{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4849513533173703,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 0.09651942026102915,
      "epoch": 0.002424756766586852,
      "grad_norm": 552.0,
      "learning_rate": 9.000000000000001e-07,
      "loss": 6.736698150634766,
      "mean_token_accuracy": 0.5771780241280794,
      "num_tokens": 36846.0,
      "step": 10
    },
    {
      "entropy": 0.10363724352791906,
      "epoch": 0.004849513533173704,
      "grad_norm": 296.0,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 6.29327507019043,
      "mean_token_accuracy": 0.6011363569647074,
      "num_tokens": 81163.0,
      "step": 20
    },
    {
      "entropy": 0.09438130774069578,
      "epoch": 0.007274270299760555,
      "grad_norm": 222.0,
      "learning_rate": 2.9e-06,
      "loss": 5.125759887695312,
      "mean_token_accuracy": 0.6545454517006875,
      "num_tokens": 123610.0,
      "step": 30
    },
    {
      "entropy": 0.10434843641705811,
      "epoch": 0.009699027066347407,
      "grad_norm": 205.0,
      "learning_rate": 3.900000000000001e-06,
      "loss": 3.081037902832031,
      "mean_token_accuracy": 0.6937500037252903,
      "num_tokens": 163947.0,
      "step": 40
    },
    {
      "entropy": 0.06968767816433682,
      "epoch": 0.012123783832934259,
      "grad_norm": 39.75,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.6972839832305908,
      "mean_token_accuracy": 0.8579545550048351,
      "num_tokens": 203175.0,
      "step": 50
    },
    {
      "entropy": 0.09051208726596087,
      "epoch": 0.01454854059952111,
      "grad_norm": 14.75,
      "learning_rate": 5.9e-06,
      "loss": 0.1402425527572632,
      "mean_token_accuracy": 0.9636363588273525,
      "num_tokens": 242004.0,
      "step": 60
    },
    {
      "entropy": 0.04463538663112558,
      "epoch": 0.01697329736610796,
      "grad_norm": 3.453125,
      "learning_rate": 6.9e-06,
      "loss": 0.059400951862335204,
      "mean_token_accuracy": 0.9761093020439148,
      "num_tokens": 282287.0,
      "step": 70
    },
    {
      "entropy": 0.03611970616912004,
      "epoch": 0.019398054132694815,
      "grad_norm": 1.78125,
      "learning_rate": 7.9e-06,
      "loss": 0.03747022747993469,
      "mean_token_accuracy": 0.9812499955296516,
      "num_tokens": 320685.0,
      "step": 80
    },
    {
      "entropy": 0.03572612675052369,
      "epoch": 0.021822810899281665,
      "grad_norm": 3.765625,
      "learning_rate": 8.900000000000001e-06,
      "loss": 0.03938577175140381,
      "mean_token_accuracy": 0.9784090861678123,
      "num_tokens": 359560.0,
      "step": 90
    },
    {
      "entropy": 0.035298487018735614,
      "epoch": 0.024247567665868518,
      "grad_norm": 11.9375,
      "learning_rate": 9.9e-06,
      "loss": 0.03732649385929108,
      "mean_token_accuracy": 0.9772456660866737,
      "num_tokens": 403870.0,
      "step": 100
    },
    {
      "entropy": 0.03189297981371055,
      "epoch": 0.02667232443245537,
      "grad_norm": 2.484375,
      "learning_rate": 1.0900000000000002e-05,
      "loss": 0.03071807622909546,
      "mean_token_accuracy": 0.9823863603174686,
      "num_tokens": 443514.0,
      "step": 110
    },
    {
      "entropy": 0.027393289806786923,
      "epoch": 0.02909708119904222,
      "grad_norm": 7.75,
      "learning_rate": 1.1900000000000001e-05,
      "loss": 0.03194451332092285,
      "mean_token_accuracy": 0.9812499962747097,
      "num_tokens": 483941.0,
      "step": 120
    },
    {
      "entropy": 0.02901054713729536,
      "epoch": 0.031521837965629074,
      "grad_norm": 2.078125,
      "learning_rate": 1.2900000000000002e-05,
      "loss": 0.04001099765300751,
      "mean_token_accuracy": 0.9801136322319508,
      "num_tokens": 524184.0,
      "step": 130
    },
    {
      "entropy": 0.036186478319723395,
      "epoch": 0.03394659473221592,
      "grad_norm": 2.453125,
      "learning_rate": 1.39e-05,
      "loss": 0.032131227850914004,
      "mean_token_accuracy": 0.977840906381607,
      "num_tokens": 565235.0,
      "step": 140
    },
    {
      "entropy": 0.02588132777964347,
      "epoch": 0.03637135149880277,
      "grad_norm": 2.59375,
      "learning_rate": 1.4900000000000001e-05,
      "loss": 0.032545623183250424,
      "mean_token_accuracy": 0.9795183934271335,
      "num_tokens": 604912.0,
      "step": 150
    },
    {
      "entropy": 0.024791866151281282,
      "epoch": 0.03879610826538963,
      "grad_norm": 4.71875,
      "learning_rate": 1.5900000000000004e-05,
      "loss": 0.029649946093559264,
      "mean_token_accuracy": 0.9835227228701114,
      "num_tokens": 643428.0,
      "step": 160
    },
    {
      "entropy": 0.035497133302123986,
      "epoch": 0.04122086503197648,
      "grad_norm": 2.421875,
      "learning_rate": 1.69e-05,
      "loss": 0.03451511561870575,
      "mean_token_accuracy": 0.977840905636549,
      "num_tokens": 682578.0,
      "step": 170
    },
    {
      "entropy": 0.032462505592047816,
      "epoch": 0.04364562179856333,
      "grad_norm": 1.3671875,
      "learning_rate": 1.79e-05,
      "loss": 0.03423822820186615,
      "mean_token_accuracy": 0.9778138473629951,
      "num_tokens": 723563.0,
      "step": 180
    },
    {
      "entropy": 0.03153426277203834,
      "epoch": 0.046070378565150186,
      "grad_norm": 1.3984375,
      "learning_rate": 1.8900000000000002e-05,
      "loss": 0.030463039875030518,
      "mean_token_accuracy": 0.9818181775510311,
      "num_tokens": 765374.0,
      "step": 190
    },
    {
      "entropy": 0.020951804842843556,
      "epoch": 0.048495135331737035,
      "grad_norm": 4.6875,
      "learning_rate": 1.9900000000000003e-05,
      "loss": 0.04620547294616699,
      "mean_token_accuracy": 0.9801136329770088,
      "num_tokens": 806149.0,
      "step": 200
    },
    {
      "entropy": 0.03319120884152653,
      "epoch": 0.050919892098323885,
      "grad_norm": 3.40625,
      "learning_rate": 1.9900000000000003e-05,
      "loss": 0.03888022005558014,
      "mean_token_accuracy": 0.9784090884029866,
      "num_tokens": 852548.0,
      "step": 210
    },
    {
      "entropy": 0.034044423646992075,
      "epoch": 0.05334464886491074,
      "grad_norm": 1.875,
      "learning_rate": 1.978888888888889e-05,
      "loss": 0.03665586411952972,
      "mean_token_accuracy": 0.9772727221250535,
      "num_tokens": 891570.0,
      "step": 220
    },
    {
      "entropy": 0.020276623886820742,
      "epoch": 0.05576940563149759,
      "grad_norm": 1.4921875,
      "learning_rate": 1.967777777777778e-05,
      "loss": 0.034099149703979495,
      "mean_token_accuracy": 0.9846590891480446,
      "num_tokens": 931221.0,
      "step": 230
    },
    {
      "entropy": 0.03405958858202211,
      "epoch": 0.05819416239808444,
      "grad_norm": 2.671875,
      "learning_rate": 1.956666666666667e-05,
      "loss": 0.03181847333908081,
      "mean_token_accuracy": 0.9812499970197678,
      "num_tokens": 973029.0,
      "step": 240
    },
    {
      "entropy": 0.0318044990162889,
      "epoch": 0.06061891916467129,
      "grad_norm": 1.71875,
      "learning_rate": 1.9455555555555555e-05,
      "loss": 0.031788495182991025,
      "mean_token_accuracy": 0.9806818142533302,
      "num_tokens": 1012763.0,
      "step": 250
    },
    {
      "entropy": 0.031602241041036906,
      "epoch": 0.06304367593125815,
