{
  "best_global_step": 300,
  "best_metric": 0.18680252134799957,
  "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": 5.0673428773880005,
      "epoch": 0.002962962962962963,
      "grad_norm": 25.440805435180664,
      "learning_rate": 0.0,
      "loss": 2.900369882583618,
      "mean_token_accuracy": 0.6152051985263824,
      "num_tokens": 16384.0,
      "step": 1
    },
    {
      "entropy": 5.123473107814789,
      "epoch": 0.005925925925925926,
      "grad_norm": 23.987611770629883,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.6741981506347656,
      "mean_token_accuracy": 0.6215632483363152,
      "num_tokens": 32768.0,
      "step": 2
    },
    {
      "entropy": 4.608297616243362,
      "epoch": 0.008888888888888889,
      "grad_norm": 23.673486709594727,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.3658032417297363,
      "mean_token_accuracy": 0.6638018935918808,
      "num_tokens": 49152.0,
      "step": 3
    },
    {
      "entropy": 5.036638975143433,
      "epoch": 0.011851851851851851,
      "grad_norm": 25.90877342224121,
      "learning_rate": 6e-06,
      "loss": 2.8426554203033447,
      "mean_token_accuracy": 0.6197315752506256,
      "num_tokens": 65536.0,
      "step": 4
    },
    {
      "entropy": 5.036393016576767,
      "epoch": 0.014814814814814815,
      "grad_norm": 29.470762252807617,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.254727363586426,
      "mean_token_accuracy": 0.665954053401947,
      "num_tokens": 81920.0,
      "step": 5
    },
    {
      "entropy": 4.822231233119965,
      "epoch": 0.017777777777777778,
      "grad_norm": 21.614458084106445,
      "learning_rate": 1e-05,
      "loss": 1.91741943359375,
      "mean_token_accuracy": 0.6739806681871414,
      "num_tokens": 98304.0,
      "step": 6
    },
    {
      "entropy": 4.474497497081757,
      "epoch": 0.02074074074074074,
      "grad_norm": 12.429451942443848,
      "learning_rate": 9.96996996996997e-06,
      "loss": 1.301264762878418,
      "mean_token_accuracy": 0.7622850760817528,
      "num_tokens": 114688.0,
      "step": 7
    },
    {
      "entropy": 4.973697364330292,
      "epoch": 0.023703703703703703,
      "grad_norm": 9.665480613708496,
      "learning_rate": 9.93993993993994e-06,
      "loss": 1.2653709650039673,
      "mean_token_accuracy": 0.7608245834708214,
      "num_tokens": 131072.0,
      "step": 8
    },
    {
      "entropy": 4.872437655925751,
      "epoch": 0.02666666666666667,
      "grad_norm": 8.410284042358398,
      "learning_rate": 9.90990990990991e-06,
      "loss": 1.4004276990890503,
      "mean_token_accuracy": 0.7534106895327568,
      "num_tokens": 147456.0,
      "step": 9
    },
    {
      "entropy": 4.694646030664444,
      "epoch": 0.02962962962962963,
      "grad_norm": 6.188877582550049,
      "learning_rate": 9.879879879879881e-06,
      "loss": 0.9507288336753845,
      "mean_token_accuracy": 0.8184276148676872,
      "num_tokens": 163840.0,
      "step": 10
    },
    {
      "entropy": 4.990409344434738,
      "epoch": 0.03259259259259259,
      "grad_norm": 5.39843225479126,
      "learning_rate": 9.849849849849851e-06,
      "loss": 0.8838430047035217,
      "mean_token_accuracy": 0.8182023242115974,
      "num_tokens": 180224.0,
      "step": 11
    },
    {
      "entropy": 5.089408040046692,
      "epoch": 0.035555555555555556,
      "grad_norm": 5.83709716796875,
      "learning_rate": 9.81981981981982e-06,
      "loss": 1.2019895315170288,
      "mean_token_accuracy": 0.7656112164258957,
      "num_tokens": 196608.0,
      "step": 12
    },
    {
      "entropy": 4.942709267139435,
      "epoch": 0.03851851851851852,
      "grad_norm": 3.295132875442505,
      "learning_rate": 9.78978978978979e-06,
      "loss": 0.6716622710227966,
      "mean_token_accuracy": 0.8567092791199684,
      "num_tokens": 212992.0,
      "step": 13
    },
    {
      "entropy": 4.812331706285477,
      "epoch": 0.04148148148148148,
      "grad_norm": 3.998014211654663,
      "learning_rate": 9.75975975975976e-06,
      "loss": 0.9384757280349731,
      "mean_token_accuracy": 0.8012193888425827,
      "num_tokens": 229376.0,
      "step": 14
    },
    {
      "entropy": 4.730546981096268,
      "epoch": 0.044444444444444446,
      "grad_norm": 3.4215450286865234,
      "learning_rate": 9.729729729729732e-06,
      "loss": 0.6778247356414795,
      "mean_token_accuracy": 0.8175079971551895,
      "num_tokens": 245760.0,
      "step": 15
    },
    {
      "entropy": 4.551107466220856,
      "epoch": 0.047407407407407405,
      "grad_norm": 3.506895065307617,
      "learning_rate": 9.699699699699701e-06,
      "loss": 0.8040162920951843,
      "mean_token_accuracy": 0.8257298544049263,
      "num_tokens": 262144.0,
      "step": 16
    },
    {
      "entropy": 4.683808833360672,
      "epoch": 0.05037037037037037,
      "grad_norm": 2.353834867477417,
      "learning_rate": 9.669669669669671e-06,
      "loss": 0.4675115644931793,
      "mean_token_accuracy": 0.8599296435713768,
      "num_tokens": 278528.0,
      "step": 17
    },
    {
      "entropy": 4.575067818164825,
      "epoch": 0.05333333333333334,
      "grad_norm": 3.1972451210021973,
      "learning_rate": 9.63963963963964e-06,
      "loss": 0.5954794883728027,
      "mean_token_accuracy": 0.8707093447446823,
      "num_tokens": 294912.0,
      "step": 18
    },
    {
      "entropy": 4.868251651525497,
      "epoch": 0.056296296296296296,
      "grad_norm": 2.378986120223999,
      "learning_rate": 9.60960960960961e-06,
      "loss": 0.5159575343132019,
      "mean_token_accuracy": 0.8798045217990875,
      "num_tokens": 311296.0,
      "step": 19
    },
    {
      "entropy": 4.48080986738205,
      "epoch": 0.05925925925925926,
      "grad_norm": 2.375600576400757,
      "learning_rate": 9.57957957957958e-06,
      "loss": 0.4439718425273895,
      "mean_token_accuracy": 0.8897516280412674,
      "num_tokens": 327680.0,
      "step": 20
    },
    {
      "entropy": 4.7171652019023895,
      "epoch": 0.06222222222222222,
      "grad_norm": 1.6935182809829712,
      "learning_rate": 9.54954954954955e-06,
      "loss": 0.3384687006473541,
      "mean_token_accuracy": 0.9077081233263016,
      "num_tokens": 344064.0,
      "step": 21
    },
    {
      "entropy": 4.822179198265076,
      "epoch": 0.06518518518518518,
      "grad_norm": 2.1172261238098145,
      "learning_rate": 9.51951951951952e-06,
      "loss": 0.4139059782028198,
      "mean_token_accuracy": 0.8918816298246384,
      "num_tokens": 360448.0,
      "step": 22
    },
    {
      "entropy": 4.77111691236496,
      "epoch": 0.06814814814814815,
      "grad_norm": 2.8216207027435303,
      "learning_rate": 9.489489489489491e-06,
      "loss": 0.5718841552734375,
      "mean_token_accuracy": 0.8634202927350998,
      "num_tokens": 376832.0,
      "step": 23
    },
    {
      "entropy": 4.7567101418972015,
      "epoch": 0.07111111111111111,
      "grad_norm": 2.0149741172790527,
      "learning_rate": 9.45945945945946e-06,
      "loss": 0.3909273147583008,
      "mean_token_accuracy": 0.8937204480171204,
      "num_tokens": 393216.0,
      "step": 24
    },
    {
      "entropy": 5.09102475643158,
      "epoch": 0.07407407407407407,
      "grad_norm": 2.1713316440582275,
      "learning_rate": 9.42942942942943e-06,
      "loss": 0.5473820567131042,
      "mean_token_accuracy": 0.8722654804587364,
      "num_tokens": 409600.0,
      "step": 25
    },
    {
      "entropy": 4.927625209093094,
      "epoch": 0.07703703703703704,
      "grad_norm": 1.8736480474472046,
      "learning_rate": 9.3993993993994e-06,
      "loss": 0.4226554036140442,
      "mean_token_accuracy": 0.8974572792649269,
      "num_tokens": 425984.0,
      "step": 26
    },
    {
      "entropy": 4.773389220237732,
      "epoch": 0.08,
      "grad_norm": 1.4882359504699707,
      "learning_rate": 9.36936936936937e-06,
      "loss": 0.2609105706214905,
      "mean_token_accuracy": 0.9335159808397293,
      "num_tokens": 442368.0,
      "step": 27
    },
    {
      "entropy": 4.801750421524048,
      "epoch": 0.08296296296296296,
      "grad_norm": 1.8475184440612793,
      "learning_rate": 9.339339339339341e-06,
      "loss": 0.37484052777290344,
      "mean_token_accuracy": 0.9150584638118744,
      "num_tokens": 458752.0,
      "step": 28
    },
    {
      "entropy": 4.914631545543671,
      "epoch": 0.08592592592592592,
      "grad_norm": 2.0742545127868652,
      "learning_rate": 9.30930930930931e-06,
      "loss": 0.41960597038269043,
      "mean_token_accuracy": 0.8818418234586716,
      "num_tokens": 475136.0,
      "step": 29
    },
    {
      "entropy": 4.78462827205658,
      "epoch": 0.08888888888888889,
      "grad_norm": 1.5355863571166992,
      "learning_rate": 9.27927927927928e-06,
      "loss": 0.31162041425704956,
      "mean_token_accuracy": 0.9063951596617699,
      "num_tokens": 491520.0,
      "step": 30
    },
    {
      "entropy": 4.981134653091431,
      "epoch": 0.09185185185185185,
      "grad_norm": 1.8707942962646484,
      "learning_rate": 9.24924924924925e-06,
      "loss": 0.4265978932380676,
      "mean_token_accuracy": 0.8898465484380722,
      "num_tokens": 507904.0,
      "step": 31
    },
    {
      "entropy": 4.555660128593445,
      "epoch": 0.09481481481481481,
      "grad_norm": 1.3826193809509277,
      "learning_rate": 9.21921921921922e-06,
      "loss": 0.301828533411026,
      "mean_token_accuracy": 0.9234981089830399,
      "num_tokens": 524288.0,
      "step": 32
    },
    {
      "entropy": 4.821201682090759,
      "epoch": 0.09777777777777778,
      "grad_norm": 1.5057733058929443,
      "learning_rate": 9.189189189189191e-06,
      "loss": 0.31418877840042114,
      "mean_token_accuracy": 0.9189433455467224,
      "num_tokens": 540672.0,
      "step": 33
    },
    {
      "entropy": 4.60411012172699,
      "epoch": 0.10074074074074074,
      "grad_norm": 1.1826591491699219,
      "learning_rate": 9.15915915915916e-06,
      "loss": 0.20352856814861298,
      "mean_token_accuracy": 0.9490585774183273,
      "num_tokens": 557056.0,
      "step": 34
    },
    {
      "entropy": 4.409397542476654,
      "epoch": 0.1037037037037037,
      "grad_norm": 1.1938457489013672,
      "learning_rate": 9.129129129129129e-06,
      "loss": 0.2377718985080719,
      "mean_token_accuracy": 0.9436894059181213,
      "num_tokens": 573440.0,
      "step": 35
    },
    {
      "entropy": 4.654195666313171,
      "epoch": 0.10666666666666667,
      "grad_norm": 1.6995933055877686,
      "learning_rate": 9.0990990990991e-06,
      "loss": 0.47783029079437256,
      "mean_token_accuracy": 0.8897658511996269,
      "num_tokens": 589824.0,
      "step": 36
    },
    {
      "entropy": 4.379251688718796,
      "epoch": 0.10962962962962963,
      "grad_norm": 1.1382306814193726,
      "learning_rate": 9.06906906906907e-06,
      "loss": 0.21475709974765778,
      "mean_token_accuracy": 0.9456312358379364,
      "num_tokens": 606208.0,
      "step": 37
    },
    {
      "entropy": 4.262269973754883,
      "epoch": 0.11259259259259259,
      "grad_norm": 1.1475385427474976,
      "learning_rate": 9.03903903903904e-06,
      "loss": 0.19059525430202484,
      "mean_token_accuracy": 0.9473334029316902,
      "num_tokens": 622592.0,
      "step": 38
    },
    {
      "entropy": 4.488475859165192,
      "epoch": 0.11555555555555555,
      "grad_norm": 1.3030574321746826,
      "learning_rate": 9.00900900900901e-06,
      "loss": 0.25836455821990967,
      "mean_token_accuracy": 0.928585596382618,
      "num_tokens": 638976.0,
      "step": 39
    },
    {
      "entropy": 4.752066731452942,
      "epoch": 0.11851851851851852,
      "grad_norm": 1.4755479097366333,
      "learning_rate": 8.97897897897898e-06,
      "loss": 0.342281311750412,
      "mean_token_accuracy": 0.9128458872437477,
      "num_tokens": 655360.0,
      "step": 40
    },
    {
      "entropy": 4.498592525720596,
      "epoch": 0.12148148148148148,
      "grad_norm": 1.222934603691101,
      "learning_rate": 8.94894894894895e-06,
      "loss": 0.3216207027435303,
      "mean_token_accuracy": 0.9307751879096031,
      "num_tokens": 671744.0,
      "step": 41
    },
    {
      "entropy": 4.6110716462135315,
      "epoch": 0.12444444444444444,
      "grad_norm": 1.5906107425689697,
      "learning_rate": 8.91891891891892e-06,
      "loss": 0.33854198455810547,
      "mean_token_accuracy": 0.9101497828960419,
      "num_tokens": 688128.0,
      "step": 42
    },
    {
      "entropy": 4.527588397264481,
      "epoch": 0.1274074074074074,
      "grad_norm": 1.4244577884674072,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.28006404638290405,
      "mean_token_accuracy": 0.929267980158329,
      "num_tokens": 704512.0,
      "step": 43
    },
    {
      "entropy": 5.190286874771118,
      "epoch": 0.13037037037037036,
      "grad_norm": 1.5435229539871216,
