": 267
    },
    {
      "entropy": 2.6539561301469803,
      "epoch": 0.794074074074074,
      "grad_norm": 0.9603286385536194,
      "learning_rate": 2.1321321321321325e-06,
      "loss": 0.15536902844905853,
      "mean_token_accuracy": 0.9468490332365036,
      "num_tokens": 4390912.0,
      "step": 268
    },
    {
      "entropy": 2.7186500430107117,
      "epoch": 0.797037037037037,
      "grad_norm": 1.0671851634979248,
      "learning_rate": 2.102102102102102e-06,
      "loss": 0.16850227117538452,
      "mean_token_accuracy": 0.9490253254771233,
      "num_tokens": 4407296.0,
      "step": 269
    },
    {
      "entropy": 2.662411704659462,
      "epoch": 0.8,
      "grad_norm": 1.1484098434448242,
      "learning_rate": 2.0720720720720723e-06,
      "loss": 0.2518634498119354,
      "mean_token_accuracy": 0.9247348234057426,
      "num_tokens": 4423680.0,
      "step": 270
    },
    {
      "entropy": 2.9073596447706223,
      "epoch": 0.802962962962963,
      "grad_norm": 1.2021108865737915,
      "learning_rate": 2.0420420420420424e-06,
      "loss": 0.21068042516708374,
      "mean_token_accuracy": 0.9385004192590714,
      "num_tokens": 4440064.0,
      "step": 271
    },
    {
      "entropy": 2.9670767188072205,
      "epoch": 0.8059259259259259,
      "grad_norm": 0.999944806098938,
      "learning_rate": 2.012012012012012e-06,
      "loss": 0.16588366031646729,
      "mean_token_accuracy": 0.9474747702479362,
      "num_tokens": 4456448.0,
      "step": 272
    },
    {
      "entropy": 2.8509768545627594,
      "epoch": 0.8088888888888889,
      "grad_norm": 0.8416431546211243,
      "learning_rate": 1.9819819819819822e-06,
      "loss": 0.11432436108589172,
      "mean_token_accuracy": 0.9629718512296677,
      "num_tokens": 4472832.0,
      "step": 273
    },
    {
      "entropy": 2.626408964395523,
      "epoch": 0.8118518518518518,
      "grad_norm": 0.9070788025856018,
      "learning_rate": 1.951951951951952e-06,
      "loss": 0.1523875892162323,
      "mean_token_accuracy": 0.9521723687648773,
      "num_tokens": 4489216.0,
      "step": 274
    },
    {
      "entropy": 2.704342484474182,
      "epoch": 0.8148148148148148,
      "grad_norm": 0.9234112501144409,
      "learning_rate": 1.921921921921922e-06,
      "loss": 0.16524599492549896,
      "mean_token_accuracy": 0.9493722841143608,
      "num_tokens": 4505600.0,
      "step": 275
    },
    {
      "entropy": 2.66214981675148,
      "epoch": 0.8177777777777778,
      "grad_norm": 0.8645657896995544,
      "learning_rate": 1.8918918918918922e-06,
      "loss": 0.15482839941978455,
      "mean_token_accuracy": 0.9581804350018501,
      "num_tokens": 4521984.0,
      "step": 276
    },
    {
      "entropy": 3.3557138442993164,
      "epoch": 0.8207407407407408,
      "grad_norm": 0.8617742657661438,
      "learning_rate": 1.861861861861862e-06,
      "loss": 0.1010599136352539,
      "mean_token_accuracy": 0.9699912890791893,
      "num_tokens": 4538368.0,
      "step": 277
    },
    {
      "entropy": 2.618903413414955,
      "epoch": 0.8237037037037037,
      "grad_norm": 0.9169335961341858,
      "learning_rate": 1.831831831831832e-06,
      "loss": 0.14013750851154327,
      "mean_token_accuracy": 0.9602667838335037,
      "num_tokens": 4554752.0,
      "step": 278
    },
    {
      "entropy": 2.9852834343910217,
      "epoch": 0.8266666666666667,
      "grad_norm": 1.1565862894058228,
      "learning_rate": 1.801801801801802e-06,
      "loss": 0.20524702966213226,
      "mean_token_accuracy": 0.9365398287773132,
      "num_tokens": 4571136.0,
      "step": 279
    },
    {
      "entropy": 2.714282900094986,