      "grad_norm": 3.546875,
      "learning_rate": 1.9344444444444446e-05,
      "loss": 0.03268867433071136,
      "mean_token_accuracy": 0.98295454159379,
      "num_tokens": 1050122.0,
      "step": 260
    },
    {
      "entropy": 0.024405305968684843,
      "epoch": 0.065468432697845,
      "grad_norm": 2.234375,
      "learning_rate": 1.9233333333333334e-05,
      "loss": 0.03167384266853333,
      "mean_token_accuracy": 0.9846590869128704,
      "num_tokens": 1092243.0,
      "step": 270
    },
    {
      "entropy": 0.026620667469615,
      "epoch": 0.06789318946443185,
      "grad_norm": 1.21875,
      "learning_rate": 1.9122222222222226e-05,
      "loss": 0.02862342298030853,
      "mean_token_accuracy": 0.9817911215126515,
      "num_tokens": 1130648.0,
      "step": 280
    },
    {
      "entropy": 0.03168776935272035,
      "epoch": 0.0703179462310187,
      "grad_norm": 2.921875,
      "learning_rate": 1.9011111111111114e-05,
      "loss": 0.02828059494495392,
      "mean_token_accuracy": 0.9823863588273525,
      "num_tokens": 1173427.0,
      "step": 290
    },
    {
      "entropy": 0.01820039304529928,
      "epoch": 0.07274270299760555,
      "grad_norm": 3.46875,
      "learning_rate": 1.8900000000000002e-05,
      "loss": 0.033576938509941104,
      "mean_token_accuracy": 0.985227269679308,
      "num_tokens": 1210875.0,
      "step": 300
    },
    {
      "entropy": 0.03017507032072899,
      "epoch": 0.07516745976419241,
      "grad_norm": 2.40625,
      "learning_rate": 1.878888888888889e-05,
      "loss": 0.03038310110569,
      "mean_token_accuracy": 0.9852272674441338,
      "num_tokens": 1250450.0,
      "step": 310
    },
    {
      "entropy": 0.031056866744256693,
      "epoch": 0.07759221653077926,
      "grad_norm": 10.5625,
      "learning_rate": 1.8677777777777778e-05,
      "loss": 0.02981090545654297,
      "mean_token_accuracy": 0.9846590854227543,
      "num_tokens": 1292161.0,
      "step": 320
    },
    {
      "entropy": 0.02875728546641767,
      "epoch": 0.08001697329736611,
      "grad_norm": 2.078125,
      "learning_rate": 1.856666666666667e-05,
      "loss": 0.028395554423332213,
      "mean_token_accuracy": 0.9857954517006874,
      "num_tokens": 1335278.0,
      "step": 330
    },
    {
      "entropy": 0.029322995853544854,
      "epoch": 0.08244173006395296,
      "grad_norm": 3.84375,
      "learning_rate": 1.8455555555555558e-05,
      "loss": 0.029434293508529663,
      "mean_token_accuracy": 0.9835227228701114,
      "num_tokens": 1373606.0,
      "step": 340
    },
    {
      "entropy": 0.027826807415840447,
      "epoch": 0.08486648683053981,
      "grad_norm": 2.21875,
      "learning_rate": 1.8344444444444446e-05,
      "loss": 0.03300779163837433,
      "mean_token_accuracy": 0.978143934905529,
      "num_tokens": 1417353.0,
      "step": 350
    },
    {
      "entropy": 0.0319683638830611,
      "epoch": 0.08729124359712666,
      "grad_norm": 2.203125,
      "learning_rate": 1.8233333333333334e-05,
      "loss": 0.03151697814464569,
      "mean_token_accuracy": 0.9823863580822945,
      "num_tokens": 1453720.0,
      "step": 360
    },
    {
      "entropy": 0.025839641344646225,
      "epoch": 0.08971600036371352,
      "grad_norm": 2.6875,
      "learning_rate": 1.8122222222222222e-05,
      "loss": 0.02670169174671173,
      "mean_token_accuracy": 0.9874999955296516,
      "num_tokens": 1496830.0,
      "step": 370
    },
    {
      "entropy": 0.02322939161786053,
      "epoch": 0.09214075713030037,
      "grad_norm": 1.921875,
      "learning_rate": 1.8011111111111113e-05,
      "loss": 0.029871052503585814,
      "mean_token_accuracy": 0.9823863603174686,
      "num_tokens": 1539244.0,
      "step": 380
    },
    {
      "entropy": 0.028905938124444218,
      "epoch": 0.09456551389688722,
      "grad_norm": 4.9375,
      "learning_rate": 1.79e-05,
      "loss": 0.031239384412765504,
      "mean_token_accuracy": 0.9874999970197678,
      "num_tokens": 1581506.0,
      "step": 390
    },
    {
      "entropy": 0.020863897850722422,
      "epoch": 0.09699027066347407,
      "grad_norm": 3.3125,
      "learning_rate": 1.778888888888889e-05,
      "loss": 0.0223777636885643,
      "mean_token_accuracy": 0.9874999970197678,
      "num_tokens": 1623446.0,
      "step": 400
    },
    {
      "entropy": 0.020877818292319716,
      "epoch": 0.09941502743006092,
      "grad_norm": 3.90625,
      "learning_rate": 1.7677777777777778e-05,
      "loss": 0.025508558750152587,
      "mean_token_accuracy": 0.9880681782960892,
      "num_tokens": 1662111.0,
      "step": 410
    },
    {
      "entropy": 0.026274691119397175,
      "epoch": 0.10183978419664777,
      "grad_norm": 1.5,
      "learning_rate": 1.756666666666667e-05,
      "loss": 0.03404468297958374,
      "mean_token_accuracy": 0.9801136307418347,
      "num_tokens": 1702735.0,
      "step": 420
    },
    {
      "entropy": 0.0291934731634683,
      "epoch": 0.10426454096323462,
      "grad_norm": 0.90625,
      "learning_rate": 1.7455555555555557e-05,
      "loss": 0.02973931431770325,
      "mean_token_accuracy": 0.9846590869128704,
      "num_tokens": 1740129.0,
      "step": 430
    },
    {
      "entropy": 0.02555293397508649,
      "epoch": 0.10668929772982148,
      "grad_norm": 0.84375,
      "learning_rate": 1.7344444444444445e-05,
      "loss": 0.02494020164012909,
      "mean_token_accuracy": 0.9869318164885044,
      "num_tokens": 1784479.0,
      "step": 440
    },
    {
      "entropy": 0.028070836007191246,
      "epoch": 0.10911405449640833,
      "grad_norm": 2.125,
      "learning_rate": 1.7233333333333337e-05,
      "loss": 0.029692673683166505,
      "mean_token_accuracy": 0.9835227228701114,
      "num_tokens": 1826847.0,
      "step": 450
    },
    {
      "entropy": 0.02322534844988695,
      "epoch": 0.11153881126299518,
      "grad_norm": 2.5,
      "learning_rate": 1.7122222222222225e-05,
      "loss": 0.02545521855354309,
      "mean_token_accuracy": 0.9874999970197678,
      "num_tokens": 1862712.0,
      "step": 460
    },
    {
      "entropy": 0.022724387578819005,
      "epoch": 0.11396356802958203,
      "grad_norm": 1.640625,
      "learning_rate": 1.7011111111111113e-05,
      "loss": 0.0265019953250885,
      "mean_token_accuracy": 0.9846590869128704,
      "num_tokens": 1903077.0,
      "step": 470
    },
    {
      "entropy": 0.02751030181589158,
      "epoch": 0.11638832479616888,
      "grad_norm": 2.03125,
      "learning_rate": 1.69e-05,
      "loss": 0.027713432908058167,
      "mean_token_accuracy": 0.9829545430839062,
      "num_tokens": 1948431.0,
      "step": 480
    },
    {
      "entropy": 0.02874135026240765,
      "epoch": 0.11881308156275573,
      "grad_norm": 0.88671875,
      "learning_rate": 1.678888888888889e-05,
      "loss": 0.03457841873168945,
      "mean_token_accuracy": 0.9806818127632141,
      "num_tokens": 1990633.0,
      "step": 490
    },
    {
      "entropy": 0.02799490848346977,
      "epoch": 0.12123783832934258,
      "grad_norm": 1.3359375,
      "learning_rate": 1.667777777777778e-05,
      "loss": 0.030140578746795654,
      "mean_token_accuracy": 0.9835227251052856,
      "num_tokens": 2028217.0,
      "step": 500
    },
    {
      "entropy": 0.025997174895564967,
      "epoch": 0.12366259509592945,
      "grad_norm": 1.71875,
      "learning_rate": 1.656666666666667e-05,