      "learning_rate": 8.85885885885886e-06,
      "loss": 0.35443374514579773,
      "mean_token_accuracy": 0.9155692234635353,
      "num_tokens": 720896.0,
      "step": 44
    },
    {
      "entropy": 4.250262916088104,
      "epoch": 0.13333333333333333,
      "grad_norm": 1.3200054168701172,
      "learning_rate": 8.82882882882883e-06,
      "loss": 0.3303371071815491,
      "mean_token_accuracy": 0.913819931447506,
      "num_tokens": 737280.0,
      "step": 45
    },
    {
      "entropy": 4.1272053718566895,
      "epoch": 0.1362962962962963,
      "grad_norm": 1.008841633796692,
      "learning_rate": 8.798798798798799e-06,
      "loss": 0.17603322863578796,
      "mean_token_accuracy": 0.9506337121129036,
      "num_tokens": 753664.0,
      "step": 46
    },
    {
      "entropy": 4.882783472537994,
      "epoch": 0.13925925925925925,
      "grad_norm": 1.5943907499313354,
      "learning_rate": 8.768768768768769e-06,
      "loss": 0.3042362332344055,
      "mean_token_accuracy": 0.9218109101057053,
      "num_tokens": 770048.0,
      "step": 47
    },
    {
      "entropy": 4.354182213544846,
      "epoch": 0.14222222222222222,
      "grad_norm": 0.9702581167221069,
      "learning_rate": 8.738738738738739e-06,
      "loss": 0.13799810409545898,
      "mean_token_accuracy": 0.9640689194202423,
      "num_tokens": 786432.0,
      "step": 48
    },
    {
      "entropy": 4.545175939798355,
      "epoch": 0.1451851851851852,
      "grad_norm": 1.3666038513183594,
      "learning_rate": 8.70870870870871e-06,
      "loss": 0.2714568078517914,
      "mean_token_accuracy": 0.9376921951770782,
      "num_tokens": 802816.0,
      "step": 49
    },
    {
      "entropy": 4.437481105327606,
      "epoch": 0.14814814814814814,
      "grad_norm": 1.278706669807434,
      "learning_rate": 8.67867867867868e-06,
      "loss": 0.3146522045135498,
      "mean_token_accuracy": 0.9113487079739571,
      "num_tokens": 819200.0,
      "step": 50
    },
    {
      "epoch": 0.14814814814814814,
      "eval_entropy": 4.500472501118978,
      "eval_loss": 0.26209747791290283,
      "eval_mean_token_accuracy": 0.9333808883031209,
      "eval_num_tokens": 819200.0,
      "eval_runtime": 41.786,
      "eval_samples_per_second": 14.359,
      "eval_steps_per_second": 1.795,
      "step": 50
    },
    {
      "entropy": 4.5342777371406555,
      "epoch": 0.1511111111111111,
      "grad_norm": 1.1738427877426147,
      "learning_rate": 8.64864864864865e-06,
      "loss": 0.24006211757659912,
      "mean_token_accuracy": 0.9387434720993042,
      "num_tokens": 835584.0,
      "step": 51
    },
    {
      "entropy": 4.773931682109833,
      "epoch": 0.15407407407407409,
      "grad_norm": 1.3778016567230225,
      "learning_rate": 8.618618618618619e-06,
      "loss": 0.28196990489959717,
      "mean_token_accuracy": 0.9180527254939079,
      "num_tokens": 851968.0,
      "step": 52
    },
    {
      "entropy": 4.472855925559998,
      "epoch": 0.15703703703703703,
      "grad_norm": 1.4825477600097656,
      "learning_rate": 8.588588588588589e-06,
      "loss": 0.36939430236816406,
      "mean_token_accuracy": 0.9117205366492271,
      "num_tokens": 868352.0,
      "step": 53
    },
    {
      "entropy": 4.932996213436127,
      "epoch": 0.16,
      "grad_norm": 1.311163306236267,
      "learning_rate": 8.55855855855856e-06,
      "loss": 0.28928816318511963,
      "mean_token_accuracy": 0.9164340123534203,
      "num_tokens": 884736.0,
      "step": 54
    },
    {
      "entropy": 4.495872259140015,
      "epoch": 0.16296296296296298,
      "grad_norm": 1.3946824073791504,
      "learning_rate": 8.52852852852853e-06,
      "loss": 0.24732254445552826,
      "mean_token_accuracy": 0.941411204636097,
      "num_tokens": 901120.0,
      "step": 55
    },
    {
      "entropy": 4.538569241762161,
      "epoch": 0.16592592592592592,
      "grad_norm": 1.0770541429519653,
      "learning_rate": 8.4984984984985e-06,
      "loss": 0.23028765618801117,
      "mean_token_accuracy": 0.9363544955849648,
      "num_tokens": 917504.0,
      "step": 56
    },
    {
      "entropy": 4.913299918174744,
      "epoch": 0.1688888888888889,
      "grad_norm": 1.426444172859192,
      "learning_rate": 8.46846846846847e-06,
      "loss": 0.3080557584762573,
      "mean_token_accuracy": 0.9213304966688156,
      "num_tokens": 933888.0,
      "step": 57
    },
    {
      "entropy": 4.198328018188477,
      "epoch": 0.17185185185185184,
      "grad_norm": 1.093044400215149,
      "learning_rate": 8.438438438438439e-06,
      "loss": 0.2499372661113739,
      "mean_token_accuracy": 0.9285173341631889,
      "num_tokens": 950272.0,
      "step": 58
    },
    {
      "entropy": 4.64614525437355,
      "epoch": 0.1748148148148148,
      "grad_norm": 1.2315367460250854,
      "learning_rate": 8.408408408408409e-06,
      "loss": 0.22362437844276428,
      "mean_token_accuracy": 0.9453909620642662,
      "num_tokens": 966656.0,
      "step": 59
    },
    {
      "entropy": 4.655407428741455,
      "epoch": 0.17777777777777778,
      "grad_norm": 1.0248944759368896,
      "learning_rate": 8.378378378378378e-06,
      "loss": 0.12162414193153381,
      "mean_token_accuracy": 0.9611344784498215,
      "num_tokens": 983040.0,
      "step": 60
    },
    {
      "entropy": 4.484728246927261,
      "epoch": 0.18074074074074073,
      "grad_norm": 1.2880580425262451,
      "learning_rate": 8.348348348348348e-06,
      "loss": 0.31024685502052307,
      "mean_token_accuracy": 0.9185083135962486,
      "num_tokens": 999424.0,
      "step": 61
    },
    {
      "entropy": 4.706384867429733,
      "epoch": 0.1837037037037037,
      "grad_norm": 1.5274730920791626,
      "learning_rate": 8.31831831831832e-06,
      "loss": 0.32353639602661133,
      "mean_token_accuracy": 0.9106626957654953,
      "num_tokens": 1015808.0,
      "step": 62
    },
    {
      "entropy": 4.800395727157593,
      "epoch": 0.18666666666666668,
      "grad_norm": 1.1703046560287476,
      "learning_rate": 8.288288288288289e-06,
      "loss": 0.2841051518917084,
      "mean_token_accuracy": 0.9192538112401962,
      "num_tokens": 1032192.0,
      "step": 63
    },
    {
      "entropy": 4.493288844823837,
      "epoch": 0.18962962962962962,
      "grad_norm": 1.2824828624725342,
      "learning_rate": 8.258258258258259e-06,
      "loss": 0.2465566098690033,
      "mean_token_accuracy": 0.9381505995988846,
      "num_tokens": 1048576.0,
      "step": 64
    },
    {
      "entropy": 4.90463325381279,
      "epoch": 0.1925925925925926,
      "grad_norm": 1.5403709411621094,
      "learning_rate": 8.228228228228229e-06,
      "loss": 0.3537861704826355,
      "mean_token_accuracy": 0.9023569896817207,
      "num_tokens": 1064960.0,
      "step": 65
    },
    {
      "entropy": 4.776595234870911,
      "epoch": 0.19555555555555557,
      "grad_norm": 1.209590196609497,
      "learning_rate": 8.198198198198198e-06,
      "loss": 0.19631054997444153,
      "mean_token_accuracy": 0.9528850540518761,
      "num_tokens": 1081344.0,
      "step": 66
    },
    {
      "entropy": 4.997211933135986,
      "epoch": 0.1985185185185185,
      "grad_norm": 1.906493067741394,
      "learning_rate": 8.16816816816817e-06,
      "loss": 0.38569140434265137,
      "mean_token_accuracy": 0.9013442024588585,
      "num_tokens": 1097728.0,
      "step": 67
    },
    {
      "entropy": 4.652494341135025,
      "epoch": 0.20148148148148148,
      "grad_norm": 1.158506155014038,
      "learning_rate": 8.13813813813814e-06,
      "loss": 0.2167915403842926,
      "mean_token_accuracy": 0.9449609816074371,
      "num_tokens": 1114112.0,
      "step": 68
    },
    {
      "entropy": 3.9529418349266052,
      "epoch": 0.20444444444444446,
      "grad_norm": 0.9355016946792603,
      "learning_rate": 8.108108108108109e-06,
      "loss": 0.1424209326505661,
      "mean_token_accuracy": 0.9587113484740257,
      "num_tokens": 1130496.0,
      "step": 69
    },
    {
      "entropy": 4.400442659854889,
      "epoch": 0.2074074074074074,
      "grad_norm": 1.4780858755111694,
      "learning_rate": 8.078078078078079e-06,
      "loss": 0.2362896054983139,
      "mean_token_accuracy": 0.9273651763796806,
      "num_tokens": 1146880.0,
      "step": 70
    },
    {
      "entropy": 4.532440781593323,
      "epoch": 0.21037037037037037,
      "grad_norm": 1.2864397764205933,
      "learning_rate": 8.048048048048048e-06,
      "loss": 0.2748740017414093,
      "mean_token_accuracy": 0.9232638701796532,
      "num_tokens": 1163264.0,
      "step": 71
    },
    {
      "entropy": 4.795024394989014,
      "epoch": 0.21333333333333335,
      "grad_norm": 1.156823754310608,
      "learning_rate": 8.018018018018018e-06,
      "loss": 0.24251019954681396,
      "mean_token_accuracy": 0.9389124140143394,
      "num_tokens": 1179648.0,
      "step": 72
    },
    {
      "entropy": 4.553127020597458,
      "epoch": 0.2162962962962963,
      "grad_norm": 0.9880868196487427,
      "learning_rate": 7.987987987987988e-06,
      "loss": 0.18323005735874176,
      "mean_token_accuracy": 0.9462512955069542,
      "num_tokens": 1196032.0,
      "step": 73
    },
    {
      "entropy": 4.950550019741058,
      "epoch": 0.21925925925925926,
      "grad_norm": 1.4889500141143799,
      "learning_rate": 7.95795795795796e-06,
      "loss": 0.3610566258430481,
      "mean_token_accuracy": 0.9149008467793465,
      "num_tokens": 1212416.0,
      "step": 74
    },
    {
      "entropy": 4.8895076513290405,
      "epoch": 0.2222222222222222,
      "grad_norm": 1.4464665651321411,
      "learning_rate": 7.927927927927929e-06,
      "loss": 0.2880629003047943,
      "mean_token_accuracy": 0.926278293132782,
      "num_tokens": 1228800.0,
      "step": 75
    },
    {
      "entropy": 4.833582162857056,
      "epoch": 0.22518518518518518,
      "grad_norm": 1.114487648010254,
      "learning_rate": 7.897897897897899e-06,
      "loss": 0.2449398636817932,
      "mean_token_accuracy": 0.936882995069027,
      "num_tokens": 1245184.0,
      "step": 76
    },
    {
      "entropy": 4.6228170692920685,
      "epoch": 0.22814814814814816,
      "grad_norm": 1.278285026550293,
      "learning_rate": 7.867867867867868e-06,
      "loss": 0.2649390697479248,
      "mean_token_accuracy": 0.9343116357922554,
      "num_tokens": 1261568.0,
      "step": 77
    },
    {
      "entropy": 4.726460009813309,
      "epoch": 0.2311111111111111,
      "grad_norm": 1.7650485038757324,
      "learning_rate": 7.837837837837838e-06,
      "loss": 0.35958024859428406,
      "mean_token_accuracy": 0.9256116822361946,
      "num_tokens": 1277952.0,
      "step": 78
    },
    {
      "entropy": 4.449074804782867,
      "epoch": 0.23407407407407407,
      "grad_norm": 1.3984698057174683,
      "learning_rate": 7.807807807807808e-06,
      "loss": 0.25653013586997986,
      "mean_token_accuracy": 0.9214952737092972,
      "num_tokens": 1294336.0,
      "step": 79
    },
    {
      "entropy": 4.68617457151413,
      "epoch": 0.23703703703703705,
      "grad_norm": 1.1441717147827148,
      "learning_rate": 7.77777777777778e-06,
      "loss": 0.22266840934753418,
      "mean_token_accuracy": 0.9355207979679108,
      "num_tokens": 1310720.0,
      "step": 80
    },
    {
      "entropy": 4.463802665472031,
      "epoch": 0.24,
      "grad_norm": 1.0529142618179321,
      "learning_rate": 7.747747747747749e-06,
      "loss": 0.1921442151069641,
      "mean_token_accuracy": 0.9416262805461884,
      "num_tokens": 1327104.0,
      "step": 81
    },
    {
      "entropy": 4.81391379237175,
      "epoch": 0.24296296296296296,
      "grad_norm": 1.2828669548034668,
      "learning_rate": 7.717717717717719e-06,
      "loss": 0.2043302357196808,
      "mean_token_accuracy": 0.9364976286888123,
      "num_tokens": 1343488.0,
      "step": 82
    },
    {
      "entropy": 4.18067741394043,
      "epoch": 0.24592592592592594,
      "grad_norm": 1.1066516637802124,
      "learning_rate": 7.687687687687688e-06,
      "loss": 0.1996747851371765,
      "mean_token_accuracy": 0.948727585375309,
      "num_tokens": 1359872.0,
      "step": 83
    },
    {
      "entropy": 4.416668742895126,
      "epoch": 0.24888888888888888,
      "grad_norm": 1.0599511861801147,
      "learning_rate": 7.657657657657658e-06,
      "loss": 0.16297127306461334,
      "mean_token_accuracy": 0.955065943300724,
      "num_tokens": 1376256.0,
      "step": 84
    },
    {
      "entropy": 4.564825862646103,
      "epoch": 0.2518518518518518,
      "grad_norm": 1.1084920167922974,
      "learning_rate": 7.6276276276276285e-06,
      "loss": 0.20912106335163116,
      "mean_token_accuracy": 0.9353301525115967,
      "num_tokens": 1392640.0,
      "step": 85
    },
    {
      "entropy": 4.509587317705154,
      "epoch": 0.2548148148148148,
      "grad_norm": 1.1490081548690796,
      "learning_rate": 7.597597597597598e-06,