      "epoch": 0.8296296296296296,
      "grad_norm": 1.0796194076538086,
      "learning_rate": 1.7717717717717719e-06,
      "loss": 0.18443220853805542,
      "mean_token_accuracy": 0.9407469630241394,
      "num_tokens": 4587520.0,
      "step": 280
    },
    {
      "entropy": 3.0053972601890564,
      "epoch": 0.8325925925925926,
      "grad_norm": 1.146272897720337,
      "learning_rate": 1.7417417417417418e-06,
      "loss": 0.1503574252128601,
      "mean_token_accuracy": 0.9579518958926201,
      "num_tokens": 4603904.0,
      "step": 281
    },
    {
      "entropy": 2.9600902795791626,
      "epoch": 0.8355555555555556,
      "grad_norm": 0.8662426471710205,
      "learning_rate": 1.711711711711712e-06,
      "loss": 0.1463131308555603,
      "mean_token_accuracy": 0.9600259512662888,
      "num_tokens": 4620288.0,
      "step": 282
    },
    {
      "entropy": 2.772203043103218,
      "epoch": 0.8385185185185186,
      "grad_norm": 0.7821983098983765,
      "learning_rate": 1.681681681681682e-06,
      "loss": 0.1397870033979416,
      "mean_token_accuracy": 0.9541897773742676,
      "num_tokens": 4636672.0,
      "step": 283
    },
    {
      "entropy": 2.568053811788559,
      "epoch": 0.8414814814814815,
      "grad_norm": 0.840868353843689,
      "learning_rate": 1.6516516516516517e-06,
      "loss": 0.16644269227981567,
      "mean_token_accuracy": 0.9549445882439613,
      "num_tokens": 4653056.0,
      "step": 284
    },
    {
      "entropy": 2.7312504649162292,
      "epoch": 0.8444444444444444,
      "grad_norm": 1.0227640867233276,
      "learning_rate": 1.6216216216216219e-06,
      "loss": 0.175675630569458,
      "mean_token_accuracy": 0.9505376815795898,
      "num_tokens": 4669440.0,
      "step": 285
    },
    {
      "entropy": 2.7323302924633026,
      "epoch": 0.8474074074074074,
      "grad_norm": 1.228810429573059,
      "learning_rate": 1.5915915915915916e-06,
      "loss": 0.19048836827278137,
      "mean_token_accuracy": 0.9527504071593285,
      "num_tokens": 4685824.0,
      "step": 286
    },
    {
      "entropy": 2.6488165259361267,
      "epoch": 0.8503703703703703,
      "grad_norm": 0.8388625979423523,
      "learning_rate": 1.5615615615615617e-06,
      "loss": 0.13485638797283173,
      "mean_token_accuracy": 0.9630501195788383,
      "num_tokens": 4702208.0,
      "step": 287
    },
    {
      "entropy": 2.64958593249321,
      "epoch": 0.8533333333333334,
      "grad_norm": 0.9264437556266785,
      "learning_rate": 1.5315315315315316e-06,
      "loss": 0.14497125148773193,
      "mean_token_accuracy": 0.9597008600831032,
      "num_tokens": 4718592.0,
      "step": 288
    },
    {
      "entropy": 2.982826054096222,
      "epoch": 0.8562962962962963,
      "grad_norm": 1.162354588508606,
      "learning_rate": 1.5015015015015017e-06,
      "loss": 0.18828660249710083,
      "mean_token_accuracy": 0.9438819661736488,
      "num_tokens": 4734976.0,
      "step": 289
    },
    {
      "entropy": 2.8330233842134476,
      "epoch": 0.8592592592592593,
      "grad_norm": 1.0217313766479492,
      "learning_rate": 1.4714714714714714e-06,
      "loss": 0.18914705514907837,
      "mean_token_accuracy": 0.9470869824290276,
      "num_tokens": 4751360.0,
      "step": 290
    },
    {
      "entropy": 2.883241593837738,
      "epoch": 0.8622222222222222,
      "grad_norm": 1.2666387557983398,
      "learning_rate": 1.4414414414414416e-06,
      "loss": 0.24541325867176056,
      "mean_token_accuracy": 0.9369024261832237,
      "num_tokens": 4767744.0,
      "step": 291
    },
    {
      "entropy": 2.940699949860573,