      "loss": 0.025564667582511903,
      "mean_token_accuracy": 0.9857954509556294,
      "num_tokens": 2072644.0,
      "step": 510
    },
    {
      "entropy": 0.023441067839212338,
      "epoch": 0.1260873518625163,
      "grad_norm": 2.34375,
      "learning_rate": 1.6455555555555556e-05,
      "loss": 0.024340197443962097,
      "mean_token_accuracy": 0.9863636337220669,
      "num_tokens": 2110975.0,
      "step": 520
    },
    {
      "entropy": 0.021592267712185274,
      "epoch": 0.12851210862910314,
      "grad_norm": 3.21875,
      "learning_rate": 1.6344444444444445e-05,
      "loss": 0.02947987914085388,
      "mean_token_accuracy": 0.9886363595724106,
      "num_tokens": 2150997.0,
      "step": 530
    },
    {
      "entropy": 0.02172533786206259,
      "epoch": 0.13093686539569,
      "grad_norm": 1.890625,
      "learning_rate": 1.6233333333333333e-05,
      "loss": 0.027436545491218566,
      "mean_token_accuracy": 0.9846590884029865,
      "num_tokens": 2190804.0,
      "step": 540
    },
    {
      "entropy": 0.02379558993197861,
      "epoch": 0.13336162216227684,
      "grad_norm": 2.21875,
      "learning_rate": 1.6122222222222224e-05,
      "loss": 0.025879621505737305,
      "mean_token_accuracy": 0.9863636337220669,
      "num_tokens": 2240073.0,
      "step": 550
    },
    {
      "entropy": 0.02549256883594353,
      "epoch": 0.1357863789288637,
      "grad_norm": 0.92578125,
      "learning_rate": 1.6011111111111112e-05,
      "loss": 0.024750103056430817,
      "mean_token_accuracy": 0.9880681797862053,
      "num_tokens": 2280612.0,
      "step": 560
    },
    {
      "entropy": 0.03142027879757734,
      "epoch": 0.13821113569545054,
      "grad_norm": 3.34375,
      "learning_rate": 1.5900000000000004e-05,
      "loss": 0.03994172811508179,
      "mean_token_accuracy": 0.9778138488531113,
      "num_tokens": 2320775.0,
      "step": 570
    },
    {
      "entropy": 0.028627974635946886,
      "epoch": 0.1406358924620374,
      "grad_norm": 0.88671875,
      "learning_rate": 1.5788888888888888e-05,
      "loss": 0.03192804753780365,
      "mean_token_accuracy": 0.9823863580822945,
      "num_tokens": 2359848.0,
      "step": 580
    },
    {
      "entropy": 0.027760646163096682,
      "epoch": 0.14306064922862424,
      "grad_norm": 1.4140625,
      "learning_rate": 1.567777777777778e-05,
      "loss": 0.023938073217868804,
      "mean_token_accuracy": 0.9863636314868927,
      "num_tokens": 2398716.0,
      "step": 590
    },
    {
      "entropy": 0.01857751572433699,
      "epoch": 0.1454854059952111,
      "grad_norm": 2.0,
      "learning_rate": 1.5566666666666668e-05,
      "loss": 0.021452714502811433,
      "mean_token_accuracy": 0.9909090891480445,
      "num_tokens": 2439221.0,
      "step": 600
    },
    {
      "entropy": 0.020551329124737096,
      "epoch": 0.14791016276179797,
      "grad_norm": 2.09375,
      "learning_rate": 1.5455555555555556e-05,
      "loss": 0.03473537266254425,
      "mean_token_accuracy": 0.9840909041464329,
      "num_tokens": 2477390.0,
      "step": 610
    },
    {
      "entropy": 0.025016580805640842,
      "epoch": 0.15033491952838482,
      "grad_norm": 1.828125,
      "learning_rate": 1.5344444444444447e-05,
      "loss": 0.022924990952014924,
      "mean_token_accuracy": 0.9863636314868927,
      "num_tokens": 2515438.0,
      "step": 620
    },
    {
      "entropy": 0.02900541319813783,
      "epoch": 0.15275967629497167,
      "grad_norm": 1.6171875,
      "learning_rate": 1.5233333333333335e-05,
      "loss": 0.030890285968780518,
      "mean_token_accuracy": 0.9829545401036739,
      "num_tokens": 2554303.0,
      "step": 630
    },
    {
      "entropy": 0.027651957263333315,
      "epoch": 0.15518443306155852,
      "grad_norm": 1.3515625,
      "learning_rate": 1.5122222222222222e-05,
      "loss": 0.026059484481811522,
      "mean_token_accuracy": 0.9874999970197678,
      "num_tokens": 2593917.0,
      "step": 640
    },
    {
      "entropy": 0.025979451155399146,
      "epoch": 0.15760918982814537,
      "grad_norm": 0.51171875,
      "learning_rate": 1.5011111111111112e-05,
      "loss": 0.026576137542724608,
      "mean_token_accuracy": 0.9846590891480446,
      "num_tokens": 2633471.0,
      "step": 650
    },
    {
      "entropy": 0.0226533382314301,
      "epoch": 0.16003394659473222,
      "grad_norm": 3.703125,
      "learning_rate": 1.4900000000000001e-05,
      "loss": 0.03458902537822724,
      "mean_token_accuracy": 0.9806818149983882,
      "num_tokens": 2673584.0,
      "step": 660
    },
    {
      "entropy": 0.026275359961891807,
      "epoch": 0.16245870336131907,
      "grad_norm": 1.7421875,
      "learning_rate": 1.4788888888888891e-05,
      "loss": 0.023309774696826935,
      "mean_token_accuracy": 0.9886363610625267,
      "num_tokens": 2718495.0,
      "step": 670
    },
    {
      "entropy": 0.02142264891299419,
      "epoch": 0.16488346012790592,
      "grad_norm": 1.40625,
      "learning_rate": 1.467777777777778e-05,
      "loss": 0.027975478768348695,
      "mean_token_accuracy": 0.9863636322319508,
      "num_tokens": 2758475.0,
      "step": 680
    },
    {
      "entropy": 0.023169603949736484,
      "epoch": 0.16730821689449277,
      "grad_norm": 3.46875,
      "learning_rate": 1.4566666666666669e-05,
      "loss": 0.027786874771118165,
      "mean_token_accuracy": 0.9846320301294327,
      "num_tokens": 2801461.0,
      "step": 690
    },
    {
      "entropy": 0.025932194874985725,
      "epoch": 0.16973297366107962,
      "grad_norm": 1.40625,
      "learning_rate": 1.4455555555555555e-05,
      "loss": 0.024787703156471254,
      "mean_token_accuracy": 0.9863636329770088,
      "num_tokens": 2843561.0,
      "step": 700
    },
    {
      "entropy": 0.02573719730553421,
      "epoch": 0.17215773042766647,
      "grad_norm": 1.1953125,
      "learning_rate": 1.4344444444444445e-05,
      "loss": 0.023608550429344177,
      "mean_token_accuracy": 0.9874999955296516,
      "num_tokens": 2883906.0,
      "step": 710
    },
    {
      "entropy": 0.023726833695400275,
      "epoch": 0.17458248719425332,
      "grad_norm": 1.828125,
      "learning_rate": 1.4233333333333335e-05,
      "loss": 0.02043382525444031,
      "mean_token_accuracy": 0.98920454159379,
      "num_tokens": 2925983.0,
      "step": 720
    },
    {
      "entropy": 0.02000678656895616,
      "epoch": 0.17700724396084017,
      "grad_norm": 4.125,
      "learning_rate": 1.4122222222222225e-05,
      "loss": 0.03365764915943146,
      "mean_token_accuracy": 0.9886363618075847,
      "num_tokens": 2964454.0,
      "step": 730
    },
    {
      "entropy": 0.02043497935082996,
      "epoch": 0.17943200072742704,
      "grad_norm": 2.859375,
      "learning_rate": 1.4011111111111113e-05,
      "loss": 0.02920505404472351,
      "mean_token_accuracy": 0.9857954502105712,
      "num_tokens": 3005068.0,
      "step": 740
    },
    {
      "entropy": 0.029566307516233792,
      "epoch": 0.1818567574940139,
      "grad_norm": 1.3828125,
      "learning_rate": 1.39e-05,
      "loss": 0.029777100682258605,
      "mean_token_accuracy": 0.9852272711694241,
      "num_tokens": 3046898.0,
      "step": 750
    },
    {
      "entropy": 0.024274932581920438,
      "epoch": 0.18428151426060074,
      "grad_norm": 1.265625,
      "learning_rate": 1.3788888888888889e-05,