      "loss": 0.1719294935464859,
      "mean_token_accuracy": 0.9419969469308853,
      "num_tokens": 1409024.0,
      "step": 86
    },
    {
      "entropy": 4.409897357225418,
      "epoch": 0.2577777777777778,
      "grad_norm": 1.3643550872802734,
      "learning_rate": 7.567567567567569e-06,
      "loss": 0.27544283866882324,
      "mean_token_accuracy": 0.9323878660798073,
      "num_tokens": 1425408.0,
      "step": 87
    },
    {
      "entropy": 4.055213838815689,
      "epoch": 0.2607407407407407,
      "grad_norm": 0.9993941187858582,
      "learning_rate": 7.5375375375375385e-06,
      "loss": 0.17965731024742126,
      "mean_token_accuracy": 0.9458681270480156,
      "num_tokens": 1441792.0,
      "step": 88
    },
    {
      "entropy": 4.460262417793274,
      "epoch": 0.2637037037037037,
      "grad_norm": 1.2199333906173706,
      "learning_rate": 7.507507507507507e-06,
      "loss": 0.21355192363262177,
      "mean_token_accuracy": 0.9442374929785728,
      "num_tokens": 1458176.0,
      "step": 89
    },
    {
      "entropy": 4.432148456573486,
      "epoch": 0.26666666666666666,
      "grad_norm": 0.9807215332984924,
      "learning_rate": 7.477477477477479e-06,
      "loss": 0.1598125398159027,
      "mean_token_accuracy": 0.955281138420105,
      "num_tokens": 1474560.0,
      "step": 90
    },
    {
      "entropy": 4.66013303399086,
      "epoch": 0.2696296296296296,
      "grad_norm": 1.1059590578079224,
      "learning_rate": 7.447447447447448e-06,
      "loss": 0.20555147528648376,
      "mean_token_accuracy": 0.9434520378708839,
      "num_tokens": 1490944.0,
      "step": 91
    },
    {
      "entropy": 4.7525100111961365,
      "epoch": 0.2725925925925926,
      "grad_norm": 1.2281633615493774,
      "learning_rate": 7.417417417417418e-06,
      "loss": 0.19257581233978271,
      "mean_token_accuracy": 0.9474203884601593,
      "num_tokens": 1507328.0,
      "step": 92
    },
    {
      "entropy": 4.724747955799103,
      "epoch": 0.27555555555555555,
      "grad_norm": 1.3266377449035645,
      "learning_rate": 7.387387387387388e-06,
      "loss": 0.3014601469039917,
      "mean_token_accuracy": 0.9271773770451546,
      "num_tokens": 1523712.0,
      "step": 93
    },
    {
      "entropy": 4.776274889707565,
      "epoch": 0.2785185185185185,
      "grad_norm": 1.2745649814605713,
      "learning_rate": 7.3573573573573575e-06,
      "loss": 0.24372521042823792,
      "mean_token_accuracy": 0.9372419118881226,
      "num_tokens": 1540096.0,
      "step": 94
    },
    {
      "entropy": 4.369442045688629,
      "epoch": 0.2814814814814815,
      "grad_norm": 0.9788532853126526,
      "learning_rate": 7.327327327327328e-06,
      "loss": 0.15342336893081665,
      "mean_token_accuracy": 0.9490785896778107,
      "num_tokens": 1556480.0,
      "step": 95
    },
    {
      "entropy": 4.46684268116951,
      "epoch": 0.28444444444444444,
      "grad_norm": 1.2184175252914429,
      "learning_rate": 7.297297297297298e-06,
      "loss": 0.1922316998243332,
      "mean_token_accuracy": 0.9433775171637535,
      "num_tokens": 1572864.0,
      "step": 96
    },
    {
      "entropy": 4.197163283824921,
      "epoch": 0.2874074074074074,
      "grad_norm": 0.8942313194274902,
      "learning_rate": 7.267267267267268e-06,
      "loss": 0.12389811873435974,
      "mean_token_accuracy": 0.967034637928009,
      "num_tokens": 1589248.0,
      "step": 97
    },
    {
      "entropy": 4.349963843822479,
      "epoch": 0.2903703703703704,
      "grad_norm": 1.1199239492416382,
      "learning_rate": 7.237237237237238e-06,
      "loss": 0.21292273700237274,
      "mean_token_accuracy": 0.9435210227966309,
      "num_tokens": 1605632.0,
      "step": 98
    },
    {
      "entropy": 4.32994818687439,
      "epoch": 0.29333333333333333,
      "grad_norm": 0.851973831653595,
      "learning_rate": 7.207207207207208e-06,
      "loss": 0.148368239402771,
      "mean_token_accuracy": 0.9601317495107651,
      "num_tokens": 1622016.0,
      "step": 99
    },
    {
      "entropy": 4.24693500995636,
      "epoch": 0.2962962962962963,
      "grad_norm": 0.9475519061088562,
      "learning_rate": 7.177177177177178e-06,
      "loss": 0.19721324741840363,
      "mean_token_accuracy": 0.9453341886401176,
      "num_tokens": 1638400.0,
      "step": 100
    },
    {
      "epoch": 0.2962962962962963,
      "eval_entropy": 4.47159772237142,
      "eval_loss": 0.21661478281021118,
      "eval_mean_token_accuracy": 0.9429468663533529,
      "eval_num_tokens": 1638400.0,
      "eval_runtime": 41.7655,
      "eval_samples_per_second": 14.366,
      "eval_steps_per_second": 1.796,
      "step": 100
    },
    {
      "entropy": 4.207014352083206,
      "epoch": 0.2992592592592593,
      "grad_norm": 0.9693189263343811,
      "learning_rate": 7.147147147147148e-06,
      "loss": 0.14343483746051788,
      "mean_token_accuracy": 0.9589961469173431,
      "num_tokens": 1654784.0,
      "step": 101
    },
    {
      "entropy": 4.128320157527924,
      "epoch": 0.3022222222222222,
      "grad_norm": 1.1413607597351074,
      "learning_rate": 7.117117117117117e-06,
      "loss": 0.18779467046260834,
      "mean_token_accuracy": 0.9469092786312103,
      "num_tokens": 1671168.0,
      "step": 102
    },
    {
      "entropy": 4.3105476796627045,
      "epoch": 0.30518518518518517,
      "grad_norm": 1.3886735439300537,
      "learning_rate": 7.087087087087087e-06,
      "loss": 0.254844605922699,
      "mean_token_accuracy": 0.9183749556541443,
      "num_tokens": 1687552.0,
      "step": 103
    },
    {
      "entropy": 4.731838554143906,
      "epoch": 0.30814814814814817,
      "grad_norm": 1.4717377424240112,
      "learning_rate": 7.057057057057057e-06,
      "loss": 0.34209197759628296,
      "mean_token_accuracy": 0.919413834810257,
      "num_tokens": 1703936.0,
      "step": 104
    },
    {
      "entropy": 4.23370036482811,
      "epoch": 0.3111111111111111,
      "grad_norm": 1.0667237043380737,
      "learning_rate": 7.027027027027028e-06,
      "loss": 0.12592408061027527,
      "mean_token_accuracy": 0.9572960957884789,
      "num_tokens": 1720320.0,
      "step": 105
    },
    {
      "entropy": 4.197715878486633,
      "epoch": 0.31407407407407406,
      "grad_norm": 1.2284098863601685,
      "learning_rate": 6.996996996996997e-06,
      "loss": 0.23633553087711334,
      "mean_token_accuracy": 0.9376436546444893,
      "num_tokens": 1736704.0,
      "step": 106
    },
    {
      "entropy": 4.500910848379135,
      "epoch": 0.31703703703703706,
      "grad_norm": 1.0759525299072266,
      "learning_rate": 6.966966966966967e-06,
      "loss": 0.14099782705307007,
      "mean_token_accuracy": 0.9613549783825874,
      "num_tokens": 1753088.0,
      "step": 107
    },
    {
      "entropy": 4.171265602111816,
      "epoch": 0.32,
      "grad_norm": 1.0097358226776123,
      "learning_rate": 6.936936936936938e-06,
      "loss": 0.2070591151714325,
      "mean_token_accuracy": 0.9426354467868805,
      "num_tokens": 1769472.0,
      "step": 108
    },
    {
      "entropy": 4.935701251029968,
      "epoch": 0.32296296296296295,
      "grad_norm": 1.482297420501709,
      "learning_rate": 6.906906906906907e-06,
      "loss": 0.30159881711006165,
      "mean_token_accuracy": 0.9308353662490845,
      "num_tokens": 1785856.0,
      "step": 109
    },
    {
      "entropy": 4.6950942277908325,
      "epoch": 0.32592592592592595,
      "grad_norm": 1.1101776361465454,
      "learning_rate": 6.876876876876878e-06,
      "loss": 0.24596142768859863,
      "mean_token_accuracy": 0.9424160867929459,
      "num_tokens": 1802240.0,
      "step": 110
    },
    {
      "entropy": 4.661284983158112,
      "epoch": 0.3288888888888889,
      "grad_norm": 1.3079828023910522,
      "learning_rate": 6.846846846846848e-06,
      "loss": 0.22717754542827606,
      "mean_token_accuracy": 0.9431795179843903,
      "num_tokens": 1818624.0,
      "step": 111
    },
    {
      "entropy": 4.5867330729961395,
      "epoch": 0.33185185185185184,
      "grad_norm": 1.2115287780761719,
      "learning_rate": 6.816816816816817e-06,
      "loss": 0.18738849461078644,
      "mean_token_accuracy": 0.948810413479805,
      "num_tokens": 1835008.0,
      "step": 112
    },
    {
      "entropy": 4.569830626249313,
      "epoch": 0.3348148148148148,
      "grad_norm": 1.156156301498413,
      "learning_rate": 6.786786786786788e-06,
      "loss": 0.20378893613815308,
      "mean_token_accuracy": 0.9407030493021011,
      "num_tokens": 1851392.0,
      "step": 113
    },
    {
      "entropy": 4.344174385070801,
      "epoch": 0.3377777777777778,
      "grad_norm": 1.1302129030227661,
      "learning_rate": 6.7567567567567575e-06,
      "loss": 0.24610473215579987,
      "mean_token_accuracy": 0.9334333911538124,
      "num_tokens": 1867776.0,
      "step": 114
    },
    {
      "entropy": 4.4949764013290405,
      "epoch": 0.34074074074074073,
      "grad_norm": 1.3018417358398438,
      "learning_rate": 6.726726726726728e-06,
      "loss": 0.25920167565345764,
      "mean_token_accuracy": 0.9228880256414413,
      "num_tokens": 1884160.0,
      "step": 115
    },
    {
      "entropy": 4.1022640615701675,
      "epoch": 0.3437037037037037,
      "grad_norm": 0.9238266348838806,
      "learning_rate": 6.696696696696697e-06,
      "loss": 0.16050559282302856,
      "mean_token_accuracy": 0.9570950344204903,
      "num_tokens": 1900544.0,
      "step": 116
    },
    {
      "entropy": 4.351706147193909,
      "epoch": 0.3466666666666667,
      "grad_norm": 1.0749512910842896,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.2006720006465912,
      "mean_token_accuracy": 0.9449799135327339,
      "num_tokens": 1916928.0,
      "step": 117
    },
    {
      "entropy": 4.2286472618579865,
      "epoch": 0.3496296296296296,
      "grad_norm": 0.9978274703025818,
      "learning_rate": 6.636636636636637e-06,
      "loss": 0.22208011150360107,
      "mean_token_accuracy": 0.9330077543854713,
      "num_tokens": 1933312.0,
      "step": 118
    },
    {
      "entropy": 4.929813742637634,
      "epoch": 0.35259259259259257,
      "grad_norm": 1.2369122505187988,
      "learning_rate": 6.606606606606607e-06,
      "loss": 0.24996542930603027,
      "mean_token_accuracy": 0.9298099726438522,
      "num_tokens": 1949696.0,
      "step": 119
    },
    {
      "entropy": 4.319897681474686,
      "epoch": 0.35555555555555557,
      "grad_norm": 0.9699567556381226,
      "learning_rate": 6.5765765765765775e-06,
      "loss": 0.1806352436542511,
      "mean_token_accuracy": 0.9387884959578514,
      "num_tokens": 1966080.0,
      "step": 120
    },
    {
      "entropy": 4.7684431076049805,
      "epoch": 0.3585185185185185,
      "grad_norm": 1.1323121786117554,
      "learning_rate": 6.546546546546547e-06,
      "loss": 0.20474456250667572,
      "mean_token_accuracy": 0.9389514252543449,
      "num_tokens": 1982464.0,
      "step": 121
    },
    {
      "entropy": 4.4029476046562195,
      "epoch": 0.36148148148148146,
      "grad_norm": 0.9564874768257141,
      "learning_rate": 6.516516516516517e-06,
      "loss": 0.14071759581565857,
      "mean_token_accuracy": 0.9662424102425575,
      "num_tokens": 1998848.0,
      "step": 122
    },
    {
      "entropy": 4.544703960418701,
      "epoch": 0.36444444444444446,
      "grad_norm": 1.0205421447753906,
      "learning_rate": 6.486486486486487e-06,
      "loss": 0.1820712685585022,
      "mean_token_accuracy": 0.9464434757828712,
      "num_tokens": 2015232.0,
      "step": 123
    },
    {
      "entropy": 4.292020469903946,
      "epoch": 0.3674074074074074,
      "grad_norm": 0.7635194063186646,
      "learning_rate": 6.456456456456457e-06,
      "loss": 0.12467391788959503,
      "mean_token_accuracy": 0.95951908826828,
      "num_tokens": 2031616.0,
      "step": 124
    },
    {
      "entropy": 3.9605874121189117,
      "epoch": 0.37037037037037035,
      "grad_norm": 0.8745735883712769,
      "learning_rate": 6.426426426426427e-06,
      "loss": 0.15197420120239258,
      "mean_token_accuracy": 0.9583142548799515,
      "num_tokens": 2048000.0,
      "step": 125
    },
    {
      "entropy": 4.56223464012146,
      "epoch": 0.37333333333333335,
      "grad_norm": 0.9845564961433411,
      "learning_rate": 6.396396396396397e-06,
      "loss": 0.1512736976146698,
      "mean_token_accuracy": 0.9498548060655594,
      "num_tokens": 2064384.0,
      "step": 126
    },
    {
      "entropy": 4.794152081012726,
      "epoch": 0.3762962962962963,
      "grad_norm": 1.3261559009552002,
      "learning_rate": 6.366366366366366e-06,
      "loss": 0.22854501008987427,
      "mean_token_accuracy": 0.9319910854101181,
      "num_tokens": 2080768.0,
      "step": 127
    },
    {
      "entropy": 4.63511598110199,
      "epoch": 0.37925925925925924,
      "grad_norm": 1.079321026802063,