      "epoch": 0.8651851851851852,
      "grad_norm": 1.0014292001724243,
      "learning_rate": 1.4114114114114117e-06,
      "loss": 0.15544487535953522,
      "mean_token_accuracy": 0.9588250517845154,
      "num_tokens": 4784128.0,
      "step": 292
    },
    {
      "entropy": 3.156528502702713,
      "epoch": 0.8681481481481481,
      "grad_norm": 1.3966485261917114,
      "learning_rate": 1.3813813813813814e-06,
      "loss": 0.19821132719516754,
      "mean_token_accuracy": 0.9371234104037285,
      "num_tokens": 4800512.0,
      "step": 293
    },
    {
      "entropy": 3.10570028424263,
      "epoch": 0.8711111111111111,
      "grad_norm": 1.1574422121047974,
      "learning_rate": 1.3513513513513515e-06,
      "loss": 0.15517574548721313,
      "mean_token_accuracy": 0.9560784548521042,
      "num_tokens": 4816896.0,
      "step": 294
    },
    {
      "entropy": 2.994854688644409,
      "epoch": 0.8740740740740741,
      "grad_norm": 1.259311556816101,
      "learning_rate": 1.3213213213213214e-06,
      "loss": 0.2651704251766205,
      "mean_token_accuracy": 0.9265646189451218,
      "num_tokens": 4833280.0,
      "step": 295
    },
    {
      "entropy": 3.040595531463623,
      "epoch": 0.8770370370370371,
      "grad_norm": 0.9883430004119873,
      "learning_rate": 1.2912912912912913e-06,
      "loss": 0.16328245401382446,
      "mean_token_accuracy": 0.9496760815382004,
      "num_tokens": 4849664.0,
      "step": 296
    },
    {
      "entropy": 2.5329011380672455,
      "epoch": 0.88,
      "grad_norm": 0.6723151206970215,
      "learning_rate": 1.2612612612612613e-06,
      "loss": 0.11081084609031677,
      "mean_token_accuracy": 0.9700026363134384,
      "num_tokens": 4866048.0,
      "step": 297
    },
    {
      "entropy": 2.563674107193947,
      "epoch": 0.882962962962963,
      "grad_norm": 1.0615836381912231,
      "learning_rate": 1.2312312312312314e-06,
      "loss": 0.19232074916362762,
      "mean_token_accuracy": 0.9422744959592819,
      "num_tokens": 4882432.0,
      "step": 298
    },
    {
      "entropy": 2.8897614181041718,
      "epoch": 0.8859259259259259,
      "grad_norm": 0.8623843789100647,
      "learning_rate": 1.2012012012012013e-06,
      "loss": 0.13093861937522888,
      "mean_token_accuracy": 0.9648579210042953,
      "num_tokens": 4898816.0,
      "step": 299
    },
    {
      "entropy": 2.9166508615016937,
      "epoch": 0.8888888888888888,
      "grad_norm": 1.2870267629623413,
      "learning_rate": 1.1711711711711712e-06,
      "loss": 0.22300012409687042,
      "mean_token_accuracy": 0.9281135722994804,
      "num_tokens": 4915200.0,
      "step": 300
    },
    {
      "epoch": 0.8888888888888888,
      "eval_entropy": 2.8255044110616048,
      "eval_loss": 0.18106622993946075,
      "eval_mean_token_accuracy": 0.9497437485059103,
      "eval_num_tokens": 4915200.0,
      "eval_runtime": 46.0454,
      "eval_samples_per_second": 13.031,
      "eval_steps_per_second": 1.629,
      "step": 300
    },
    {
      "entropy": 3.0466695725917816,
      "epoch": 0.8918518518518519,
      "grad_norm": 1.3157471418380737,
      "learning_rate": 1.1411411411411411e-06,
      "loss": 0.3266354501247406,
      "mean_token_accuracy": 0.9144779592752457,
      "num_tokens": 4931584.0,
      "step": 301
    },
    {
      "entropy": 2.7269774675369263,
      "epoch": 0.8948148148148148,
      "grad_norm": 0.8248907923698425,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.14082956314086914,
      "mean_token_accuracy": 0.9545648768544197,
      "num_tokens": 4947968.0,
      "step": 302
    },
    {
      "entropy": 3.25367334485054,