      "loss": 0.0237515389919281,
      "mean_token_accuracy": 0.9869318157434464,
      "num_tokens": 3088514.0,
      "step": 760
    },
    {
      "entropy": 0.02470423321283306,
      "epoch": 0.1867062710271876,
      "grad_norm": 1.125,
      "learning_rate": 1.3677777777777779e-05,
      "loss": 0.030222481489181517,
      "mean_token_accuracy": 0.9846590884029865,
      "num_tokens": 3129575.0,
      "step": 770
    },
    {
      "entropy": 0.028905244976613176,
      "epoch": 0.18913102779377444,
      "grad_norm": 1.7421875,
      "learning_rate": 1.3566666666666668e-05,
      "loss": 0.0264810711145401,
      "mean_token_accuracy": 0.985227269679308,
      "num_tokens": 3170056.0,
      "step": 780
    },
    {
      "entropy": 0.024943281395098892,
      "epoch": 0.1915557845603613,
      "grad_norm": 1.7578125,
      "learning_rate": 1.3455555555555556e-05,
      "loss": 0.029622524976730347,
      "mean_token_accuracy": 0.9823863595724106,
      "num_tokens": 3210101.0,
      "step": 790
    },
    {
      "entropy": 0.03016873870701602,
      "epoch": 0.19398054132694814,
      "grad_norm": 1.3046875,
      "learning_rate": 1.3344444444444446e-05,
      "loss": 0.035280358791351316,
      "mean_token_accuracy": 0.9806818120181561,
      "num_tokens": 3252969.0,
      "step": 800
    },
    {
      "entropy": 0.0313864836067296,
      "epoch": 0.196405298093535,
      "grad_norm": 2.046875,
      "learning_rate": 1.3233333333333334e-05,
      "loss": 0.02824702560901642,
      "mean_token_accuracy": 0.9823863610625267,
      "num_tokens": 3293318.0,
      "step": 810
    },
    {
      "entropy": 0.031964658715151015,
      "epoch": 0.19883005486012184,
      "grad_norm": 0.83984375,
      "learning_rate": 1.3122222222222222e-05,
      "loss": 0.02947427034378052,
      "mean_token_accuracy": 0.9863636322319508,
      "num_tokens": 3338604.0,
      "step": 820
    },
    {
      "entropy": 0.024338746762441587,
      "epoch": 0.2012548116267087,
      "grad_norm": 1.9609375,
      "learning_rate": 1.3011111111111112e-05,
      "loss": 0.023631574213504793,
      "mean_token_accuracy": 0.984090905636549,
      "num_tokens": 3376591.0,
      "step": 830
    },
    {
      "entropy": 0.022459183935188774,
      "epoch": 0.20367956839329554,
      "grad_norm": 2.390625,
      "learning_rate": 1.2900000000000002e-05,
      "loss": 0.029895466566085816,
      "mean_token_accuracy": 0.9846590891480446,
      "num_tokens": 3417635.0,
      "step": 840
    },
    {
      "entropy": 0.022913600040919847,
      "epoch": 0.2061043251598824,
      "grad_norm": 2.296875,
      "learning_rate": 1.278888888888889e-05,
      "loss": 0.02521446943283081,
      "mean_token_accuracy": 0.9857954502105712,
      "num_tokens": 3459294.0,
      "step": 850
    },
    {
      "entropy": 0.025265456292163434,
      "epoch": 0.20852908192646924,
      "grad_norm": 1.4765625,
      "learning_rate": 1.2677777777777778e-05,
      "loss": 0.022895295917987824,
      "mean_token_accuracy": 0.9886363618075847,
      "num_tokens": 3500919.0,
      "step": 860
    },
    {
      "entropy": 0.026354352129692414,
      "epoch": 0.2109538386930561,
      "grad_norm": 1.2265625,
      "learning_rate": 1.2566666666666668e-05,
      "loss": 0.02891395688056946,
      "mean_token_accuracy": 0.9857954509556294,
      "num_tokens": 3542022.0,
      "step": 870
    },
    {
      "entropy": 0.026587546153950824,
      "epoch": 0.21337859545964297,
      "grad_norm": 2.15625,
      "learning_rate": 1.2455555555555556e-05,
      "loss": 0.025007185339927674,
      "mean_token_accuracy": 0.9880681782960892,
      "num_tokens": 3580600.0,
      "step": 880
    },
    {
      "entropy": 0.021694945237868524,
      "epoch": 0.21580335222622982,
      "grad_norm": 1.578125,
      "learning_rate": 1.2344444444444446e-05,
      "loss": 0.025805145502090454,
      "mean_token_accuracy": 0.9874999970197678,
      "num_tokens": 3620910.0,
      "step": 890
    },
    {
      "entropy": 0.02144139350803016,
      "epoch": 0.21822810899281667,
      "grad_norm": 1.90625,
      "learning_rate": 1.2233333333333335e-05,
      "loss": 0.020484310388565064,
      "mean_token_accuracy": 0.9892045423388481,
      "num_tokens": 3660844.0,
      "step": 900
    },
    {
      "entropy": 0.02579846643930068,
      "epoch": 0.22065286575940352,
      "grad_norm": 1.6640625,
      "learning_rate": 1.2122222222222223e-05,
      "loss": 0.03399317860603333,
      "mean_token_accuracy": 0.9812499992549419,
      "num_tokens": 3702701.0,
      "step": 910
    },
    {
      "entropy": 0.025248465246386333,
      "epoch": 0.22307762252599037,
      "grad_norm": 1.4140625,
      "learning_rate": 1.2011111111111111e-05,
      "loss": 0.025516167283058167,
      "mean_token_accuracy": 0.9846590876579284,
      "num_tokens": 3744019.0,
      "step": 920
    },
    {
      "entropy": 0.02416455528455117,
      "epoch": 0.22550237929257722,
      "grad_norm": 1.234375,
      "learning_rate": 1.1900000000000001e-05,
      "loss": 0.0251032292842865,
      "mean_token_accuracy": 0.9857683956623078,
      "num_tokens": 3783437.0,
      "step": 930
    },
    {
      "entropy": 0.024388600795100503,
      "epoch": 0.22792713605916406,
      "grad_norm": 5.96875,
      "learning_rate": 1.178888888888889e-05,
      "loss": 0.026069679856300355,
      "mean_token_accuracy": 0.984090905636549,
      "num_tokens": 3823298.0,
      "step": 940
    },
    {
      "entropy": 0.02505718378297388,
      "epoch": 0.23035189282575091,
      "grad_norm": 1.03125,
      "learning_rate": 1.1677777777777779e-05,
      "loss": 0.025333744287490845,
      "mean_token_accuracy": 0.9874999970197678,
      "num_tokens": 3863057.0,
      "step": 950
    },
    {
      "entropy": 0.02888807645626912,
      "epoch": 0.23277664959233776,
      "grad_norm": 2.640625,
      "learning_rate": 1.1566666666666669e-05,
      "loss": 0.02982906103134155,
      "mean_token_accuracy": 0.9852272674441338,
      "num_tokens": 3902174.0,
      "step": 960
    },
    {
      "entropy": 0.026199724662365042,
      "epoch": 0.23520140635892461,
      "grad_norm": 1.53125,
      "learning_rate": 1.1455555555555557e-05,
      "loss": 0.03483479619026184,
      "mean_token_accuracy": 0.9795454509556294,
      "num_tokens": 3940800.0,
      "step": 970
    },
    {
      "entropy": 0.02737266756594181,
      "epoch": 0.23762616312551146,
      "grad_norm": 2.1875,
      "learning_rate": 1.1344444444444445e-05,
      "loss": 0.024009138345718384,
      "mean_token_accuracy": 0.9874999970197678,
      "num_tokens": 3980912.0,
      "step": 980
    },
    {
      "entropy": 0.02674058018892538,
      "epoch": 0.2400509198920983,
      "grad_norm": 0.83203125,
      "learning_rate": 1.1233333333333333e-05,
      "loss": 0.02231251299381256,
      "mean_token_accuracy": 0.9880681790411472,
      "num_tokens": 4020042.0,
      "step": 990
    },
    {
      "entropy": 0.023251514391176897,
      "epoch": 0.24247567665868516,
      "grad_norm": 1.8515625,
      "learning_rate": 1.1122222222222223e-05,
      "loss": 0.02937265336513519,
      "mean_token_accuracy": 0.9801136322319508,
      "num_tokens": 4056666.0,
      "step": 1000
    },
    {
      "entropy": 0.023756788905575376,
      "epoch": 0.24490043342527204,
      "grad_norm": 2.875,