      "learning_rate": 6.336336336336338e-06,
      "loss": 0.1720743030309677,
      "mean_token_accuracy": 0.957709938287735,
      "num_tokens": 2097152.0,
      "step": 128
    },
    {
      "entropy": 4.407420843839645,
      "epoch": 0.38222222222222224,
      "grad_norm": 1.0738903284072876,
      "learning_rate": 6.3063063063063065e-06,
      "loss": 0.2169407606124878,
      "mean_token_accuracy": 0.934087373316288,
      "num_tokens": 2113536.0,
      "step": 129
    },
    {
      "entropy": 4.5498992800712585,
      "epoch": 0.3851851851851852,
      "grad_norm": 1.213025689125061,
      "learning_rate": 6.276276276276276e-06,
      "loss": 0.21649494767189026,
      "mean_token_accuracy": 0.9468272924423218,
      "num_tokens": 2129920.0,
      "step": 130
    },
    {
      "entropy": 4.264434903860092,
      "epoch": 0.38814814814814813,
      "grad_norm": 1.1310464143753052,
      "learning_rate": 6.246246246246247e-06,
      "loss": 0.2593050003051758,
      "mean_token_accuracy": 0.9187903478741646,
      "num_tokens": 2146304.0,
      "step": 131
    },
    {
      "entropy": 4.4609885811805725,
      "epoch": 0.39111111111111113,
      "grad_norm": 1.2307900190353394,
      "learning_rate": 6.2162162162162164e-06,
      "loss": 0.20887784659862518,
      "mean_token_accuracy": 0.9533038958907127,
      "num_tokens": 2162688.0,
      "step": 132
    },
    {
      "entropy": 4.1560070514678955,
      "epoch": 0.3940740740740741,
      "grad_norm": 0.9243044257164001,
      "learning_rate": 6.186186186186187e-06,
      "loss": 0.1576833426952362,
      "mean_token_accuracy": 0.9526226669549942,
      "num_tokens": 2179072.0,
      "step": 133
    },
    {
      "entropy": 4.444341331720352,
      "epoch": 0.397037037037037,
      "grad_norm": 1.1948330402374268,
      "learning_rate": 6.156156156156157e-06,
      "loss": 0.19747555255889893,
      "mean_token_accuracy": 0.9430738091468811,
      "num_tokens": 2195456.0,
      "step": 134
    },
    {
      "entropy": 4.401091754436493,
      "epoch": 0.4,
      "grad_norm": 1.1566073894500732,
      "learning_rate": 6.126126126126126e-06,
      "loss": 0.21883514523506165,
      "mean_token_accuracy": 0.9299560263752937,
      "num_tokens": 2211840.0,
      "step": 135
    },
    {
      "entropy": 4.302798196673393,
      "epoch": 0.40296296296296297,
      "grad_norm": 1.0565509796142578,
      "learning_rate": 6.096096096096097e-06,
      "loss": 0.19299334287643433,
      "mean_token_accuracy": 0.9438915625214577,
      "num_tokens": 2228224.0,
      "step": 136
    },
    {
      "entropy": 4.350885450839996,
      "epoch": 0.4059259259259259,
      "grad_norm": 1.1246678829193115,
      "learning_rate": 6.066066066066067e-06,
      "loss": 0.19274839758872986,
      "mean_token_accuracy": 0.9512739032506943,
      "num_tokens": 2244608.0,
      "step": 137
    },
    {
      "entropy": 4.625757545232773,
      "epoch": 0.4088888888888889,
      "grad_norm": 1.2307966947555542,
      "learning_rate": 6.036036036036037e-06,
      "loss": 0.20485444366931915,
      "mean_token_accuracy": 0.9429539069533348,
      "num_tokens": 2260992.0,
      "step": 138
    },
    {
      "entropy": 4.32535794377327,
      "epoch": 0.41185185185185186,
      "grad_norm": 1.2642028331756592,
      "learning_rate": 6.006006006006007e-06,
      "loss": 0.25205034017562866,
      "mean_token_accuracy": 0.932308979332447,
      "num_tokens": 2277376.0,
      "step": 139
    },
    {
      "entropy": 4.17849925160408,
      "epoch": 0.4148148148148148,
      "grad_norm": 0.955399215221405,
      "learning_rate": 5.975975975975976e-06,
      "loss": 0.1536632478237152,
      "mean_token_accuracy": 0.9589925780892372,
      "num_tokens": 2293760.0,
      "step": 140
    },
    {
      "entropy": 4.372272342443466,
      "epoch": 0.4177777777777778,
      "grad_norm": 0.9778451919555664,
      "learning_rate": 5.945945945945947e-06,
      "loss": 0.1569301038980484,
      "mean_token_accuracy": 0.9550120830535889,
      "num_tokens": 2310144.0,
      "step": 141
    },
    {
      "entropy": 4.745310127735138,
      "epoch": 0.42074074074074075,
      "grad_norm": 1.2513574361801147,
      "learning_rate": 5.915915915915916e-06,
      "loss": 0.2559147775173187,
      "mean_token_accuracy": 0.9207377508282661,
      "num_tokens": 2326528.0,
      "step": 142
    },
    {
      "entropy": 4.450577944517136,
      "epoch": 0.4237037037037037,
      "grad_norm": 1.0372724533081055,
      "learning_rate": 5.885885885885886e-06,
      "loss": 0.15306635200977325,
      "mean_token_accuracy": 0.9559603407979012,
      "num_tokens": 2342912.0,
      "step": 143
    },
    {
      "entropy": 4.487079828977585,
      "epoch": 0.4266666666666667,
      "grad_norm": 1.4338313341140747,
      "learning_rate": 5.855855855855856e-06,
      "loss": 0.30096203088760376,
      "mean_token_accuracy": 0.9216851592063904,
      "num_tokens": 2359296.0,
      "step": 144
    },
    {
      "entropy": 4.670129269361496,
      "epoch": 0.42962962962962964,
      "grad_norm": 1.4709210395812988,
      "learning_rate": 5.825825825825826e-06,
      "loss": 0.3111795485019684,
      "mean_token_accuracy": 0.9231788516044617,
      "num_tokens": 2375680.0,
      "step": 145
    },
    {
      "entropy": 3.8934575021266937,
      "epoch": 0.4325925925925926,
      "grad_norm": 0.8137439489364624,
      "learning_rate": 5.7957957957957965e-06,
      "loss": 0.10548266023397446,
      "mean_token_accuracy": 0.9705552309751511,
      "num_tokens": 2392064.0,
      "step": 146
    },
    {
      "entropy": 4.808850020170212,
      "epoch": 0.43555555555555553,
      "grad_norm": 1.1599018573760986,
      "learning_rate": 5.765765765765766e-06,
      "loss": 0.2138221561908722,
      "mean_token_accuracy": 0.9295399561524391,
      "num_tokens": 2408448.0,
      "step": 147
    },
    {
      "entropy": 4.4261142164468765,
      "epoch": 0.43851851851851853,
      "grad_norm": 1.013484239578247,
      "learning_rate": 5.735735735735736e-06,
      "loss": 0.1929900050163269,
      "mean_token_accuracy": 0.9362559765577316,
      "num_tokens": 2424832.0,
      "step": 148
    },
    {
      "entropy": 4.577237486839294,
      "epoch": 0.4414814814814815,
      "grad_norm": 0.987130343914032,
      "learning_rate": 5.7057057057057065e-06,
      "loss": 0.15615130960941315,
      "mean_token_accuracy": 0.9573905169963837,
      "num_tokens": 2441216.0,
      "step": 149
    },
    {
      "entropy": 4.548634827136993,
      "epoch": 0.4444444444444444,
      "grad_norm": 0.9747557640075684,
      "learning_rate": 5.675675675675676e-06,
      "loss": 0.13734684884548187,
      "mean_token_accuracy": 0.9566464349627495,
      "num_tokens": 2457600.0,
      "step": 150
    },
    {
      "epoch": 0.4444444444444444,
      "eval_entropy": 4.47318754196167,
      "eval_loss": 0.2031925767660141,
      "eval_mean_token_accuracy": 0.9450729378064473,
      "eval_num_tokens": 2457600.0,
      "eval_runtime": 41.7974,
      "eval_samples_per_second": 14.355,
      "eval_steps_per_second": 1.794,
      "step": 150
    },
    {
      "entropy": 4.4322622418403625,
      "epoch": 0.4474074074074074,
      "grad_norm": 1.0352574586868286,
      "learning_rate": 5.645645645645647e-06,
      "loss": 0.16834309697151184,
      "mean_token_accuracy": 0.957368515431881,
      "num_tokens": 2473984.0,
      "step": 151
    },
    {
      "entropy": 4.8277173936367035,
      "epoch": 0.45037037037037037,
      "grad_norm": 1.1564394235610962,
      "learning_rate": 5.615615615615616e-06,
      "loss": 0.19843925535678864,
      "mean_token_accuracy": 0.9456110745668411,
      "num_tokens": 2490368.0,
      "step": 152
    },
    {
      "entropy": 4.3622554540634155,
      "epoch": 0.4533333333333333,
      "grad_norm": 1.0009597539901733,
      "learning_rate": 5.585585585585585e-06,
      "loss": 0.15888887643814087,
      "mean_token_accuracy": 0.9618657007813454,
      "num_tokens": 2506752.0,
      "step": 153
    },
    {
      "entropy": 4.047450542449951,
      "epoch": 0.4562962962962963,
      "grad_norm": 0.8274773955345154,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.10196447372436523,
      "mean_token_accuracy": 0.9689462706446648,
      "num_tokens": 2523136.0,
      "step": 154
    },
    {
      "entropy": 4.7075022757053375,
      "epoch": 0.45925925925925926,
      "grad_norm": 1.2134307622909546,
      "learning_rate": 5.5255255255255255e-06,
      "loss": 0.20454266667366028,
      "mean_token_accuracy": 0.9368318691849709,
      "num_tokens": 2539520.0,
      "step": 155
    },
    {
      "entropy": 4.807535111904144,
      "epoch": 0.4622222222222222,
      "grad_norm": 1.235094666481018,
      "learning_rate": 5.495495495495496e-06,
      "loss": 0.20193201303482056,
      "mean_token_accuracy": 0.9449198022484779,
      "num_tokens": 2555904.0,
      "step": 156
    },
    {
      "entropy": 4.4695963859558105,
      "epoch": 0.4651851851851852,
      "grad_norm": 1.0803649425506592,
      "learning_rate": 5.465465465465466e-06,
      "loss": 0.17315573990345,
      "mean_token_accuracy": 0.947208046913147,
      "num_tokens": 2572288.0,
      "step": 157
    },
    {
      "entropy": 4.616615951061249,
      "epoch": 0.46814814814814815,
      "grad_norm": 1.227483868598938,
      "learning_rate": 5.4354354354354355e-06,
      "loss": 0.23030991852283478,
      "mean_token_accuracy": 0.9416991248726845,
      "num_tokens": 2588672.0,
      "step": 158
    },
    {
      "entropy": 4.387355536222458,
      "epoch": 0.4711111111111111,
      "grad_norm": 0.8713464140892029,
      "learning_rate": 5.405405405405406e-06,
      "loss": 0.1236598938703537,
      "mean_token_accuracy": 0.9607251882553101,
      "num_tokens": 2605056.0,
      "step": 159
    },
    {
      "entropy": 5.015526354312897,
      "epoch": 0.4740740740740741,
      "grad_norm": 1.4982430934906006,
      "learning_rate": 5.375375375375376e-06,
      "loss": 0.31277599930763245,
      "mean_token_accuracy": 0.9063860177993774,
      "num_tokens": 2621440.0,
      "step": 160
    },
    {
      "entropy": 4.156640291213989,
      "epoch": 0.47703703703703704,
      "grad_norm": 1.0972727537155151,
      "learning_rate": 5.345345345345346e-06,
      "loss": 0.21312345564365387,
      "mean_token_accuracy": 0.953573189675808,
      "num_tokens": 2637824.0,
      "step": 161
    },
    {
      "entropy": 4.109684631228447,
      "epoch": 0.48,
      "grad_norm": 1.2694456577301025,
      "learning_rate": 5.315315315315316e-06,
      "loss": 0.16456085443496704,
      "mean_token_accuracy": 0.9567013755440712,
      "num_tokens": 2654208.0,
      "step": 162
    },
    {
      "entropy": 4.538438081741333,
      "epoch": 0.482962962962963,
      "grad_norm": 1.1064934730529785,
      "learning_rate": 5.285285285285286e-06,
      "loss": 0.2004067748785019,
      "mean_token_accuracy": 0.9418867826461792,
      "num_tokens": 2670592.0,
      "step": 163
    },
    {
      "entropy": 4.703399509191513,
      "epoch": 0.48592592592592593,
      "grad_norm": 1.2677124738693237,
      "learning_rate": 5.255255255255256e-06,
      "loss": 0.2093324512243271,
      "mean_token_accuracy": 0.9343851506710052,
      "num_tokens": 2686976.0,
      "step": 164
    },
    {
      "entropy": 4.521232604980469,
      "epoch": 0.4888888888888889,
      "grad_norm": 0.9266942739486694,
      "learning_rate": 5.225225225225226e-06,
      "loss": 0.12813670933246613,
      "mean_token_accuracy": 0.9629350677132607,
      "num_tokens": 2703360.0,
      "step": 165
    },
    {
      "entropy": 4.625990688800812,
      "epoch": 0.4918518518518519,
      "grad_norm": 1.1284843683242798,
      "learning_rate": 5.195195195195195e-06,
      "loss": 0.16044031083583832,
      "mean_token_accuracy": 0.9474668651819229,
      "num_tokens": 2719744.0,
      "step": 166
    },
    {
      "entropy": 4.294618993997574,
      "epoch": 0.4948148148148148,
      "grad_norm": 1.0471121072769165,
      "learning_rate": 5.165165165165165e-06,
      "loss": 0.1588296741247177,
      "mean_token_accuracy": 0.9490892142057419,
      "num_tokens": 2736128.0,
      "step": 167
    },
    {
      "entropy": 3.69912526011467,
      "epoch": 0.49777777777777776,
      "grad_norm": 0.7754703164100647,
      "learning_rate": 5.135135135135135e-06,
      "loss": 0.09100496768951416,
      "mean_token_accuracy": 0.9716495275497437,
      "num_tokens": 2752512.0,
      "step": 168
    },
    {
      "entropy": 4.571734249591827,
      "epoch": 0.5007407407407407,
      "grad_norm": 0.9708716869354248,
      "learning_rate": 5.105105105105106e-06,
      "loss": 0.11646466702222824,
      "mean_token_accuracy": 0.9648235365748405,
      "num_tokens": 2768896.0,
      "step": 169
    },
    {
      "entropy": 4.400004804134369,
      "epoch": 0.5037037037037037,