      "epoch": 0.8977777777777778,
      "grad_norm": 1.4216176271438599,
      "learning_rate": 1.0810810810810812e-06,
      "loss": 0.22889986634254456,
      "mean_token_accuracy": 0.9246480539441109,
      "num_tokens": 4964352.0,
      "step": 303
    },
    {
      "entropy": 2.9509867429733276,
      "epoch": 0.9007407407407407,
      "grad_norm": 1.2879687547683716,
      "learning_rate": 1.051051051051051e-06,
      "loss": 0.24101980030536652,
      "mean_token_accuracy": 0.9254022687673569,
      "num_tokens": 4980736.0,
      "step": 304
    },
    {
      "entropy": 2.6613685488700867,
      "epoch": 0.9037037037037037,
      "grad_norm": 0.8715454339981079,
      "learning_rate": 1.0210210210210212e-06,
      "loss": 0.13674139976501465,
      "mean_token_accuracy": 0.9558299705386162,
      "num_tokens": 4997120.0,
      "step": 305
    },
    {
      "entropy": 3.0323977172374725,
      "epoch": 0.9066666666666666,
      "grad_norm": 1.2636479139328003,
      "learning_rate": 9.909909909909911e-07,
      "loss": 0.22365128993988037,
      "mean_token_accuracy": 0.9214575439691544,
      "num_tokens": 5013504.0,
      "step": 306
    },
    {
      "entropy": 2.7798266112804413,
      "epoch": 0.9096296296296297,
      "grad_norm": 1.071028709411621,
      "learning_rate": 9.60960960960961e-07,
      "loss": 0.25200745463371277,
      "mean_token_accuracy": 0.9370187669992447,
      "num_tokens": 5029888.0,
      "step": 307
    },
    {
      "entropy": 2.966661810874939,
      "epoch": 0.9125925925925926,
      "grad_norm": 0.9755561947822571,
      "learning_rate": 9.30930930930931e-07,
      "loss": 0.1364070028066635,
      "mean_token_accuracy": 0.9633779376745224,
      "num_tokens": 5046272.0,
      "step": 308
    },
    {
      "entropy": 2.8475814908742905,
      "epoch": 0.9155555555555556,
      "grad_norm": 1.1182949542999268,
      "learning_rate": 9.00900900900901e-07,
      "loss": 0.2216225266456604,
      "mean_token_accuracy": 0.9494200721383095,
      "num_tokens": 5062656.0,
      "step": 309
    },
    {
      "entropy": 3.083686202764511,
      "epoch": 0.9185185185185185,
      "grad_norm": 1.019321084022522,
      "learning_rate": 8.708708708708709e-07,
      "loss": 0.17073342204093933,
      "mean_token_accuracy": 0.9476535245776176,
      "num_tokens": 5079040.0,
      "step": 310
    },
    {
      "entropy": 2.7442470490932465,
      "epoch": 0.9214814814814815,
      "grad_norm": 0.9397963881492615,
      "learning_rate": 8.40840840840841e-07,
      "loss": 0.16900652647018433,
      "mean_token_accuracy": 0.956196092069149,
      "num_tokens": 5095424.0,
      "step": 311
    },
    {
      "entropy": 2.976117357611656,
      "epoch": 0.9244444444444444,
      "grad_norm": 0.9691776633262634,
      "learning_rate": 8.108108108108109e-07,
      "loss": 0.14978164434432983,
      "mean_token_accuracy": 0.9600504860281944,
      "num_tokens": 5111808.0,
      "step": 312
    },
    {
      "entropy": 3.0268036127090454,
      "epoch": 0.9274074074074075,
      "grad_norm": 1.3066312074661255,
      "learning_rate": 7.807807807807808e-07,
      "loss": 0.2101222723722458,
      "mean_token_accuracy": 0.9468629732728004,
      "num_tokens": 5128192.0,
      "step": 313
    },
    {
      "entropy": 2.8280453979969025,
      "epoch": 0.9303703703703704,
      "grad_norm": 1.1003689765930176,
      "learning_rate": 7.507507507507509e-07,
      "loss": 0.2189224362373352,
      "mean_token_accuracy": 0.942246712744236,
      "num_tokens": 5144576.0,
      "step": 314
    },
    {
      "entropy": 2.976786032319069,
      "epoch": 0.9333333333333333,