      "learning_rate": 1.1011111111111113e-05,
      "loss": 0.021810129284858704,
      "mean_token_accuracy": 0.9880681782960892,
      "num_tokens": 4100066.0,
      "step": 1010
    },
    {
      "entropy": 0.023163225039979807,
      "epoch": 0.2473251901918589,
      "grad_norm": 1.6875,
      "learning_rate": 1.0900000000000002e-05,
      "loss": 0.02502768337726593,
      "mean_token_accuracy": 0.9874999955296516,
      "num_tokens": 4139561.0,
      "step": 1020
    },
    {
      "entropy": 0.02136901097264854,
      "epoch": 0.24974994695844574,
      "grad_norm": 1.609375,
      "learning_rate": 1.0788888888888889e-05,
      "loss": 0.019149383902549742,
      "mean_token_accuracy": 0.9897456675767898,
      "num_tokens": 4177327.0,
      "step": 1030
    },
    {
      "entropy": 0.02029619142294905,
      "epoch": 0.2521747037250326,
      "grad_norm": 2.328125,
      "learning_rate": 1.0677777777777779e-05,
      "loss": 0.031534403562545776,
      "mean_token_accuracy": 0.984090905636549,
      "num_tokens": 4218398.0,
      "step": 1040
    },
    {
      "entropy": 0.02094364000367932,
      "epoch": 0.2545994604916194,
      "grad_norm": 2.546875,
      "learning_rate": 1.0566666666666667e-05,
      "loss": 0.03177621364593506,
      "mean_token_accuracy": 0.9869318149983883,
      "num_tokens": 4261625.0,
      "step": 1050
    },
    {
      "entropy": 0.025660122126373607,
      "epoch": 0.2570242172582063,
      "grad_norm": 1.03125,
      "learning_rate": 1.0455555555555556e-05,
      "loss": 0.029423072934150696,
      "mean_token_accuracy": 0.9846590876579284,
      "num_tokens": 4300872.0,
      "step": 1060
    },
    {
      "entropy": 0.027266277611670376,
      "epoch": 0.2594489740247931,
      "grad_norm": 1.3515625,
      "learning_rate": 1.0344444444444446e-05,
      "loss": 0.028110802173614502,
      "mean_token_accuracy": 0.9840909071266651,
      "num_tokens": 4342744.0,
      "step": 1070
    },
    {
      "entropy": 0.02724119477079512,
      "epoch": 0.26187373079138,
      "grad_norm": 1.0390625,
      "learning_rate": 1.0233333333333336e-05,
      "loss": 0.023805345594882964,
      "mean_token_accuracy": 0.9863636329770088,
      "num_tokens": 4382536.0,
      "step": 1080
    },
    {
      "entropy": 0.02513553084399973,
      "epoch": 0.26429848755796687,
      "grad_norm": 1.265625,
      "learning_rate": 1.0122222222222222e-05,
      "loss": 0.026082912087440492,
      "mean_token_accuracy": 0.9874999962747097,
      "num_tokens": 4427451.0,
      "step": 1090
    },
    {
      "entropy": 0.024046725058724405,
      "epoch": 0.2667232443245537,
      "grad_norm": 1.84375,
      "learning_rate": 1.0011111111111112e-05,
      "loss": 0.025588032603263856,
      "mean_token_accuracy": 0.9869318135082722,
      "num_tokens": 4464955.0,
      "step": 1100
    },
    {
      "entropy": 0.022660849900512404,
      "epoch": 0.26914800109114057,
      "grad_norm": 1.421875,
      "learning_rate": 9.9e-06,
      "loss": 0.021866945922374724,
      "mean_token_accuracy": 0.9880681782960892,
      "num_tokens": 4505114.0,
      "step": 1110
    },
    {
      "entropy": 0.02450128130503799,
      "epoch": 0.2715727578577274,
      "grad_norm": 1.28125,
      "learning_rate": 9.78888888888889e-06,
      "loss": 0.024013830721378325,
      "mean_token_accuracy": 0.9869318149983883,
      "num_tokens": 4545307.0,
      "step": 1120
    },
    {
      "entropy": 0.023239308406709826,
      "epoch": 0.27399751462431426,
      "grad_norm": 1.9296875,
      "learning_rate": 9.677777777777778e-06,
      "loss": 0.026031026244163515,
      "mean_token_accuracy": 0.9869318135082722,
      "num_tokens": 4583174.0,
      "step": 1130
    },
    {
      "entropy": 0.02481268622832431,
      "epoch": 0.2764222713909011,
      "grad_norm": 0.859375,
      "learning_rate": 9.566666666666668e-06,
      "loss": 0.028152582049369813,
      "mean_token_accuracy": 0.98295454159379,
      "num_tokens": 4623363.0,
      "step": 1140
    },
    {
      "entropy": 0.02144863496614562,
      "epoch": 0.27884702815748796,
      "grad_norm": 1.8515625,
      "learning_rate": 9.455555555555557e-06,
      "loss": 0.018676747381687165,
      "mean_token_accuracy": 0.9897727243602276,
      "num_tokens": 4663530.0,
      "step": 1150
    },
    {
      "entropy": 0.021853329628083885,
      "epoch": 0.2812717849240748,
      "grad_norm": 1.03125,
      "learning_rate": 9.344444444444446e-06,
      "loss": 0.023481160402297974,
      "mean_token_accuracy": 0.9880681790411472,
      "num_tokens": 4703751.0,
      "step": 1160
    },
    {
      "entropy": 0.020845431580073636,
      "epoch": 0.28369654169066166,
      "grad_norm": 1.828125,
      "learning_rate": 9.233333333333334e-06,
      "loss": 0.025833714008331298,
      "mean_token_accuracy": 0.9892045423388481,
      "num_tokens": 4742890.0,
      "step": 1170
    },
    {
      "entropy": 0.020562034113299887,
      "epoch": 0.2861212984572485,
      "grad_norm": 3.359375,
      "learning_rate": 9.122222222222223e-06,
      "loss": 0.02122436910867691,
      "mean_token_accuracy": 0.9869318149983883,
      "num_tokens": 4786008.0,
      "step": 1180
    },
    {
      "entropy": 0.0207685565927477,
      "epoch": 0.28854605522383536,
      "grad_norm": 0.92578125,
      "learning_rate": 9.011111111111111e-06,
      "loss": 0.016763387620449065,
      "mean_token_accuracy": 0.9914772696793079,
      "num_tokens": 4829612.0,
      "step": 1190
    },
    {
      "entropy": 0.022811952112988365,
      "epoch": 0.2909708119904222,
      "grad_norm": 2.5625,
      "learning_rate": 8.900000000000001e-06,
      "loss": 0.025601431727409363,
      "mean_token_accuracy": 0.985227269679308,
      "num_tokens": 4871202.0,
      "step": 1200
    },
    {
      "entropy": 0.021590291650136352,
      "epoch": 0.29339556875700906,
      "grad_norm": 2.09375,
      "learning_rate": 8.788888888888891e-06,
      "loss": 0.02018619328737259,
      "mean_token_accuracy": 0.9886363603174686,
      "num_tokens": 4911931.0,
      "step": 1210
    },
    {
      "entropy": 0.02163595509409788,
      "epoch": 0.29582032552359594,
      "grad_norm": 1.8984375,
      "learning_rate": 8.677777777777779e-06,
      "loss": 0.025696516036987305,
      "mean_token_accuracy": 0.9863636329770088,
      "num_tokens": 4951183.0,
      "step": 1220
    },
    {
      "entropy": 0.02311627586459508,
      "epoch": 0.29824508229018276,
      "grad_norm": 0.83203125,
      "learning_rate": 8.566666666666667e-06,
      "loss": 0.028527703881263734,
      "mean_token_accuracy": 0.9835227243602276,
      "num_tokens": 4994778.0,
      "step": 1230
    },
    {
      "entropy": 0.020521293884030456,
      "epoch": 0.30066983905676964,
      "grad_norm": 0.7421875,
      "learning_rate": 8.455555555555555e-06,
      "loss": 0.021116530895233153,
      "mean_token_accuracy": 0.9914772696793079,
      "num_tokens": 5034446.0,
      "step": 1240
    },
    {
      "entropy": 0.021558842277681834,
      "epoch": 0.30309459582335646,
      "grad_norm": 1.1953125,
      "learning_rate": 8.344444444444445e-06,
      "loss": 0.022282668948173524,
      "mean_token_accuracy": 0.9897727251052857,
      "num_tokens": 5074406.0,
      "step": 1250
    },
    {
      "entropy": 0.01967097778419884,
      "epoch": 0.30551935258994334,
      "grad_norm": 1.1953125,