      "grad_norm": 1.1822189092636108,
      "learning_rate": 5.075075075075075e-06,
      "loss": 0.21681709587574005,
      "mean_token_accuracy": 0.9444219768047333,
      "num_tokens": 2785280.0,
      "step": 170
    },
    {
      "entropy": 5.048867702484131,
      "epoch": 0.5066666666666667,
      "grad_norm": 1.394973635673523,
      "learning_rate": 5.045045045045045e-06,
      "loss": 0.22330451011657715,
      "mean_token_accuracy": 0.9336277917027473,
      "num_tokens": 2801664.0,
      "step": 171
    },
    {
      "entropy": 4.8033424615859985,
      "epoch": 0.5096296296296297,
      "grad_norm": 1.3380451202392578,
      "learning_rate": 5.0150150150150156e-06,
      "loss": 0.22821509838104248,
      "mean_token_accuracy": 0.9419275671243668,
      "num_tokens": 2818048.0,
      "step": 172
    },
    {
      "entropy": 5.28538578748703,
      "epoch": 0.5125925925925926,
      "grad_norm": 1.3695138692855835,
      "learning_rate": 4.984984984984985e-06,
      "loss": 0.30885204672813416,
      "mean_token_accuracy": 0.9175591617822647,
      "num_tokens": 2834432.0,
      "step": 173
    },
    {
      "entropy": 4.509493857622147,
      "epoch": 0.5155555555555555,
      "grad_norm": 1.2185577154159546,
      "learning_rate": 4.954954954954955e-06,
      "loss": 0.1723558008670807,
      "mean_token_accuracy": 0.9549972116947174,
      "num_tokens": 2850816.0,
      "step": 174
    },
    {
      "entropy": 4.428200513124466,
      "epoch": 0.5185185185185185,
      "grad_norm": 1.0344725847244263,
      "learning_rate": 4.9249249249249255e-06,
      "loss": 0.18210822343826294,
      "mean_token_accuracy": 0.9371201321482658,
      "num_tokens": 2867200.0,
      "step": 175
    },
    {
      "entropy": 4.199506938457489,
      "epoch": 0.5214814814814814,
      "grad_norm": 0.9780181050300598,
      "learning_rate": 4.894894894894895e-06,
      "loss": 0.15261413156986237,
      "mean_token_accuracy": 0.9657870754599571,
      "num_tokens": 2883584.0,
      "step": 176
    },
    {
      "entropy": 4.598901391029358,
      "epoch": 0.5244444444444445,
      "grad_norm": 1.1018935441970825,
      "learning_rate": 4.864864864864866e-06,
      "loss": 0.13766585290431976,
      "mean_token_accuracy": 0.9611618220806122,
      "num_tokens": 2899968.0,
      "step": 177
    },
    {
      "entropy": 4.50525039434433,
      "epoch": 0.5274074074074074,
      "grad_norm": 1.0960816144943237,
      "learning_rate": 4.8348348348348355e-06,
      "loss": 0.19349969923496246,
      "mean_token_accuracy": 0.9485724493861198,
      "num_tokens": 2916352.0,
      "step": 178
    },
    {
      "entropy": 4.279501169919968,
      "epoch": 0.5303703703703704,
      "grad_norm": 0.9863625764846802,
      "learning_rate": 4.804804804804805e-06,
      "loss": 0.2030724287033081,
      "mean_token_accuracy": 0.9458435997366905,
      "num_tokens": 2932736.0,
      "step": 179
    },
    {
      "entropy": 5.122234761714935,
      "epoch": 0.5333333333333333,
      "grad_norm": 1.640047550201416,
      "learning_rate": 4.774774774774775e-06,
      "loss": 0.29480722546577454,
      "mean_token_accuracy": 0.9210398867726326,
      "num_tokens": 2949120.0,
      "step": 180
    },
    {
      "entropy": 4.030216574668884,
      "epoch": 0.5362962962962963,
      "grad_norm": 0.7872567772865295,
      "learning_rate": 4.7447447447447454e-06,
      "loss": 0.1057211309671402,
      "mean_token_accuracy": 0.9713046625256538,
      "num_tokens": 2965504.0,
      "step": 181
    },
    {
      "entropy": 4.142587408423424,
      "epoch": 0.5392592592592592,
      "grad_norm": 1.0884994268417358,
      "learning_rate": 4.714714714714715e-06,
      "loss": 0.1927669793367386,
      "mean_token_accuracy": 0.9500064626336098,
      "num_tokens": 2981888.0,
      "step": 182
    },
    {
      "entropy": 4.687072217464447,
      "epoch": 0.5422222222222223,
      "grad_norm": 1.2485096454620361,
      "learning_rate": 4.684684684684685e-06,
      "loss": 0.10758813470602036,
      "mean_token_accuracy": 0.962133914232254,
      "num_tokens": 2998272.0,
      "step": 183
    },
    {
      "entropy": 5.1369404792785645,
      "epoch": 0.5451851851851852,
      "grad_norm": 1.3939979076385498,
      "learning_rate": 4.654654654654655e-06,
      "loss": 0.30590641498565674,
      "mean_token_accuracy": 0.918172724545002,
      "num_tokens": 3014656.0,
      "step": 184
    },
    {
      "entropy": 4.4005067348480225,
      "epoch": 0.5481481481481482,
      "grad_norm": 1.104361653327942,
      "learning_rate": 4.624624624624625e-06,
      "loss": 0.16491568088531494,
      "mean_token_accuracy": 0.9591816738247871,
      "num_tokens": 3031040.0,
      "step": 185
    },
    {
      "entropy": 4.227125704288483,
      "epoch": 0.5511111111111111,
      "grad_norm": 0.7754931449890137,
      "learning_rate": 4.594594594594596e-06,
      "loss": 0.08171460777521133,
      "mean_token_accuracy": 0.9774893298745155,
      "num_tokens": 3047424.0,
      "step": 186
    },
    {
      "entropy": 4.507180631160736,
      "epoch": 0.554074074074074,
      "grad_norm": 0.9401944279670715,
      "learning_rate": 4.5645645645645645e-06,
      "loss": 0.14690890908241272,
      "mean_token_accuracy": 0.9507097601890564,
      "num_tokens": 3063808.0,
      "step": 187
    },
    {
      "entropy": 4.42718380689621,
      "epoch": 0.557037037037037,
      "grad_norm": 1.1169674396514893,
      "learning_rate": 4.534534534534535e-06,
      "loss": 0.19810277223587036,
      "mean_token_accuracy": 0.9506569430232048,
      "num_tokens": 3080192.0,
      "step": 188
    },
    {
      "entropy": 4.601956844329834,
      "epoch": 0.56,
      "grad_norm": 1.144074559211731,
      "learning_rate": 4.504504504504505e-06,
      "loss": 0.2215370535850525,
      "mean_token_accuracy": 0.9346054270863533,
      "num_tokens": 3096576.0,
      "step": 189
    },
    {
      "entropy": 4.288637429475784,
      "epoch": 0.562962962962963,
      "grad_norm": 1.078635811805725,
      "learning_rate": 4.474474474474475e-06,
      "loss": 0.16822373867034912,
      "mean_token_accuracy": 0.9575443118810654,
      "num_tokens": 3112960.0,
      "step": 190
    },
    {
      "entropy": 4.341667145490646,
      "epoch": 0.5659259259259259,
      "grad_norm": 0.7250970005989075,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.08381730318069458,
      "mean_token_accuracy": 0.9752118736505508,
      "num_tokens": 3129344.0,
      "step": 191
    },
    {
      "entropy": 5.112811267375946,
      "epoch": 0.5688888888888889,
      "grad_norm": 1.5165245532989502,
      "learning_rate": 4.414414414414415e-06,
      "loss": 0.21411152184009552,
      "mean_token_accuracy": 0.9332496151328087,
      "num_tokens": 3145728.0,
      "step": 192
    },
    {
      "entropy": 5.112403213977814,
      "epoch": 0.5718518518518518,
      "grad_norm": 1.6484229564666748,
      "learning_rate": 4.384384384384384e-06,
      "loss": 0.3096240758895874,
      "mean_token_accuracy": 0.90643060952425,
      "num_tokens": 3162112.0,
      "step": 193
    },
    {
      "entropy": 4.62445867061615,
      "epoch": 0.5748148148148148,
      "grad_norm": 1.1302015781402588,
      "learning_rate": 4.354354354354355e-06,
      "loss": 0.20545849204063416,
      "mean_token_accuracy": 0.9418515488505363,
      "num_tokens": 3178496.0,
      "step": 194
    },
    {
      "entropy": 3.8342756927013397,
      "epoch": 0.5777777777777777,
      "grad_norm": 0.9575071930885315,
      "learning_rate": 4.324324324324325e-06,
      "loss": 0.14582830667495728,
      "mean_token_accuracy": 0.9646051079034805,
      "num_tokens": 3194880.0,
      "step": 195
    },
    {
      "entropy": 4.761662006378174,
      "epoch": 0.5807407407407408,
      "grad_norm": 1.2014862298965454,
      "learning_rate": 4.294294294294294e-06,
      "loss": 0.20800161361694336,
      "mean_token_accuracy": 0.936536893248558,
      "num_tokens": 3211264.0,
      "step": 196
    },
    {
      "entropy": 4.695729613304138,
      "epoch": 0.5837037037037037,
      "grad_norm": 1.216070532798767,
      "learning_rate": 4.264264264264265e-06,
      "loss": 0.19924479722976685,
      "mean_token_accuracy": 0.9393056556582451,
      "num_tokens": 3227648.0,
      "step": 197
    },
    {
      "entropy": 4.349013060331345,
      "epoch": 0.5866666666666667,
      "grad_norm": 1.1972628831863403,
      "learning_rate": 4.234234234234235e-06,
      "loss": 0.1833992451429367,
      "mean_token_accuracy": 0.9455256313085556,
      "num_tokens": 3244032.0,
      "step": 198
    },
    {
      "entropy": 4.20524337887764,
      "epoch": 0.5896296296296296,
      "grad_norm": 1.0359379053115845,
      "learning_rate": 4.204204204204204e-06,
      "loss": 0.16859027743339539,
      "mean_token_accuracy": 0.9528629556298256,
      "num_tokens": 3260416.0,
      "step": 199
    },
    {
      "entropy": 4.638263285160065,
      "epoch": 0.5925925925925926,
      "grad_norm": 1.406262993812561,
      "learning_rate": 4.174174174174174e-06,
      "loss": 0.21396510303020477,
      "mean_token_accuracy": 0.9318485632538795,
      "num_tokens": 3276800.0,
      "step": 200
    },
    {
      "epoch": 0.5925925925925926,
      "eval_entropy": 4.493113050460815,
      "eval_loss": 0.1943897157907486,
      "eval_mean_token_accuracy": 0.9472449056307475,
      "eval_num_tokens": 3276800.0,
      "eval_runtime": 41.7961,
      "eval_samples_per_second": 14.355,
      "eval_steps_per_second": 1.794,
      "step": 200
    },
    {
      "entropy": 4.721972018480301,
      "epoch": 0.5955555555555555,
      "grad_norm": 1.2206627130508423,
      "learning_rate": 4.1441441441441446e-06,
      "loss": 0.26079389452934265,
      "mean_token_accuracy": 0.9290014132857323,
      "num_tokens": 3293184.0,
      "step": 201
    },
    {
      "entropy": 4.651299446821213,
      "epoch": 0.5985185185185186,
      "grad_norm": 1.0173994302749634,
      "learning_rate": 4.114114114114114e-06,
      "loss": 0.17843106389045715,
      "mean_token_accuracy": 0.9478134959936142,
      "num_tokens": 3309568.0,
      "step": 202
    },
    {
      "entropy": 4.191153109073639,
      "epoch": 0.6014814814814815,
      "grad_norm": 1.1586952209472656,
      "learning_rate": 4.084084084084085e-06,
      "loss": 0.20614223182201385,
      "mean_token_accuracy": 0.9499139413237572,
      "num_tokens": 3325952.0,
      "step": 203
    },
    {
      "entropy": 5.034147620201111,
      "epoch": 0.6044444444444445,
      "grad_norm": 1.3782150745391846,
      "learning_rate": 4.0540540540540545e-06,
      "loss": 0.320191353559494,
      "mean_token_accuracy": 0.9129593074321747,
      "num_tokens": 3342336.0,
      "step": 204
    },
    {
      "entropy": 4.099581271409988,
      "epoch": 0.6074074074074074,
      "grad_norm": 0.9795379042625427,
      "learning_rate": 4.024024024024024e-06,
      "loss": 0.17633658647537231,
      "mean_token_accuracy": 0.9418376907706261,
      "num_tokens": 3358720.0,
      "step": 205
    },
    {
      "entropy": 4.193884015083313,
      "epoch": 0.6103703703703703,
      "grad_norm": 1.0334500074386597,
      "learning_rate": 3.993993993993994e-06,
      "loss": 0.12188932299613953,
      "mean_token_accuracy": 0.9651204645633698,
      "num_tokens": 3375104.0,
      "step": 206
    },
    {
      "entropy": 4.6628334522247314,
      "epoch": 0.6133333333333333,
      "grad_norm": 1.1743429899215698,
      "learning_rate": 3.9639639639639645e-06,
      "loss": 0.24152298271656036,
      "mean_token_accuracy": 0.9141566455364227,
      "num_tokens": 3391488.0,
      "step": 207
    },
    {
      "entropy": 4.851721405982971,
      "epoch": 0.6162962962962963,
      "grad_norm": 1.2436100244522095,
      "learning_rate": 3.933933933933934e-06,
      "loss": 0.2596300542354584,
      "mean_token_accuracy": 0.9323165863752365,
      "num_tokens": 3407872.0,
      "step": 208
    },
    {
      "entropy": 4.460325717926025,
      "epoch": 0.6192592592592593,
      "grad_norm": 1.0915201902389526,
      "learning_rate": 3.903903903903904e-06,
      "loss": 0.22234000265598297,
      "mean_token_accuracy": 0.9414825439453125,
      "num_tokens": 3424256.0,
      "step": 209
    },
    {
      "entropy": 4.80398628115654,
      "epoch": 0.6222222222222222,
      "grad_norm": 1.035743236541748,
      "learning_rate": 3.8738738738738744e-06,
      "loss": 0.14587096869945526,
      "mean_token_accuracy": 0.9518779441714287,
      "num_tokens": 3440640.0,
      "step": 210
    },
    {
      "entropy": 3.747168242931366,
      "epoch": 0.6251851851851852,
      "grad_norm": 1.6041151285171509,
      "learning_rate": 3.843843843843844e-06,
      "loss": 0.16117823123931885,
      "mean_token_accuracy": 0.9578864499926567,
      "num_tokens": 3457024.0,
      "step": 211
    },
    {
      "entropy": 4.582811236381531,