      "grad_norm": 0.9063469767570496,
      "learning_rate": 7.207207207207208e-07,
      "loss": 0.14953091740608215,
      "mean_token_accuracy": 0.9534271508455276,
      "num_tokens": 5160960.0,
      "step": 315
    },
    {
      "entropy": 2.9797759354114532,
      "epoch": 0.9362962962962963,
      "grad_norm": 1.2507872581481934,
      "learning_rate": 6.906906906906907e-07,
      "loss": 0.23357172310352325,
      "mean_token_accuracy": 0.9364637657999992,
      "num_tokens": 5177344.0,
      "step": 316
    },
    {
      "entropy": 2.994144380092621,
      "epoch": 0.9392592592592592,
      "grad_norm": 1.2454637289047241,
      "learning_rate": 6.606606606606607e-07,
      "loss": 0.22540506720542908,
      "mean_token_accuracy": 0.9319886118173599,
      "num_tokens": 5193728.0,
      "step": 317
    },
    {
      "entropy": 3.1280756294727325,
      "epoch": 0.9422222222222222,
      "grad_norm": 0.9932249784469604,
      "learning_rate": 6.306306306306306e-07,
      "loss": 0.11862220615148544,
      "mean_token_accuracy": 0.9601308852434158,
      "num_tokens": 5210112.0,
      "step": 318
    },
    {
      "entropy": 2.7858297526836395,
      "epoch": 0.9451851851851852,
      "grad_norm": 0.8466230034828186,
      "learning_rate": 6.006006006006006e-07,
      "loss": 0.15886148810386658,
      "mean_token_accuracy": 0.9521259963512421,
      "num_tokens": 5226496.0,
      "step": 319
    },
    {
      "entropy": 2.7214138507843018,
      "epoch": 0.9481481481481482,
      "grad_norm": 1.0153858661651611,
      "learning_rate": 5.705705705705706e-07,
      "loss": 0.13641257584095,
      "mean_token_accuracy": 0.9631103426218033,
      "num_tokens": 5242880.0,
      "step": 320
    },
    {
      "entropy": 2.760718896985054,
      "epoch": 0.9511111111111111,
      "grad_norm": 0.7906075119972229,
      "learning_rate": 5.405405405405406e-07,
      "loss": 0.12025375664234161,
      "mean_token_accuracy": 0.9681820869445801,
      "num_tokens": 5259264.0,
      "step": 321
    },
    {
      "entropy": 3.0152215659618378,
      "epoch": 0.9540740740740741,
      "grad_norm": 1.3097968101501465,
      "learning_rate": 5.105105105105106e-07,
      "loss": 0.20160530507564545,
      "mean_token_accuracy": 0.9353676289319992,
      "num_tokens": 5275648.0,
      "step": 322
    },
    {
      "entropy": 2.9761607944965363,
      "epoch": 0.957037037037037,
      "grad_norm": 0.9927833080291748,
      "learning_rate": 4.804804804804805e-07,
      "loss": 0.15498238801956177,
      "mean_token_accuracy": 0.9501162618398666,
      "num_tokens": 5292032.0,
      "step": 323
    },
    {
      "entropy": 2.993096560239792,
      "epoch": 0.96,
      "grad_norm": 1.4152226448059082,
      "learning_rate": 4.504504504504505e-07,
      "loss": 0.28583815693855286,
      "mean_token_accuracy": 0.9212516695261002,
      "num_tokens": 5308416.0,
      "step": 324
    },
    {
      "entropy": 3.1000093817710876,
      "epoch": 0.9629629629629629,
      "grad_norm": 1.1058512926101685,
      "learning_rate": 4.204204204204205e-07,
      "loss": 0.1754665970802307,
      "mean_token_accuracy": 0.9442660957574844,
      "num_tokens": 5324800.0,
      "step": 325
    },
    {
      "entropy": 3.0041013956069946,
      "epoch": 0.965925925925926,
      "grad_norm": 0.7665889859199524,
      "learning_rate": 3.903903903903904e-07,
      "loss": 0.09065846353769302,
      "mean_token_accuracy": 0.9687816500663757,
      "num_tokens": 5341184.0,
      "step": 326
    },
    {
      "entropy": 2.494102329015732,
      "epoch": 0.9688888888888889,
      "grad_norm": 0.7940945625305176,