      "learning_rate": 8.233333333333335e-06,
      "loss": 0.0207993745803833,
      "mean_token_accuracy": 0.9920454524457455,
      "num_tokens": 5121427.0,
      "step": 1260
    },
    {
      "entropy": 0.02370198413136677,
      "epoch": 0.30794410935653016,
      "grad_norm": 0.85546875,
      "learning_rate": 8.122222222222223e-06,
      "loss": 0.028493967652320863,
      "mean_token_accuracy": 0.9846590861678124,
      "num_tokens": 5160392.0,
      "step": 1270
    },
    {
      "entropy": 0.023408127912307463,
      "epoch": 0.31036886612311704,
      "grad_norm": 3.046875,
      "learning_rate": 8.011111111111113e-06,
      "loss": 0.024442848563194276,
      "mean_token_accuracy": 0.9857954502105712,
      "num_tokens": 5199665.0,
      "step": 1280
    },
    {
      "entropy": 0.018637936454115333,
      "epoch": 0.31279362288970386,
      "grad_norm": 1.59375,
      "learning_rate": 7.9e-06,
      "loss": 0.02052522748708725,
      "mean_token_accuracy": 0.9869318135082722,
      "num_tokens": 5237561.0,
      "step": 1290
    },
    {
      "entropy": 0.02345360331178199,
      "epoch": 0.31521837965629074,
      "grad_norm": 2.234375,
      "learning_rate": 7.788888888888889e-06,
      "loss": 0.03127675354480743,
      "mean_token_accuracy": 0.9840909071266651,
      "num_tokens": 5276996.0,
      "step": 1300
    },
    {
      "entropy": 0.024497390102942516,
      "epoch": 0.31764313642287756,
      "grad_norm": 1.3984375,
      "learning_rate": 7.677777777777778e-06,
      "loss": 0.02247738242149353,
      "mean_token_accuracy": 0.9874999955296516,
      "num_tokens": 5318441.0,
      "step": 1310
    },
    {
      "entropy": 0.018883616356470158,
      "epoch": 0.32006789318946444,
      "grad_norm": 3.203125,
      "learning_rate": 7.566666666666667e-06,
      "loss": 0.01997789889574051,
      "mean_token_accuracy": 0.9897727236151695,
      "num_tokens": 5358209.0,
      "step": 1320
    },
    {
      "entropy": 0.024151582821377814,
      "epoch": 0.32249264995605126,
      "grad_norm": 2.078125,
      "learning_rate": 7.455555555555556e-06,
      "loss": 0.028815129399299623,
      "mean_token_accuracy": 0.9823863588273525,
      "num_tokens": 5397538.0,
      "step": 1330
    },
    {
      "entropy": 0.020284308634154515,
      "epoch": 0.32491740672263814,
      "grad_norm": 1.046875,
      "learning_rate": 7.344444444444445e-06,
      "loss": 0.019358839094638824,
      "mean_token_accuracy": 0.9897727258503437,
      "num_tokens": 5435900.0,
      "step": 1340
    },
    {
      "entropy": 0.02354962546473871,
      "epoch": 0.327342163489225,
      "grad_norm": 1.3515625,
      "learning_rate": 7.233333333333334e-06,
      "loss": 0.031699517369270326,
      "mean_token_accuracy": 0.9835227228701114,
      "num_tokens": 5476573.0,
      "step": 1350
    },
    {
      "entropy": 0.02446860164272948,
      "epoch": 0.32976692025581184,
      "grad_norm": 0.92578125,
      "learning_rate": 7.122222222222222e-06,
      "loss": 0.023002330958843232,
      "mean_token_accuracy": 0.9869318142533302,
      "num_tokens": 5513779.0,
      "step": 1360
    },
    {
      "entropy": 0.022691848923705037,
      "epoch": 0.3321916770223987,
      "grad_norm": 3.484375,
      "learning_rate": 7.011111111111112e-06,
      "loss": 0.028314879536628722,
      "mean_token_accuracy": 0.984090904891491,
      "num_tokens": 5555000.0,
      "step": 1370
    },
    {
      "entropy": 0.019440177543947355,
      "epoch": 0.33461643378898553,
      "grad_norm": 0.59375,
      "learning_rate": 6.9e-06,
      "loss": 0.01770046055316925,
      "mean_token_accuracy": 0.9914772711694241,
      "num_tokens": 5593016.0,
      "step": 1380
    },
    {
      "entropy": 0.023765943434227665,
      "epoch": 0.3370411905555724,
      "grad_norm": 1.09375,
      "learning_rate": 6.788888888888889e-06,
      "loss": 0.02077813595533371,
      "mean_token_accuracy": 0.9903409063816071,
      "num_tokens": 5635379.0,
      "step": 1390
    },
    {
      "entropy": 0.01887716218980131,
      "epoch": 0.33946594732215923,
      "grad_norm": 1.90625,
      "learning_rate": 6.677777777777779e-06,
      "loss": 0.021604160964488982,
      "mean_token_accuracy": 0.9880681782960892,
      "num_tokens": 5673576.0,
      "step": 1400
    },
    {
      "entropy": 0.023854930577999766,
      "epoch": 0.3418907040887461,
      "grad_norm": 4.21875,
      "learning_rate": 6.566666666666667e-06,
      "loss": 0.025638672709465026,
      "mean_token_accuracy": 0.9846590854227543,
      "num_tokens": 5714108.0,
      "step": 1410
    },
    {
      "entropy": 0.020290096460576022,
      "epoch": 0.34431546085533293,
      "grad_norm": 0.9609375,
      "learning_rate": 6.455555555555556e-06,
      "loss": 0.020940312743186952,
      "mean_token_accuracy": 0.9909090884029865,
      "num_tokens": 5755604.0,
      "step": 1420
    },
    {
      "entropy": 0.019387308810064497,
      "epoch": 0.3467402176219198,
      "grad_norm": 3.09375,
      "learning_rate": 6.3444444444444454e-06,
      "loss": 0.02511414885520935,
      "mean_token_accuracy": 0.9880681790411472,
      "num_tokens": 5793500.0,
      "step": 1430
    },
    {
      "entropy": 0.02229226140134415,
      "epoch": 0.34916497438850663,
      "grad_norm": 0.8359375,
      "learning_rate": 6.2333333333333335e-06,
      "loss": 0.0231765940785408,
      "mean_token_accuracy": 0.9869318157434464,
      "num_tokens": 5833950.0,
      "step": 1440
    },
    {
      "entropy": 0.02084504545036907,
      "epoch": 0.3515897311550935,
      "grad_norm": 1.734375,
      "learning_rate": 6.1222222222222224e-06,
      "loss": 0.03049103319644928,
      "mean_token_accuracy": 0.9835227228701114,
      "num_tokens": 5872957.0,
      "step": 1450
    },
    {
      "entropy": 0.01889653230196018,
      "epoch": 0.35401448792168033,
      "grad_norm": 1.4765625,
      "learning_rate": 6.011111111111112e-06,
      "loss": 0.019928357005119322,
      "mean_token_accuracy": 0.990340905636549,
      "num_tokens": 5913588.0,
      "step": 1460
    },
    {
      "entropy": 0.019694303501910326,
      "epoch": 0.3564392446882672,
      "grad_norm": 3.5,
      "learning_rate": 5.9e-06,
      "loss": 0.01947687119245529,
      "mean_token_accuracy": 0.9909090884029865,
      "num_tokens": 5953428.0,
      "step": 1470
    },
    {
      "entropy": 0.019155515848706273,
      "epoch": 0.3588640014548541,
      "grad_norm": 2.671875,
      "learning_rate": 5.788888888888889e-06,
      "loss": 0.022465848922729494,
      "mean_token_accuracy": 0.9897727243602276,
      "num_tokens": 5993281.0,
      "step": 1480
    },
    {
      "entropy": 0.021128093433253524,
      "epoch": 0.3612887582214409,
      "grad_norm": 1.8203125,
      "learning_rate": 5.677777777777779e-06,
      "loss": 0.027405399084091186,
      "mean_token_accuracy": 0.9835227221250534,
      "num_tokens": 6035357.0,
      "step": 1490
    },
    {
      "entropy": 0.022639484232195174,
      "epoch": 0.3637135149880278,
      "grad_norm": 2.265625,
      "learning_rate": 5.566666666666667e-06,
      "loss": 0.021980252861976624,
      "mean_token_accuracy": 0.9892045430839062,
      "num_tokens": 6077438.0,
      "step": 1500
    },
    {
      "entropy": 0.021943628234339485,
      "epoch": 0.3661382717546146,
      "grad_norm": 3.640625,
      "learning_rate": 5.455555555555556e-06,