      "epoch": 0.6281481481481481,
      "grad_norm": 1.146173119544983,
      "learning_rate": 3.8138138138138143e-06,
      "loss": 0.1815604269504547,
      "mean_token_accuracy": 0.9542162492871284,
      "num_tokens": 3473408.0,
      "step": 212
    },
    {
      "entropy": 4.734657645225525,
      "epoch": 0.6311111111111111,
      "grad_norm": 1.439591884613037,
      "learning_rate": 3.7837837837837844e-06,
      "loss": 0.2559170126914978,
      "mean_token_accuracy": 0.9301478862762451,
      "num_tokens": 3489792.0,
      "step": 213
    },
    {
      "entropy": 4.904492408037186,
      "epoch": 0.6340740740740741,
      "grad_norm": 1.5149108171463013,
      "learning_rate": 3.7537537537537537e-06,
      "loss": 0.2521807849407196,
      "mean_token_accuracy": 0.9222758710384369,
      "num_tokens": 3506176.0,
      "step": 214
    },
    {
      "entropy": 4.850836426019669,
      "epoch": 0.6370370370370371,
      "grad_norm": 1.2862434387207031,
      "learning_rate": 3.723723723723724e-06,
      "loss": 0.2918914258480072,
      "mean_token_accuracy": 0.916606992483139,
      "num_tokens": 3522560.0,
      "step": 215
    },
    {
      "entropy": 4.7798881232738495,
      "epoch": 0.64,
      "grad_norm": 1.3682328462600708,
      "learning_rate": 3.693693693693694e-06,
      "loss": 0.27479875087738037,
      "mean_token_accuracy": 0.9360896050930023,
      "num_tokens": 3538944.0,
      "step": 216
    },
    {
      "entropy": 4.9106133580207825,
      "epoch": 0.642962962962963,
      "grad_norm": 1.1826263666152954,
      "learning_rate": 3.663663663663664e-06,
      "loss": 0.22357432544231415,
      "mean_token_accuracy": 0.9431413114070892,
      "num_tokens": 3555328.0,
      "step": 217
    },
    {
      "entropy": 4.957390755414963,
      "epoch": 0.6459259259259259,
      "grad_norm": 1.407552719116211,
      "learning_rate": 3.633633633633634e-06,
      "loss": 0.2573108971118927,
      "mean_token_accuracy": 0.9116180762648582,
      "num_tokens": 3571712.0,
      "step": 218
    },
    {
      "entropy": 4.5682536363601685,
      "epoch": 0.6488888888888888,
      "grad_norm": 1.198503017425537,
      "learning_rate": 3.603603603603604e-06,
      "loss": 0.20289891958236694,
      "mean_token_accuracy": 0.9426070526242256,
      "num_tokens": 3588096.0,
      "step": 219
    },
    {
      "entropy": 4.498745650053024,
      "epoch": 0.6518518518518519,
      "grad_norm": 1.0328853130340576,
      "learning_rate": 3.573573573573574e-06,
      "loss": 0.165367990732193,
      "mean_token_accuracy": 0.9520335793495178,
      "num_tokens": 3604480.0,
      "step": 220
    },
    {
      "entropy": 4.466074496507645,
      "epoch": 0.6548148148148148,
      "grad_norm": 1.149162769317627,
      "learning_rate": 3.5435435435435437e-06,
      "loss": 0.17273250222206116,
      "mean_token_accuracy": 0.949529230594635,
      "num_tokens": 3620864.0,
      "step": 221
    },
    {
      "entropy": 5.377149283885956,
      "epoch": 0.6577777777777778,
      "grad_norm": 1.4259369373321533,
      "learning_rate": 3.513513513513514e-06,
      "loss": 0.26014643907546997,
      "mean_token_accuracy": 0.9218437299132347,
      "num_tokens": 3637248.0,
      "step": 222
    },
    {
      "entropy": 4.532017946243286,
      "epoch": 0.6607407407407407,
      "grad_norm": 0.9938780069351196,
      "learning_rate": 3.4834834834834835e-06,
      "loss": 0.18300899863243103,
      "mean_token_accuracy": 0.9539739042520523,
      "num_tokens": 3653632.0,
      "step": 223
    },
    {
      "entropy": 4.6572571992874146,
      "epoch": 0.6637037037037037,
      "grad_norm": 1.3851807117462158,
      "learning_rate": 3.4534534534534537e-06,
      "loss": 0.21601992845535278,
      "mean_token_accuracy": 0.935951441526413,
      "num_tokens": 3670016.0,
      "step": 224
    },
    {
      "entropy": 4.810268461704254,
      "epoch": 0.6666666666666666,
      "grad_norm": 1.25748610496521,
      "learning_rate": 3.423423423423424e-06,
      "loss": 0.19575180113315582,
      "mean_token_accuracy": 0.9432429894804955,
      "num_tokens": 3686400.0,
      "step": 225
    },
    {
      "entropy": 4.724693238735199,
      "epoch": 0.6696296296296296,
      "grad_norm": 1.1689661741256714,
      "learning_rate": 3.393393393393394e-06,
      "loss": 0.19534890353679657,
      "mean_token_accuracy": 0.9337200000882149,
      "num_tokens": 3702784.0,
      "step": 226
    },
    {
      "entropy": 4.160109043121338,
      "epoch": 0.6725925925925926,
      "grad_norm": 0.9268734455108643,
      "learning_rate": 3.363363363363364e-06,
      "loss": 0.1599588692188263,
      "mean_token_accuracy": 0.9591795429587364,
      "num_tokens": 3719168.0,
      "step": 227
    },
    {
      "entropy": 4.595972537994385,
      "epoch": 0.6755555555555556,
      "grad_norm": 1.1711211204528809,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.14186200499534607,
      "mean_token_accuracy": 0.9561845883727074,
      "num_tokens": 3735552.0,
      "step": 228
    },
    {
      "entropy": 4.655384361743927,
      "epoch": 0.6785185185185185,
      "grad_norm": 1.3928577899932861,
      "learning_rate": 3.3033033033033035e-06,
      "loss": 0.26774126291275024,
      "mean_token_accuracy": 0.9372854456305504,
      "num_tokens": 3751936.0,
      "step": 229
    },
    {
      "entropy": 4.9599509835243225,
      "epoch": 0.6814814814814815,
      "grad_norm": 1.4189996719360352,
      "learning_rate": 3.2732732732732736e-06,
      "loss": 0.24391934275627136,
      "mean_token_accuracy": 0.9354264885187149,
      "num_tokens": 3768320.0,
      "step": 230
    },
    {
      "entropy": 4.2409583032131195,
      "epoch": 0.6844444444444444,
      "grad_norm": 0.8994541168212891,
      "learning_rate": 3.2432432432432437e-06,
      "loss": 0.13000833988189697,
      "mean_token_accuracy": 0.9605458900332451,
      "num_tokens": 3784704.0,
      "step": 231
    },
    {
      "entropy": 5.001427173614502,
      "epoch": 0.6874074074074074,
      "grad_norm": 1.2159807682037354,
      "learning_rate": 3.2132132132132134e-06,
      "loss": 0.22633376717567444,
      "mean_token_accuracy": 0.9305596277117729,
      "num_tokens": 3801088.0,
      "step": 232
    },
    {
      "entropy": 4.597889393568039,
      "epoch": 0.6903703703703704,
      "grad_norm": 1.168084740638733,
      "learning_rate": 3.183183183183183e-06,
      "loss": 0.21986490488052368,
      "mean_token_accuracy": 0.9350046962499619,
      "num_tokens": 3817472.0,
      "step": 233
    },
    {
      "entropy": 4.029878079891205,
      "epoch": 0.6933333333333334,
      "grad_norm": 0.7615580558776855,
      "learning_rate": 3.1531531531531532e-06,
      "loss": 0.08273021876811981,
      "mean_token_accuracy": 0.9739208221435547,
      "num_tokens": 3833856.0,
      "step": 234
    },
    {
      "entropy": 4.588170945644379,
      "epoch": 0.6962962962962963,
      "grad_norm": 1.1632869243621826,
      "learning_rate": 3.1231231231231234e-06,
      "loss": 0.19196635484695435,
      "mean_token_accuracy": 0.9486162513494492,
      "num_tokens": 3850240.0,
      "step": 235
    },
    {
      "entropy": 4.604072004556656,
      "epoch": 0.6992592592592592,
      "grad_norm": 1.2127914428710938,
      "learning_rate": 3.0930930930930935e-06,
      "loss": 0.19198723137378693,
      "mean_token_accuracy": 0.9417206943035126,
      "num_tokens": 3866624.0,
      "step": 236
    },
    {
      "entropy": 4.094269037246704,
      "epoch": 0.7022222222222222,
      "grad_norm": 1.0208652019500732,
      "learning_rate": 3.063063063063063e-06,
      "loss": 0.1620928794145584,
      "mean_token_accuracy": 0.9501628130674362,
      "num_tokens": 3883008.0,
      "step": 237
    },
    {
      "entropy": 4.765289276838303,
      "epoch": 0.7051851851851851,
      "grad_norm": 1.2969844341278076,
      "learning_rate": 3.0330330330330333e-06,
      "loss": 0.2629225254058838,
      "mean_token_accuracy": 0.9354538172483444,
      "num_tokens": 3899392.0,
      "step": 238
    },
    {
      "entropy": 4.631817579269409,
      "epoch": 0.7081481481481482,
      "grad_norm": 1.3574633598327637,
      "learning_rate": 3.0030030030030034e-06,
      "loss": 0.2812841534614563,
      "mean_token_accuracy": 0.9247230440378189,
      "num_tokens": 3915776.0,
      "step": 239
    },
    {
      "entropy": 4.724785268306732,
      "epoch": 0.7111111111111111,
      "grad_norm": 1.6501286029815674,
      "learning_rate": 2.9729729729729736e-06,
      "loss": 0.3619709312915802,
      "mean_token_accuracy": 0.9125356450676918,
      "num_tokens": 3932160.0,
      "step": 240
    },
    {
      "entropy": 4.020456701517105,
      "epoch": 0.7140740740740741,
      "grad_norm": 0.8593155741691589,
      "learning_rate": 2.942942942942943e-06,
      "loss": 0.12883971631526947,
      "mean_token_accuracy": 0.9684963598847389,
      "num_tokens": 3948544.0,
      "step": 241
    },
    {
      "entropy": 4.411629885435104,
      "epoch": 0.717037037037037,
      "grad_norm": 1.3800005912780762,
      "learning_rate": 2.912912912912913e-06,
      "loss": 0.26712891459465027,
      "mean_token_accuracy": 0.9271191209554672,
      "num_tokens": 3964928.0,
      "step": 242
    },
    {
      "entropy": 4.682328164577484,
      "epoch": 0.72,
      "grad_norm": 1.2916998863220215,
      "learning_rate": 2.882882882882883e-06,
      "loss": 0.1764688491821289,
      "mean_token_accuracy": 0.9463462606072426,
      "num_tokens": 3981312.0,
      "step": 243
    },
    {
      "entropy": 4.445810675621033,
      "epoch": 0.7229629629629629,
      "grad_norm": 1.2792280912399292,
      "learning_rate": 2.8528528528528532e-06,
      "loss": 0.21392151713371277,
      "mean_token_accuracy": 0.940187118947506,
      "num_tokens": 3997696.0,
      "step": 244
    },
    {
      "entropy": 4.090399906039238,
      "epoch": 0.725925925925926,
      "grad_norm": 0.9300408363342285,
      "learning_rate": 2.8228228228228234e-06,
      "loss": 0.14719796180725098,
      "mean_token_accuracy": 0.9563947021961212,
      "num_tokens": 4014080.0,
      "step": 245
    },
    {
      "entropy": 4.908781349658966,
      "epoch": 0.7288888888888889,
      "grad_norm": 1.3864506483078003,
      "learning_rate": 2.7927927927927926e-06,
      "loss": 0.32241857051849365,
      "mean_token_accuracy": 0.9273277744650841,
      "num_tokens": 4030464.0,
      "step": 246
    },
    {
      "entropy": 4.440890371799469,
      "epoch": 0.7318518518518519,
      "grad_norm": 1.017882227897644,
      "learning_rate": 2.7627627627627628e-06,
      "loss": 0.1510559320449829,
      "mean_token_accuracy": 0.956509493291378,
      "num_tokens": 4046848.0,
      "step": 247
    },
    {
      "entropy": 4.220070123672485,
      "epoch": 0.7348148148148148,
      "grad_norm": 1.0297763347625732,
      "learning_rate": 2.732732732732733e-06,
      "loss": 0.13601014018058777,
      "mean_token_accuracy": 0.9613732844591141,
      "num_tokens": 4063232.0,
      "step": 248
    },
    {
      "entropy": 4.682860463857651,
      "epoch": 0.7377777777777778,
      "grad_norm": 1.2604784965515137,
      "learning_rate": 2.702702702702703e-06,
      "loss": 0.18585912883281708,
      "mean_token_accuracy": 0.9522142335772514,
      "num_tokens": 4079616.0,
      "step": 249
    },
    {
      "entropy": 4.788404762744904,
      "epoch": 0.7407407407407407,
      "grad_norm": 1.173094391822815,
      "learning_rate": 2.672672672672673e-06,
      "loss": 0.17388513684272766,
      "mean_token_accuracy": 0.9485967382788658,
      "num_tokens": 4096000.0,
      "step": 250
    },
    {
      "epoch": 0.7407407407407407,
      "eval_entropy": 4.462542368570964,
      "eval_loss": 0.18963900208473206,
      "eval_mean_token_accuracy": 0.948217601776123,
      "eval_num_tokens": 4096000.0,
      "eval_runtime": 41.7722,
      "eval_samples_per_second": 14.364,
      "eval_steps_per_second": 1.795,
      "step": 250
    },
    {
      "entropy": 4.589096903800964,
      "epoch": 0.7437037037037038,
      "grad_norm": 1.1348332166671753,
      "learning_rate": 2.642642642642643e-06,
      "loss": 0.16592100262641907,
      "mean_token_accuracy": 0.950493760406971,
      "num_tokens": 4112384.0,
      "step": 251
    },
    {
      "entropy": 4.6115531623363495,
      "epoch": 0.7466666666666667,
      "grad_norm": 1.274220585823059,
      "learning_rate": 2.612612612612613e-06,
      "loss": 0.23926684260368347,
      "mean_token_accuracy": 0.941401943564415,
      "num_tokens": 4128768.0,
      "step": 252
    },
    {
      "entropy": 3.9944489896297455,
      "epoch": 0.7496296296296296,
      "grad_norm": 0.865063488483429,
      "learning_rate": 2.5825825825825827e-06,
      "loss": 0.10147737711668015,
      "mean_token_accuracy": 0.9678801819682121,
      "num_tokens": 4145152.0,
      "step": 253
    },
    {