      "learning_rate": 3.603603603603604e-07,
      "loss": 0.1419333815574646,
      "mean_token_accuracy": 0.9551639035344124,
      "num_tokens": 5357568.0,
      "step": 327
    },
    {
      "entropy": 3.124778002500534,
      "epoch": 0.9718518518518519,
      "grad_norm": 1.1697940826416016,
      "learning_rate": 3.3033033033033036e-07,
      "loss": 0.20611074566841125,
      "mean_token_accuracy": 0.9439872354269028,
      "num_tokens": 5373952.0,
      "step": 328
    },
    {
      "entropy": 2.7635384500026703,
      "epoch": 0.9748148148148148,
      "grad_norm": 1.0099842548370361,
      "learning_rate": 3.003003003003003e-07,
      "loss": 0.19179926812648773,
      "mean_token_accuracy": 0.9434934630990028,
      "num_tokens": 5390336.0,
      "step": 329
    },
    {
      "entropy": 2.8500694930553436,
      "epoch": 0.9777777777777777,
      "grad_norm": 0.8996491432189941,
      "learning_rate": 2.702702702702703e-07,
      "loss": 0.1620231419801712,
      "mean_token_accuracy": 0.9505789652466774,
      "num_tokens": 5406720.0,
      "step": 330
    },
    {
      "entropy": 2.9804718792438507,
      "epoch": 0.9807407407407407,
      "grad_norm": 1.110015869140625,
      "learning_rate": 2.4024024024024026e-07,
      "loss": 0.21477189660072327,
      "mean_token_accuracy": 0.9368481785058975,
      "num_tokens": 5423104.0,
      "step": 331
    },
    {
      "entropy": 2.7032893896102905,
      "epoch": 0.9837037037037037,
      "grad_norm": 0.8577917218208313,
      "learning_rate": 2.1021021021021025e-07,
      "loss": 0.1445036679506302,
      "mean_token_accuracy": 0.9551440551877022,
      "num_tokens": 5439488.0,
      "step": 332
    },
    {
      "entropy": 2.7800481021404266,
      "epoch": 0.9866666666666667,
      "grad_norm": 0.9222099184989929,
      "learning_rate": 1.801801801801802e-07,
      "loss": 0.1509193778038025,
      "mean_token_accuracy": 0.9618949517607689,
      "num_tokens": 5455872.0,
      "step": 333
    },
    {
      "entropy": 2.795774072408676,
      "epoch": 0.9896296296296296,
      "grad_norm": 1.091988205909729,
      "learning_rate": 1.5015015015015016e-07,
      "loss": 0.185760959982872,
      "mean_token_accuracy": 0.9529979601502419,
      "num_tokens": 5472256.0,
      "step": 334
    },
    {
      "entropy": 3.014428496360779,
      "epoch": 0.9925925925925926,
      "grad_norm": 1.1064698696136475,
      "learning_rate": 1.2012012012012013e-07,
      "loss": 0.1730300635099411,
      "mean_token_accuracy": 0.9544021263718605,
      "num_tokens": 5488640.0,
      "step": 335
    },
    {
      "entropy": 2.7403995990753174,
      "epoch": 0.9955555555555555,
      "grad_norm": 1.0951151847839355,
      "learning_rate": 9.00900900900901e-08,
      "loss": 0.19785557687282562,
      "mean_token_accuracy": 0.9353931546211243,
      "num_tokens": 5505024.0,
      "step": 336
    },
    {
      "entropy": 3.0471527874469757,
      "epoch": 0.9985185185185185,
      "grad_norm": 1.2756664752960205,
      "learning_rate": 6.006006006006006e-08,
      "loss": 0.24869737029075623,
      "mean_token_accuracy": 0.9270326271653175,
      "num_tokens": 5521408.0,
      "step": 337
    },
    {
      "entropy": 2.9130772948265076,
      "epoch": 1.0,
      "grad_norm": 2.2841439247131348,
      "learning_rate": 3.003003003003003e-08,
      "loss": 0.2699161171913147,
      "mean_token_accuracy": 0.9327885955572128,
      "num_tokens": 5529600.0,
      "step": 338
    }
  ],
  "logging_steps": 1,
  "max_steps": 338,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.53130594336768e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}