      "loss": 0.03260737359523773,
      "mean_token_accuracy": 0.9852272681891918,
      "num_tokens": 6117468.0,
      "step": 1510
    },
    {
      "entropy": 0.025014979231627876,
      "epoch": 0.3685630285212015,
      "grad_norm": 1.0546875,
      "learning_rate": 5.344444444444446e-06,
      "loss": 0.023974181711673738,
      "mean_token_accuracy": 0.9869318142533302,
      "num_tokens": 6157286.0,
      "step": 1520
    },
    {
      "entropy": 0.01908629078325248,
      "epoch": 0.3709877852877883,
      "grad_norm": 1.15625,
      "learning_rate": 5.233333333333334e-06,
      "loss": 0.015591822564601898,
      "mean_token_accuracy": 0.9926136344671249,
      "num_tokens": 6198312.0,
      "step": 1530
    },
    {
      "entropy": 0.020979891011893414,
      "epoch": 0.3734125420543752,
      "grad_norm": 1.8515625,
      "learning_rate": 5.122222222222223e-06,
      "loss": 0.022910787165164946,
      "mean_token_accuracy": 0.989559655636549,
      "num_tokens": 6240280.0,
      "step": 1540
    },
    {
      "entropy": 0.025256870118482766,
      "epoch": 0.375837298820962,
      "grad_norm": 3.25,
      "learning_rate": 5.011111111111111e-06,
      "loss": 0.03658706545829773,
      "mean_token_accuracy": 0.9801136322319508,
      "num_tokens": 6284875.0,
      "step": 1550
    },
    {
      "entropy": 0.02342147537783603,
      "epoch": 0.3782620555875489,
      "grad_norm": 1.2734375,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.025882428884506224,
      "mean_token_accuracy": 0.9738636329770088,
      "num_tokens": 6329857.0,
      "step": 1560
    },
    {
      "entropy": 0.0226567227995929,
      "epoch": 0.3806868123541357,
      "grad_norm": 1.265625,
      "learning_rate": 4.7888888888888894e-06,
      "loss": 0.018890340626239777,
      "mean_token_accuracy": 0.990340905636549,
      "num_tokens": 6370302.0,
      "step": 1570
    },
    {
      "entropy": 0.022661514219635137,
      "epoch": 0.3831115691207226,
      "grad_norm": 1.921875,
      "learning_rate": 4.677777777777778e-06,
      "loss": 0.021921022236347197,
      "mean_token_accuracy": 0.9874999985098839,
      "num_tokens": 6407249.0,
      "step": 1580
    },
    {
      "entropy": 0.022505474984882312,
      "epoch": 0.3855363258873094,
      "grad_norm": 4.125,
      "learning_rate": 4.566666666666667e-06,
      "loss": 0.0322005420923233,
      "mean_token_accuracy": 0.9846590884029865,
      "num_tokens": 6446978.0,
      "step": 1590
    },
    {
      "entropy": 0.022300617951987078,
      "epoch": 0.3879610826538963,
      "grad_norm": 3.046875,
      "learning_rate": 4.455555555555555e-06,
      "loss": 0.023643006384372712,
      "mean_token_accuracy": 0.98920454159379,
      "num_tokens": 6486520.0,
      "step": 1600
    },
    {
      "entropy": 0.02587687493778503,
      "epoch": 0.3903858394204831,
      "grad_norm": 2.78125,
      "learning_rate": 4.344444444444445e-06,
      "loss": 0.027235689759254455,
      "mean_token_accuracy": 0.985227269679308,
      "num_tokens": 6528137.0,
      "step": 1610
    },
    {
      "entropy": 0.021340829327573373,
      "epoch": 0.39281059618707,
      "grad_norm": 0.609375,
      "learning_rate": 4.233333333333334e-06,
      "loss": 0.021160535514354706,
      "mean_token_accuracy": 0.9880681797862053,
      "num_tokens": 6565831.0,
      "step": 1620
    },
    {
      "entropy": 0.02321274704027019,
      "epoch": 0.39523535295365686,
      "grad_norm": 1.609375,
      "learning_rate": 4.122222222222222e-06,
      "loss": 0.026870492100715637,
      "mean_token_accuracy": 0.9869318164885044,
      "num_tokens": 6605622.0,
      "step": 1630
    },
    {
      "entropy": 0.024162391825620945,
      "epoch": 0.3976601097202437,
      "grad_norm": 1.15625,
      "learning_rate": 4.011111111111111e-06,
      "loss": 0.025860154628753663,
      "mean_token_accuracy": 0.9874999970197678,
      "num_tokens": 6647782.0,
      "step": 1640
    },
    {
      "entropy": 0.024304146416397998,
      "epoch": 0.40008486648683056,
      "grad_norm": 2.5625,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.02480360269546509,
      "mean_token_accuracy": 0.9857954509556294,
      "num_tokens": 6688774.0,
      "step": 1650
    },
    {
      "entropy": 0.024905602099897804,
      "epoch": 0.4025096232534174,
      "grad_norm": 0.828125,
      "learning_rate": 3.7888888888888893e-06,
      "loss": 0.02738499641418457,
      "mean_token_accuracy": 0.9863636329770088,
      "num_tokens": 6731921.0,
      "step": 1660
    },
    {
      "entropy": 0.022832987389074334,
      "epoch": 0.40493438002000426,
      "grad_norm": 2.390625,
      "learning_rate": 3.6777777777777778e-06,
      "loss": 0.027183347940444948,
      "mean_token_accuracy": 0.9840909034013748,
      "num_tokens": 6771015.0,
      "step": 1670
    },
    {
      "entropy": 0.024453511039519073,
      "epoch": 0.4073591367865911,
      "grad_norm": 2.515625,
      "learning_rate": 3.566666666666667e-06,
      "loss": 0.027268880605697633,
      "mean_token_accuracy": 0.9886363595724106,
      "num_tokens": 6811078.0,
      "step": 1680
    },
    {
      "entropy": 0.02214456081451317,
      "epoch": 0.40978389355317796,
      "grad_norm": 0.98828125,
      "learning_rate": 3.455555555555556e-06,
      "loss": 0.02482033222913742,
      "mean_token_accuracy": 0.9874999970197678,
      "num_tokens": 6848500.0,
      "step": 1690
    },
    {
      "entropy": 0.02487115224266745,
      "epoch": 0.4122086503197648,
      "grad_norm": 2.5,
      "learning_rate": 3.3444444444444445e-06,
      "loss": 0.021296854317188262,
      "mean_token_accuracy": 0.9897727251052857,
      "num_tokens": 6894353.0,
      "step": 1700
    },
    {
      "entropy": 0.025811467690800782,
      "epoch": 0.41463340708635166,
      "grad_norm": 0.9296875,
      "learning_rate": 3.2333333333333334e-06,
      "loss": 0.032750701904296874,
      "mean_token_accuracy": 0.9823863610625267,
      "num_tokens": 6937570.0,
      "step": 1710
    },
    {
      "entropy": 0.01883704749238859,
      "epoch": 0.4170581638529385,
      "grad_norm": 1.7734375,
      "learning_rate": 3.1222222222222228e-06,
      "loss": 0.023237355053424835,
      "mean_token_accuracy": 0.987472939491272,
      "num_tokens": 6979578.0,
      "step": 1720
    },
    {
      "entropy": 0.022163561169008972,
      "epoch": 0.41948292061952536,
      "grad_norm": 2.359375,
      "learning_rate": 3.0111111111111113e-06,
      "loss": 0.0246141716837883,
      "mean_token_accuracy": 0.9874999977648258,
      "num_tokens": 7015596.0,
      "step": 1730
    },
    {
      "entropy": 0.026256153864324006,
      "epoch": 0.4219076773861122,
      "grad_norm": 2.203125,
      "learning_rate": 2.9e-06,
      "loss": 0.028805309534072877,
      "mean_token_accuracy": 0.9823863595724106,
      "num_tokens": 7055852.0,
      "step": 1740
    },
    {
      "entropy": 0.02054688634939339,
      "epoch": 0.42433243415269906,
      "grad_norm": 1.3125,
      "learning_rate": 2.788888888888889e-06,
      "loss": 0.018951752781867982,
      "mean_token_accuracy": 0.991477270424366,
      "num_tokens": 7092886.0,
      "step": 1750
    },
    {
      "entropy": 0.024736029255382163,
      "epoch": 0.42675719091928593,
      "grad_norm": 1.1640625,
      "learning_rate": 2.677777777777778e-06,