      "entropy": 3.8060972094535828,
      "epoch": 0.7525925925925926,
      "grad_norm": 0.9605916142463684,
      "learning_rate": 2.552552552552553e-06,
      "loss": 0.15311875939369202,
      "mean_token_accuracy": 0.9615834206342697,
      "num_tokens": 4161536.0,
      "step": 254
    },
    {
      "entropy": 4.757985502481461,
      "epoch": 0.7555555555555555,
      "grad_norm": 1.014810562133789,
      "learning_rate": 2.5225225225225225e-06,
      "loss": 0.16364140808582306,
      "mean_token_accuracy": 0.9574583321809769,
      "num_tokens": 4177920.0,
      "step": 255
    },
    {
      "entropy": 4.692818701267242,
      "epoch": 0.7585185185185185,
      "grad_norm": 1.1628785133361816,
      "learning_rate": 2.4924924924924926e-06,
      "loss": 0.16226403415203094,
      "mean_token_accuracy": 0.9519179984927177,
      "num_tokens": 4194304.0,
      "step": 256
    },
    {
      "entropy": 4.45292067527771,
      "epoch": 0.7614814814814815,
      "grad_norm": 1.0610793828964233,
      "learning_rate": 2.4624624624624628e-06,
      "loss": 0.16758939623832703,
      "mean_token_accuracy": 0.9456373304128647,
      "num_tokens": 4210688.0,
      "step": 257
    },
    {
      "entropy": 4.858905553817749,
      "epoch": 0.7644444444444445,
      "grad_norm": 1.1992101669311523,
      "learning_rate": 2.432432432432433e-06,
      "loss": 0.24088703095912933,
      "mean_token_accuracy": 0.9358232468366623,
      "num_tokens": 4227072.0,
      "step": 258
    },
    {
      "entropy": 3.652865171432495,
      "epoch": 0.7674074074074074,
      "grad_norm": 0.6357795596122742,
      "learning_rate": 2.4024024024024026e-06,
      "loss": 0.06520264595746994,
      "mean_token_accuracy": 0.9755329266190529,
      "num_tokens": 4243456.0,
      "step": 259
    },
    {
      "entropy": 4.366178065538406,
      "epoch": 0.7703703703703704,
      "grad_norm": 1.0377697944641113,
      "learning_rate": 2.3723723723723727e-06,
      "loss": 0.21136701107025146,
      "mean_token_accuracy": 0.9377603009343147,
      "num_tokens": 4259840.0,
      "step": 260
    },
    {
      "entropy": 4.124667555093765,
      "epoch": 0.7733333333333333,
      "grad_norm": 1.0690512657165527,
      "learning_rate": 2.3423423423423424e-06,
      "loss": 0.19677580893039703,
      "mean_token_accuracy": 0.9424155652523041,
      "num_tokens": 4276224.0,
      "step": 261
    },
    {
      "entropy": 3.813077688217163,
      "epoch": 0.7762962962962963,
      "grad_norm": 0.8186977505683899,
      "learning_rate": 2.3123123123123125e-06,
      "loss": 0.13669820129871368,
      "mean_token_accuracy": 0.9639555439352989,
      "num_tokens": 4292608.0,
      "step": 262
    },
    {
      "entropy": 4.723409593105316,
      "epoch": 0.7792592592592592,
      "grad_norm": 1.2509578466415405,
      "learning_rate": 2.2822822822822822e-06,
      "loss": 0.2619772255420685,
      "mean_token_accuracy": 0.9272322133183479,
      "num_tokens": 4308992.0,
      "step": 263
    },
    {
      "entropy": 4.524006426334381,
      "epoch": 0.7822222222222223,
      "grad_norm": 1.4938315153121948,
      "learning_rate": 2.2522522522522524e-06,
      "loss": 0.23966805636882782,
      "mean_token_accuracy": 0.9376035183668137,
      "num_tokens": 4325376.0,
      "step": 264
    },
    {
      "entropy": 4.929118633270264,
      "epoch": 0.7851851851851852,
      "grad_norm": 1.0435031652450562,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.18770577013492584,
      "mean_token_accuracy": 0.9413028880953789,
      "num_tokens": 4341760.0,
      "step": 265
    },
    {
      "entropy": 4.579936176538467,
      "epoch": 0.7881481481481482,
      "grad_norm": 1.4159420728683472,
      "learning_rate": 2.192192192192192e-06,
      "loss": 0.2776143550872803,
      "mean_token_accuracy": 0.9255271106958389,
      "num_tokens": 4358144.0,
      "step": 266
    },
    {
      "entropy": 4.820716619491577,
      "epoch": 0.7911111111111111,
      "grad_norm": 1.613826870918274,
      "learning_rate": 2.1621621621621623e-06,
      "loss": 0.3104412853717804,
      "mean_token_accuracy": 0.9057983085513115,
      "num_tokens": 4374528.0,
      "step": 267
    },
    {
      "entropy": 4.2236190140247345,
      "epoch": 0.794074074074074,
      "grad_norm": 1.0366666316986084,
      "learning_rate": 2.1321321321321325e-06,
      "loss": 0.15785987675189972,
      "mean_token_accuracy": 0.9481320306658745,
      "num_tokens": 4390912.0,
      "step": 268
    },
    {
      "entropy": 4.53969943523407,
      "epoch": 0.797037037037037,
      "grad_norm": 1.2214860916137695,
      "learning_rate": 2.102102102102102e-06,
      "loss": 0.17463329434394836,
      "mean_token_accuracy": 0.9488485306501389,
      "num_tokens": 4407296.0,
      "step": 269
    },
    {
      "entropy": 4.513545721769333,
      "epoch": 0.8,
      "grad_norm": 1.253744125366211,
      "learning_rate": 2.0720720720720723e-06,
      "loss": 0.2571987509727478,
      "mean_token_accuracy": 0.9238199889659882,
      "num_tokens": 4423680.0,
      "step": 270
    },
    {
      "entropy": 4.627400696277618,
      "epoch": 0.802962962962963,
      "grad_norm": 1.261799931526184,
      "learning_rate": 2.0420420420420424e-06,
      "loss": 0.21333637833595276,
      "mean_token_accuracy": 0.9401093125343323,
      "num_tokens": 4440064.0,
      "step": 271
    },
    {
      "entropy": 4.711275100708008,
      "epoch": 0.8059259259259259,
      "grad_norm": 1.0754541158676147,
      "learning_rate": 2.012012012012012e-06,
      "loss": 0.17184728384017944,
      "mean_token_accuracy": 0.9402873814105988,
      "num_tokens": 4456448.0,
      "step": 272
    },
    {
      "entropy": 4.300737351179123,
      "epoch": 0.8088888888888889,
      "grad_norm": 0.9222189784049988,
      "learning_rate": 1.9819819819819822e-06,
      "loss": 0.11910435557365417,
      "mean_token_accuracy": 0.96168452501297,
      "num_tokens": 4472832.0,
      "step": 273
    },
    {
      "entropy": 4.132241189479828,
      "epoch": 0.8118518518518518,
      "grad_norm": 0.9851776361465454,
      "learning_rate": 1.951951951951952e-06,
      "loss": 0.1541721224784851,
      "mean_token_accuracy": 0.9520291164517403,
      "num_tokens": 4489216.0,
      "step": 274
    },
    {
      "entropy": 4.443509578704834,
      "epoch": 0.8148148148148148,
      "grad_norm": 1.000207543373108,
      "learning_rate": 1.921921921921922e-06,
      "loss": 0.1677963137626648,
      "mean_token_accuracy": 0.9470869600772858,
      "num_tokens": 4505600.0,
      "step": 275
    },
    {
      "entropy": 4.449289679527283,
      "epoch": 0.8177777777777778,
      "grad_norm": 0.9717920422554016,
      "learning_rate": 1.8918918918918922e-06,
      "loss": 0.16066408157348633,
      "mean_token_accuracy": 0.9579946920275688,
      "num_tokens": 4521984.0,
      "step": 276
    },
    {
      "entropy": 4.996759414672852,
      "epoch": 0.8207407407407408,
      "grad_norm": 0.934596598148346,
      "learning_rate": 1.861861861861862e-06,
      "loss": 0.1031251847743988,
      "mean_token_accuracy": 0.9705559313297272,
      "num_tokens": 4538368.0,
      "step": 277
    },
    {
      "entropy": 4.149277359247208,
      "epoch": 0.8237037037037037,
      "grad_norm": 0.9842032790184021,
      "learning_rate": 1.831831831831832e-06,
      "loss": 0.14273987710475922,
      "mean_token_accuracy": 0.9624664485454559,
      "num_tokens": 4554752.0,
      "step": 278
    },
    {
      "entropy": 4.724992364645004,
      "epoch": 0.8266666666666667,
      "grad_norm": 1.2098346948623657,
      "learning_rate": 1.801801801801802e-06,
      "loss": 0.21289005875587463,
      "mean_token_accuracy": 0.9379382282495499,
      "num_tokens": 4571136.0,
      "step": 279
    },
    {
      "entropy": 4.560792237520218,
      "epoch": 0.8296296296296296,
      "grad_norm": 1.127171277999878,
      "learning_rate": 1.7717717717717719e-06,
      "loss": 0.19056805968284607,
      "mean_token_accuracy": 0.9425921738147736,
      "num_tokens": 4587520.0,
      "step": 280
    },
    {
      "entropy": 4.740212678909302,
      "epoch": 0.8325925925925926,
      "grad_norm": 1.2597956657409668,
      "learning_rate": 1.7417417417417418e-06,
      "loss": 0.15070828795433044,
      "mean_token_accuracy": 0.9584286734461784,
      "num_tokens": 4603904.0,
      "step": 281
    },
    {
      "entropy": 4.721891224384308,
      "epoch": 0.8355555555555556,
      "grad_norm": 0.9663627743721008,
      "learning_rate": 1.711711711711712e-06,
      "loss": 0.15181522071361542,
      "mean_token_accuracy": 0.9599451720714569,
      "num_tokens": 4620288.0,
      "step": 282
    },
    {
      "entropy": 4.407726615667343,
      "epoch": 0.8385185185185186,
      "grad_norm": 0.8520182371139526,
      "learning_rate": 1.681681681681682e-06,
      "loss": 0.144394189119339,
      "mean_token_accuracy": 0.9539293721318245,
      "num_tokens": 4636672.0,
      "step": 283
    },
    {
      "entropy": 4.145166367292404,
      "epoch": 0.8414814814814815,
      "grad_norm": 0.9270368814468384,
      "learning_rate": 1.6516516516516517e-06,
      "loss": 0.1719590574502945,
      "mean_token_accuracy": 0.9544447138905525,
      "num_tokens": 4653056.0,
      "step": 284
    },
    {
      "entropy": 4.479555249214172,
      "epoch": 0.8444444444444444,
      "grad_norm": 1.0755841732025146,
      "learning_rate": 1.6216216216216219e-06,
      "loss": 0.17867596447467804,
      "mean_token_accuracy": 0.9488372579216957,
      "num_tokens": 4669440.0,
      "step": 285
    },
    {
      "entropy": 4.480806350708008,
      "epoch": 0.8474074074074074,
      "grad_norm": 1.2898063659667969,
      "learning_rate": 1.5915915915915916e-06,
      "loss": 0.19736865162849426,
      "mean_token_accuracy": 0.950205609202385,
      "num_tokens": 4685824.0,
      "step": 286
    },
    {
      "entropy": 4.170996308326721,
      "epoch": 0.8503703703703703,
      "grad_norm": 0.9429665207862854,
      "learning_rate": 1.5615615615615617e-06,
      "loss": 0.14257119596004486,
      "mean_token_accuracy": 0.9622804448008537,
      "num_tokens": 4702208.0,
      "step": 287
    },
    {
      "entropy": 4.297813713550568,
      "epoch": 0.8533333333333334,
      "grad_norm": 1.0085495710372925,
      "learning_rate": 1.5315315315315316e-06,
      "loss": 0.14964225888252258,
      "mean_token_accuracy": 0.9610767140984535,
      "num_tokens": 4718592.0,
      "step": 288
    },
    {
      "entropy": 4.700436592102051,
      "epoch": 0.8562962962962963,
      "grad_norm": 1.283758521080017,
      "learning_rate": 1.5015015015015017e-06,
      "loss": 0.18858590722084045,
      "mean_token_accuracy": 0.941825196146965,
      "num_tokens": 4734976.0,
      "step": 289
    },
    {
      "entropy": 4.504350274801254,
      "epoch": 0.8592592592592593,
      "grad_norm": 1.0768287181854248,
      "learning_rate": 1.4714714714714714e-06,
      "loss": 0.19429968297481537,
      "mean_token_accuracy": 0.9455336704850197,
      "num_tokens": 4751360.0,
      "step": 290
    },
    {
      "entropy": 4.9205896854400635,
      "epoch": 0.8622222222222222,
      "grad_norm": 1.3085178136825562,
      "learning_rate": 1.4414414414414416e-06,
      "loss": 0.24937868118286133,
      "mean_token_accuracy": 0.9369010552763939,
      "num_tokens": 4767744.0,
      "step": 291
    },
    {
      "entropy": 4.477796494960785,
      "epoch": 0.8651851851851852,
      "grad_norm": 1.0749484300613403,
      "learning_rate": 1.4114114114114117e-06,
      "loss": 0.15879273414611816,
      "mean_token_accuracy": 0.9605017602443695,
      "num_tokens": 4784128.0,
      "step": 292
    },
    {
      "entropy": 5.006076633930206,
      "epoch": 0.8681481481481481,
      "grad_norm": 1.504447102546692,
      "learning_rate": 1.3813813813813814e-06,
      "loss": 0.20927917957305908,
      "mean_token_accuracy": 0.9365587085485458,
      "num_tokens": 4800512.0,
      "step": 293
    },
    {
      "entropy": 4.842982888221741,
      "epoch": 0.8711111111111111,
      "grad_norm": 1.1934711933135986,
      "learning_rate": 1.3513513513513515e-06,
      "loss": 0.16013149917125702,
      "mean_token_accuracy": 0.9554265588521957,
      "num_tokens": 4816896.0,
      "step": 294
    },
    {
      "entropy": 5.128973603248596,
      "epoch": 0.8740740740740741,
      "grad_norm": 1.3415642976760864,
      "learning_rate": 1.3213213213213214e-06,
      "loss": 0.2703382074832916,
      "mean_token_accuracy": 0.9249211624264717,
      "num_tokens": 4833280.0,
      "step": 295
    },
    {
      "entropy": 4.623582750558853,
      "epoch": 0.8770370370370371,
      "grad_norm": 1.0999549627304077,
      "learning_rate": 1.2912912912912913e-06,
      "loss": 0.171325221657753,
      "mean_token_accuracy": 0.9484955444931984,
      "num_tokens": 4849664.0,
      "step": 296
    },
    {
      "entropy": 3.944809466600418,
      "epoch": 0.88,
      "grad_norm": 0.7336201667785645,
      "learning_rate": 1.2612612612612613e-06,
      "loss": 0.11464260518550873,
      "mean_token_accuracy": 0.9696908593177795,
      "num_tokens": 4866048.0,
      "step": 297
    },
    {
      "entropy": 4.276359975337982,
      "epoch": 0.882962962962963,
      "grad_norm": 1.1169263124465942,
      "learning_rate": 1.2312312312312314e-06,
      "loss": 0.20006108283996582,
      "mean_token_accuracy": 0.9414397776126862,
      "num_tokens": 4882432.0,
      "step": 298
    },
    {
      "entropy": 4.439151048660278,
      "epoch": 0.8859259259259259,
      "grad_norm": 0.9406883120536804,
      "learning_rate": 1.2012012012012013e-06,
      "loss": 0.1345605105161667,
      "mean_token_accuracy": 0.9641304537653923,
      "num_tokens": 4898816.0,
      "step": 299
    },
    {
      "entropy": 4.791579067707062,
      "epoch": 0.8888888888888888,
      "grad_norm": 1.3605457544326782,
      "learning_rate": 1.1711711711711712e-06,
      "loss": 0.22707751393318176,
      "mean_token_accuracy": 0.9276619628071785,
      "num_tokens": 4915200.0,
      "step": 300
    },
    {
      "epoch": 0.8888888888888888,
      "eval_entropy": 4.468447434107462,
      "eval_loss": 0.18680252134799957,
      "eval_mean_token_accuracy": 0.9488305298487345,
      "eval_num_tokens": 4915200.0,
      "eval_runtime": 41.7774,
      "eval_samples_per_second": 14.362,
      "eval_steps_per_second": 1.795,
      "step": 300
    },
    {
      "entropy": 5.0460052490234375,
      "epoch": 0.8918518518518519,
      "grad_norm": 1.672562599182129,
      "learning_rate": 1.1411411411411411e-06,
      "loss": 0.3344571590423584,
      "mean_token_accuracy": 0.9075080156326294,
      "num_tokens": 4931584.0,
      "step": 301
    },
    {
      "entropy": 4.424874842166901,
      "epoch": 0.8948148148148148,
      "grad_norm": 0.8936777710914612,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.14294463396072388,
      "mean_token_accuracy": 0.9543017521500587,
      "num_tokens": 4947968.0,
      "step": 302
    },
    {
      "entropy": 5.2555792927742,
      "epoch": 0.8977777777777778,
      "grad_norm": 1.6100314855575562,
      "learning_rate": 1.0810810810810812e-06,
      "loss": 0.23647087812423706,
      "mean_token_accuracy": 0.9238611906766891,
      "num_tokens": 4964352.0,
      "step": 303
    },
    {
      "entropy": 4.815032601356506,
      "epoch": 0.9007407407407407,
      "grad_norm": 1.3806898593902588,
      "learning_rate": 1.051051051051051e-06,
      "loss": 0.24262529611587524,
      "mean_token_accuracy": 0.9265779107809067,
      "num_tokens": 4980736.0,
      "step": 304
    },
    {
      "entropy": 4.329847127199173,
      "epoch": 0.9037037037037037,
      "grad_norm": 0.9695173501968384,
      "learning_rate": 1.0210210210210212e-06,
      "loss": 0.14157229661941528,
      "mean_token_accuracy": 0.9537581130862236,
      "num_tokens": 4997120.0,
      "step": 305
    },
    {
      "entropy": 4.7898417711257935,
      "epoch": 0.9066666666666666,
      "grad_norm": 1.4385104179382324,
      "learning_rate": 9.909909909909911e-07,
      "loss": 0.23195311427116394,
      "mean_token_accuracy": 0.923228420317173,
      "num_tokens": 5013504.0,
      "step": 306
    },
    {
      "entropy": 4.603113412857056,
      "epoch": 0.9096296296296297,
      "grad_norm": 1.1466540098190308,
      "learning_rate": 9.60960960960961e-07,
      "loss": 0.2606795132160187,
      "mean_token_accuracy": 0.9337464720010757,
      "num_tokens": 5029888.0,
      "step": 307
    },
    {
      "entropy": 4.548976570367813,
      "epoch": 0.9125925925925926,
      "grad_norm": 1.063750147819519,
      "learning_rate": 9.30930930930931e-07,
      "loss": 0.13953007757663727,
      "mean_token_accuracy": 0.9625963196158409,
      "num_tokens": 5046272.0,
      "step": 308
    },
    {
      "entropy": 4.353551208972931,
      "epoch": 0.9155555555555556,
      "grad_norm": 1.182592749595642,
      "learning_rate": 9.00900900900901e-07,
      "loss": 0.22516204416751862,
      "mean_token_accuracy": 0.9505400955677032,
      "num_tokens": 5062656.0,
      "step": 309
    },
    {
      "entropy": 4.589662879705429,
      "epoch": 0.9185185185185185,
      "grad_norm": 1.1263549327850342,
      "learning_rate": 8.708708708708709e-07,
      "loss": 0.17610906064510345,
      "mean_token_accuracy": 0.9492174163460732,
      "num_tokens": 5079040.0,
      "step": 310
    },
    {
      "entropy": 4.288213074207306,
      "epoch": 0.9214814814814815,
      "grad_norm": 1.0428308248519897,
      "learning_rate": 8.40840840840841e-07,
      "loss": 0.17399916052818298,
      "mean_token_accuracy": 0.957172192633152,
      "num_tokens": 5095424.0,
      "step": 311
    },
    {
      "entropy": 4.563187420368195,
      "epoch": 0.9244444444444444,
      "grad_norm": 1.0395258665084839,
      "learning_rate": 8.108108108108109e-07,
      "loss": 0.15612207353115082,
      "mean_token_accuracy": 0.9563304558396339,
      "num_tokens": 5111808.0,
      "step": 312
    },
    {
      "entropy": 4.641351521015167,
      "epoch": 0.9274074074074075,
      "grad_norm": 1.378040075302124,
      "learning_rate": 7.807807807807808e-07,
      "loss": 0.21732960641384125,
      "mean_token_accuracy": 0.9494708776473999,
      "num_tokens": 5128192.0,
      "step": 313
    },
    {
      "entropy": 4.642271339893341,
      "epoch": 0.9303703703703704,
      "grad_norm": 1.2148431539535522,
      "learning_rate": 7.507507507507509e-07,
      "loss": 0.22766008973121643,
      "mean_token_accuracy": 0.9394201561808586,
      "num_tokens": 5144576.0,
      "step": 314
    },
    {
      "entropy": 4.512627899646759,
      "epoch": 0.9333333333333333,
      "grad_norm": 1.0244803428649902,
      "learning_rate": 7.207207207207208e-07,
      "loss": 0.1585380584001541,
      "mean_token_accuracy": 0.9516338929533958,
      "num_tokens": 5160960.0,
      "step": 315
    },
    {
      "entropy": 4.724791347980499,
      "epoch": 0.9362962962962963,
      "grad_norm": 1.3785619735717773,
      "learning_rate": 6.906906906906907e-07,
      "loss": 0.2422824501991272,
      "mean_token_accuracy": 0.935348816215992,
      "num_tokens": 5177344.0,
      "step": 316
    },
    {
      "entropy": 4.733954519033432,
      "epoch": 0.9392592592592592,
      "grad_norm": 1.3217679262161255,
      "learning_rate": 6.606606606606607e-07,
      "loss": 0.22770625352859497,
      "mean_token_accuracy": 0.9305073171854019,
      "num_tokens": 5193728.0,
      "step": 317
    },
    {
      "entropy": 4.7431544959545135,
      "epoch": 0.9422222222222222,
      "grad_norm": 1.0940858125686646,
      "learning_rate": 6.306306306306306e-07,
      "loss": 0.12314062565565109,
      "mean_token_accuracy": 0.9600890204310417,
      "num_tokens": 5210112.0,
      "step": 318
    },
    {
      "entropy": 4.324954122304916,
      "epoch": 0.9451851851851852,
      "grad_norm": 0.9080625772476196,
      "learning_rate": 6.006006006006006e-07,
      "loss": 0.16247284412384033,
      "mean_token_accuracy": 0.9540717229247093,
      "num_tokens": 5226496.0,
      "step": 319
    },
    {
      "entropy": 4.2534666657447815,
      "epoch": 0.9481481481481482,
      "grad_norm": 1.1102370023727417,
      "learning_rate": 5.705705705705706e-07,
      "loss": 0.1454768031835556,
      "mean_token_accuracy": 0.9645774811506271,
      "num_tokens": 5242880.0,
      "step": 320
    },
    {
      "entropy": 4.176030099391937,
      "epoch": 0.9511111111111111,
      "grad_norm": 0.8763568997383118,
      "learning_rate": 5.405405405405406e-07,
      "loss": 0.12408076971769333,
      "mean_token_accuracy": 0.9674194008111954,
      "num_tokens": 5259264.0,
      "step": 321
    },
    {
      "entropy": 4.658020377159119,
      "epoch": 0.9540740740740741,
      "grad_norm": 1.3769937753677368,
      "learning_rate": 5.105105105105106e-07,
      "loss": 0.20787939429283142,
      "mean_token_accuracy": 0.9395479783415794,
      "num_tokens": 5275648.0,
      "step": 322
    },
    {
      "entropy": 4.637662440538406,
      "epoch": 0.957037037037037,
      "grad_norm": 1.0618715286254883,
      "learning_rate": 4.804804804804805e-07,
      "loss": 0.15968897938728333,
      "mean_token_accuracy": 0.9517290517687798,
      "num_tokens": 5292032.0,
      "step": 323
    },
    {
      "entropy": 4.746343225240707,
      "epoch": 0.96,
      "grad_norm": 1.552250623703003,
      "learning_rate": 4.504504504504505e-07,
      "loss": 0.29586261510849,
      "mean_token_accuracy": 0.9229539930820465,
      "num_tokens": 5308416.0,
      "step": 324
    },
    {
      "entropy": 4.7195650935173035,
      "epoch": 0.9629629629629629,
      "grad_norm": 1.1863840818405151,
      "learning_rate": 4.204204204204205e-07,
      "loss": 0.18401281535625458,
      "mean_token_accuracy": 0.9443574622273445,
      "num_tokens": 5324800.0,
      "step": 325
    },
    {
      "entropy": 4.679966747760773,
      "epoch": 0.965925925925926,
      "grad_norm": 0.794443666934967,
      "learning_rate": 3.903903903903904e-07,
      "loss": 0.09033509343862534,
      "mean_token_accuracy": 0.9667606800794601,
      "num_tokens": 5341184.0,
      "step": 326
    },
    {
      "entropy": 3.9146403670310974,
      "epoch": 0.9688888888888889,
      "grad_norm": 0.8764992356300354,
      "learning_rate": 3.603603603603604e-07,
      "loss": 0.14650200307369232,
      "mean_token_accuracy": 0.9536269530653954,
      "num_tokens": 5357568.0,
      "step": 327
    },
    {
      "entropy": 4.918329060077667,
      "epoch": 0.9718518518518519,
      "grad_norm": 1.2639726400375366,
      "learning_rate": 3.3033033033033036e-07,
      "loss": 0.21113714575767517,
      "mean_token_accuracy": 0.9443980157375336,
      "num_tokens": 5373952.0,
      "step": 328
    },
    {
      "entropy": 4.445641905069351,
      "epoch": 0.9748148148148148,
      "grad_norm": 1.0556613206863403,
      "learning_rate": 3.003003003003003e-07,
      "loss": 0.19537337124347687,
      "mean_token_accuracy": 0.9416832104325294,
      "num_tokens": 5390336.0,
      "step": 329
    },
    {
      "entropy": 4.553157716989517,
      "epoch": 0.9777777777777777,
      "grad_norm": 0.9739739894866943,
      "learning_rate": 2.702702702702703e-07,
      "loss": 0.16301685571670532,
      "mean_token_accuracy": 0.952493242919445,
      "num_tokens": 5406720.0,
      "step": 330
    },
    {
      "entropy": 4.54691219329834,
      "epoch": 0.9807407407407407,
      "grad_norm": 1.193598985671997,
      "learning_rate": 2.4024024024024026e-07,
      "loss": 0.21868029236793518,
      "mean_token_accuracy": 0.9394289404153824,
      "num_tokens": 5423104.0,
      "step": 331
    },
    {
      "entropy": 4.176305890083313,
      "epoch": 0.9837037037037037,
      "grad_norm": 0.9158902168273926,
      "learning_rate": 2.1021021021021025e-07,
      "loss": 0.1457735002040863,
      "mean_token_accuracy": 0.9562234431505203,
      "num_tokens": 5439488.0,
      "step": 332
    },
    {
      "entropy": 4.15862163901329,
      "epoch": 0.9866666666666667,
      "grad_norm": 1.0964727401733398,
      "learning_rate": 1.801801801801802e-07,
      "loss": 0.15685325860977173,
      "mean_token_accuracy": 0.9591808095574379,
      "num_tokens": 5455872.0,
      "step": 333
    },
    {
      "entropy": 4.425233960151672,
      "epoch": 0.9896296296296296,
      "grad_norm": 1.1989465951919556,
      "learning_rate": 1.5015015015015016e-07,
      "loss": 0.1884509027004242,
      "mean_token_accuracy": 0.9532799497246742,
      "num_tokens": 5472256.0,
      "step": 334
    },
    {
      "entropy": 4.831894814968109,
      "epoch": 0.9925925925925926,
      "grad_norm": 1.1758801937103271,
      "learning_rate": 1.2012012012012013e-07,
      "loss": 0.17529185116291046,
      "mean_token_accuracy": 0.954959362745285,
      "num_tokens": 5488640.0,
      "step": 335
    },
    {
      "entropy": 4.57403039932251,
      "epoch": 0.9955555555555555,
      "grad_norm": 1.1927546262741089,
      "learning_rate": 9.00900900900901e-08,
      "loss": 0.20289713144302368,
      "mean_token_accuracy": 0.9340859204530716,
      "num_tokens": 5505024.0,
      "step": 336
    },
    {
      "entropy": 4.882504642009735,
      "epoch": 0.9985185185185185,
      "grad_norm": 1.341749906539917,
      "learning_rate": 6.006006006006006e-08,
      "loss": 0.2592087686061859,
      "mean_token_accuracy": 0.9262190982699394,
      "num_tokens": 5521408.0,
      "step": 337
    },
    {
      "entropy": 5.048959851264954,
      "epoch": 1.0,
      "grad_norm": 2.596557140350342,
      "learning_rate": 3.003003003003003e-08,
      "loss": 0.27569669485092163,
      "mean_token_accuracy": 0.9293998777866364,
      "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
}