      "loss": 0.026023554801940917,
      "mean_token_accuracy": 0.9874999970197678,
      "num_tokens": 7134382.0,
      "step": 1760
    },
    {
      "entropy": 0.023883560596868848,
      "epoch": 0.42918194768587276,
      "grad_norm": 2.171875,
      "learning_rate": 2.566666666666667e-06,
      "loss": 0.022034844756126402,
      "mean_token_accuracy": 0.9874999970197678,
      "num_tokens": 7175153.0,
      "step": 1770
    },
    {
      "entropy": 0.023680539545011925,
      "epoch": 0.43160670445245963,
      "grad_norm": 2.65625,
      "learning_rate": 2.455555555555556e-06,
      "loss": 0.021315962076187134,
      "mean_token_accuracy": 0.9880681790411472,
      "num_tokens": 7216758.0,
      "step": 1780
    },
    {
      "entropy": 0.025864069088902396,
      "epoch": 0.43403146121904645,
      "grad_norm": 3.640625,
      "learning_rate": 2.3444444444444448e-06,
      "loss": 0.023368486762046815,
      "mean_token_accuracy": 0.990340905636549,
      "num_tokens": 7257744.0,
      "step": 1790
    },
    {
      "entropy": 0.021552615765995143,
      "epoch": 0.43645621798563333,
      "grad_norm": 2.359375,
      "learning_rate": 2.2333333333333333e-06,
      "loss": 0.02335650324821472,
      "mean_token_accuracy": 0.9863636337220669,
      "num_tokens": 7293067.0,
      "step": 1800
    },
    {
      "entropy": 0.023214141586140612,
      "epoch": 0.43888097475222015,
      "grad_norm": 4.21875,
      "learning_rate": 2.1222222222222226e-06,
      "loss": 0.028553092479705812,
      "mean_token_accuracy": 0.9852272689342498,
      "num_tokens": 7334737.0,
      "step": 1810
    },
    {
      "entropy": 0.019011126364603115,
      "epoch": 0.44130573151880703,
      "grad_norm": 1.6796875,
      "learning_rate": 2.011111111111111e-06,
      "loss": 0.027167025208473205,
      "mean_token_accuracy": 0.9852272689342498,
      "num_tokens": 7377606.0,
      "step": 1820
    },
    {
      "entropy": 0.022004775167988554,
      "epoch": 0.44373048828539385,
      "grad_norm": 0.96484375,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 0.03003118634223938,
      "mean_token_accuracy": 0.981818176060915,
      "num_tokens": 7417763.0,
      "step": 1830
    },
    {
      "entropy": 0.022360649610618566,
      "epoch": 0.44615524505198073,
      "grad_norm": 1.0078125,
      "learning_rate": 1.788888888888889e-06,
      "loss": 0.02517186403274536,
      "mean_token_accuracy": 0.9863636329770088,
      "num_tokens": 7459114.0,
      "step": 1840
    },
    {
      "entropy": 0.02276606035607074,
      "epoch": 0.44858000181856755,
      "grad_norm": 1.390625,
      "learning_rate": 1.6777777777777779e-06,
      "loss": 0.020943556725978852,
      "mean_token_accuracy": 0.98920454159379,
      "num_tokens": 7498162.0,
      "step": 1850
    },
    {
      "entropy": 0.027342419610067736,
      "epoch": 0.45100475858515443,
      "grad_norm": 2.1875,
      "learning_rate": 1.566666666666667e-06,
      "loss": 0.02874574959278107,
      "mean_token_accuracy": 0.9846590876579284,
      "num_tokens": 7534837.0,
      "step": 1860
    },
    {
      "entropy": 0.02010235228917736,
      "epoch": 0.45342951535174125,
      "grad_norm": 1.375,
      "learning_rate": 1.4555555555555557e-06,
      "loss": 0.029143670201301576,
      "mean_token_accuracy": 0.9863636314868927,
      "num_tokens": 7575437.0,
      "step": 1870
    },
    {
      "entropy": 0.019995948372934436,
      "epoch": 0.45585427211832813,
      "grad_norm": 1.8203125,
      "learning_rate": 1.3444444444444446e-06,
      "loss": 0.026229032874107362,
      "mean_token_accuracy": 0.9846590861678124,
      "num_tokens": 7615212.0,
      "step": 1880
    },
    {
      "entropy": 0.019795780479307724,
      "epoch": 0.458279028884915,
      "grad_norm": 2.90625,
      "learning_rate": 1.2333333333333335e-06,
      "loss": 0.020921708643436433,
      "mean_token_accuracy": 0.9880681775510312,
      "num_tokens": 7661315.0,
      "step": 1890
    },
    {
      "entropy": 0.021343442313173,
      "epoch": 0.46070378565150183,
      "grad_norm": 0.703125,
      "learning_rate": 1.1222222222222222e-06,
      "loss": 0.02051656097173691,
      "mean_token_accuracy": 0.9880681790411472,
      "num_tokens": 7704690.0,
      "step": 1900
    },
    {
      "entropy": 0.02385749567615676,
      "epoch": 0.4631285424180887,
      "grad_norm": 3.765625,
      "learning_rate": 1.0111111111111111e-06,
      "loss": 0.029207244515419006,
      "mean_token_accuracy": 0.9880681790411472,
      "num_tokens": 7745332.0,
      "step": 1910
    },
    {
      "entropy": 0.024054063166568086,
      "epoch": 0.46555329918467553,
      "grad_norm": 2.1875,
      "learning_rate": 9.000000000000001e-07,
      "loss": 0.023930326104164124,
      "mean_token_accuracy": 0.9863636322319508,
      "num_tokens": 7785168.0,
      "step": 1920
    },
    {
      "entropy": 0.026037648288092897,
      "epoch": 0.4679780559512624,
      "grad_norm": 0.92578125,
      "learning_rate": 7.888888888888889e-07,
      "loss": 0.022882741689682008,
      "mean_token_accuracy": 0.98920454159379,
      "num_tokens": 7826084.0,
      "step": 1930
    },
    {
      "entropy": 0.02522805301673543,
      "epoch": 0.47040281271784923,
      "grad_norm": 1.046875,
      "learning_rate": 6.777777777777779e-07,
      "loss": 0.02458210289478302,
      "mean_token_accuracy": 0.9857954509556294,
      "num_tokens": 7868131.0,
      "step": 1940
    },
    {
      "entropy": 0.022025877615305944,
      "epoch": 0.4728275694844361,
      "grad_norm": 2.1875,
      "learning_rate": 5.666666666666667e-07,
      "loss": 0.021030083298683167,
      "mean_token_accuracy": 0.98920454159379,
      "num_tokens": 7910667.0,
      "step": 1950
    },
    {
      "entropy": 0.022395222801515048,
      "epoch": 0.4752523262510229,
      "grad_norm": 1.5390625,
      "learning_rate": 4.5555555555555563e-07,
      "loss": 0.022039563953876497,
      "mean_token_accuracy": 0.98920454159379,
      "num_tokens": 7951358.0,
      "step": 1960
    },
    {
      "entropy": 0.021953649059105372,
      "epoch": 0.4776770830176098,
      "grad_norm": 0.921875,
      "learning_rate": 3.444444444444445e-07,
      "loss": 0.022923427820205688,
      "mean_token_accuracy": 0.9863636337220669,
      "num_tokens": 7989788.0,
      "step": 1970
    },
    {
      "entropy": 0.02497839363559251,
      "epoch": 0.4801018397841966,
      "grad_norm": 1.3125,
      "learning_rate": 2.3333333333333336e-07,
      "loss": 0.02633815407752991,
      "mean_token_accuracy": 0.9869318149983883,
      "num_tokens": 8034242.0,
      "step": 1980
    },
    {
      "entropy": 0.024830547462852338,
      "epoch": 0.4825265965507835,
      "grad_norm": 1.375,
      "learning_rate": 1.2222222222222225e-07,
      "loss": 0.02468583881855011,
      "mean_token_accuracy": 0.984090904891491,
      "num_tokens": 8077885.0,
      "step": 1990
    },
    {
      "entropy": 0.02569021463914396,
      "epoch": 0.4849513533173703,
      "grad_norm": 1.046875,
      "learning_rate": 1.1111111111111112e-08,
      "loss": 0.02596776783466339,
      "mean_token_accuracy": 0.9857954502105712,
      "num_tokens": 8117574.0,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 2000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.864313676398182e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}